Search

In order to address the challenges of complex process and low precision in traditional device modeling, double hidden layer back propagation neural network (BPNN) are trained using the conjugate gradient (CG) algorithm and the Levenberg–Marquardt (LM) algorithm, the CG‐BPNN and LM‐BPNN models of small signal for gallium arsenide (GaAs) pseudomorphic high electron mobility transistor (pHEMT) are obtained and analyzed here. At first, the scattering parameters (S‐parameters) of GaAs pHEMT are divided into training set and test set randomly. Experimental results show that the CG‐BPNN model is better than another S‐parameters when predicting ImS12 with mean square error (MSE) of 7.6632e‐06, while LM‐BPNN model predicts ImS12 with MSE of 2.4672e‐06. Meanwhile, the MSE of CG‐BPNN model is higher than LM‐BPNN model when predicting all the S‐parameters. In addition, it shows a smaller fluctuation range for the error curve of LM‐BPNN model, which is more stable than the CG‐BPNN model. Therefore, the double hidden layer LM‐BPNN model is the better choice to characterize the small signal of GaAs pHEMT. [ABSTRACT FROM AUTHOR]

Path loss is important for the unmanned aerial vehicle (UAV) placement, trajectory optimization, and power allocation in UAV-aided communications. By considering both the factors of path delay and reflection angle (RA) in the non-line-of-sight (NLoS) paths, a new machine learning-based delay–angle-joint path loss prediction method for UAV mmWave channels is proposed. The one-input back propagation based neural network (BPNN) and two-input BPNN are built to predict the path power for line-of-sight (LoS) case and NLoS case, respectively. Meanwhile, a data acquisition method is developed to obtain massive data set for training the BPNN. According to the geometric information of digital map, a calculation method for path delay and RA is also proposed to drive the BPNN. The proposed method is simulated and analyzed based on ray-tracing (RT) simulated data under a typical urban scenario at 28 GHz. The 2D relationship of power–delay for the LoS case and 3D relationship of power–delay–RA for the NLoS case are obtained through the trained BPNN, which are well consistent with the validation set of RT data and outperform the traditional methods, i.e., 3GPP model and exponential model. Moreover, it's found that the path power rapidly decreases when RA is 65 ∘ – 75 ∘ under the simulation scenario, which could be an important reference for transceiver placement and route planning to reduce the impact of NLoS paths in the UAV channel. [ABSTRACT FROM AUTHOR]

The accurate recognition of unknown terrain by robots plays a vital role in completing walking tasks. Haptic-based terrain recognition methods have their own advantages over visual, audio, and vibration-based terrain recognition methods. With the increasing complexity of robot application scenarios, there are still many technical difficulties in using tactile systems to recognize terrain. For real-time terrain detection technology, how to improve the accuracy of terrain recognition and shorten recognition time based on tactile systems is an urgent problem to be solved. In this paper, the digital elevation model (DEM) geomorphic factor terrain recognition algorithm using probabilistic neural networks (DGFTRA-PNN) based on tactile force-feedback system is proposed to solve this problem. First, the tactile force-feedback system is mounted on the foot ends of the hexapod robot to collect terrain data. Second, the feasibility of using DEM geomorphic factors as terrain features to identify terrain was demonstrated. Third, DGFTRA-PNNs are established based on the different walking gait of the hexapod robot, achieving terrain recognition based on DEM feature values. Finally, the accuracy and speed of DGFTRA-PNN were verified through field experiments. Compared with the terrain recognition algorithm based on back propagation neural network (BP) and extreme learning machine (ELM) algorithm, the DGFTRA-PNN has advantages in terms of terrain recognition accuracy and recognition time: the average recognition accuracy for various types of terrain is 96%, and the calculation time is 0.04 seconds. [ABSTRACT FROM AUTHOR]

For multiphase batch process, the characteristics of process data under various batches differ. Consequently, the soft sensor model built for a particular working condition is inapplicable to other working conditions. Besides, each batch can be divided into several phases whose characteristics are probably different. To address the above problems, a soft sensor model based on phase division and transfer learning strategy is proposed. First, transfer learning strategy is adopted to construct a soft sensor model applicable to various working conditions. Specifically, geodesic flow kernel based on linear local tangent space alignment (LLTSA-GFK) algorithm is designed. By projecting process data to the common manifold subspace, the distribution difference of data between various batches is reduced and the performance of the soft sensor model is enhanced. In addition, sequence-based fuzzy clustering and just-in-time learning (JITL) are adopted to solve the multistage characteristic for batch process. The root-mean-square error (RMSE), coefficient of determination (R 2) , and mean absolute error (MAE) are adopted to compare the conventional soft sensing approach (i.e., partial least-square regression based on JITL, support vector regression, and back propagation neural network) with the proposed approach. The superiority of the proposed model is verified by a fed-batch penicillin fermentation process. [ABSTRACT FROM AUTHOR]

Various faults of photovoltaic (PV) modules inevitably occur in the work process, since PV modules are installed in hostile situation. To obtain the types of failure, a novel fault diagnosis method based on back propagation (BP) neural network with Levenberg-Marquardt (L-M) algorithm for PV modules is proposed. Through the in-depth analysis the output of PV modules under normal and fault conditions, the input variables of the diagnosis model are acquired. The high-speed and real-time fault diagnosis model for PV modules is first designed based on TMS320VC5402 DSP and long-distance wireless fault diagnosis is realized by Zigbee technology. The simulation and experimental results show that the fault diagnosis method for PV modules based on BP network with L-M algorithm can effectively detect four types of fault for PV modules such as open circuit, short circuit, partial shading and abnormal degradation. The numerical results verify the effectiveness and correctness of the proposed method, which can provide a great educational benefit of PV operation technology. [ABSTRACT FROM AUTHOR]

In this article, we analyze the dynamics of the non-linear tumor-immune delayed (TID) model illustrating the interaction among tumor cells and the immune system (cytotoxic T lymphocytes, T helper cells), where the delays portray the times required for molecule formation, cell growth, segregation, and transportation, among other factors by exploiting the knacks of soft computing paradigm utilizing neural networks with back propagation Levenberg Marquardt approach (NNLMA). The governing differential delayed system of non-linear TID, which comprised the densities of the tumor population, cytotoxic T lymphocytes and T helper cells, is represented by non-linear delay ordinary differential equations with three classes. The baseline data is formulated by exploiting the explicit Runge-Kutta method (RKM) by diverting the transmutation rate of Tc to Th of the Tc population, transmutation rate of Tc to Th of the Th population, eradication of tumor cells through Tc cells, eradication of tumor cells through Th cells, Tc cells' natural mortality rate, Th cells' natural mortality rate as well as time delay. The approximated solution of the non-linear TID model is determined by randomly subdividing the formulated data samples for training, testing, as well as validation sets in the network formulation and learning procedures. The strength, reliability, and efficacy of the designed NNLMA for solving non-linear TID model are endorsed by small/negligible absolute errors, error histogram studies, mean squared errors based convergence and close to optimal modeling index for regression measurements. [ABSTRACT FROM AUTHOR]

This paper aims to analyze the influencing factors of supply chain risk management under FinTech innovation. It takes Enterprise A as the research object. To begin with, the basic knowledge of supply chain risk management is introduced, and its five influencing factors are discussed. Furthermore, the relevant theories of the cloud model are analyzed and three kinds of cloud generators are compared. Combined with Back Propagation Fuzzy Neural Network (BPFNN), a supply chain management model is constructed based on the fuzzy cloud model - BPFNN. Eventually, the middle-level and above managers at Enterprise A and experts are provided with questionnaires to assess the various factors impacting supply chain risk. The model's effectiveness is verified through error convergence iii training samples and the fitting of test samples. The results show that the test set sample's mean square error is stable at 0.0137, indicating the model's strong generalization ability. The cloud fuzzy neural network model for financial supply chain collaborative innovation's risk assessment is successfully established. According to the weight analysis, the highest risk factors in the financial industry's supply chain collaborative innovation management are trust risk, capability risk. loss of key customers, and changes in customer demand. This research on the influencing factors of supply chain risk management under collaborative innovation has certain practical significance for the risk management of collaborative innovation. [ABSTRACT FROM AUTHOR]

This study evaluates the effectiveness of artificial Kandelia obovata forests in wave attenuation through physical model experiments conducted in a wave flume. The experiments meticulously replicated real-world hydrodynamic conditions and mangrove movement responses using the principles of gravitational and motion similarity, with a scaled 1:10 model of Kandelia obovata. Our approach included comparative experiments against a 1:100 gradient concrete slope to isolate the effects of seabed friction and flume wall reflections. The wave height was measured using strategically placed wave gauges. The findings indicated that the artificial Kandelia obovata forests significantly attenuated waves, with a decrease in the total attenuation capacity as the water depth increased from 2.75 m to 3.28 m under both regular and irregular waves. The elastic mangrove model with a canopy effect led to a 15% increase in wave attenuation over cylindrical models. Predictive models using multivariate nonlinear regression and back propagation neural networks showed that the latter provided a superior accuracy in estimating wave transmission coefficients [ABSTRACT FROM AUTHOR]

To achieve efficient structural design, it is crucial to reduce the cost of materials while ensuring structural safety. This study proposes an optimization method for design parameters (DPs) in a prestressed steel structure driven by multi-factor coupling. To accomplish this, a numerical proxy model of prestressed steel structures is established with integration of DPs and mechanical parameters (MPs). A data association-parameter analysis-optimization selection system is established. A correlation between DPs and MPs is established using a back propagation (BP) neural network. This correlation provides samples for parameter analysis and optimization selection. MPs are used to characterize the safety of the structure. Based on the safety grade analysis, the key DPs that affect the mechanical properties of the structure are obtained. A mapping function is created to match the MPs and the key DPs. The optimal structural DPs are obtained by setting structural materials as the optimization objective and safety energy as the constraint condition. The theoretical model is applied to an 80-m-span gymnasium and verified with a scale test physical model. The MPs obtained using the proposed method are in good agreement with the experimental results. Compared with the traditional design method, the design cycle can be shortened by more than 90%. Driven by the optimal selection method, a saving of more than 20% can be achieved through reduction of structural material quantities. Moreover, the cross-sectional dimensions of radial cables have a substantial influence on vertical displacement. The initial tension and cross-sectional size of the upper radial cable exhibit the most pronounced impact on the stress distribution in that cable. The initial tension and cross-sectional size of the lower radial cable hold the greatest sway over the stress distribution in that cable. [ABSTRACT FROM AUTHOR]

This study proposes a novel approach to estimate canopy density in Picea Schrenkiana var. Tianschanica forest sub-compartments by integrating optical and radar satellite data. This effort is aimed at enhancing methodologies for forest resource surveys and monitoring, particularly vital for the sustainable development of semi-arid mountainous areas with fragile ecological environments. The study area is the West Tianshan Mountain Nature Reserve in Xinjiang, which is characterized by its unique dominant tree species, Picea Schrenkiana. A total of 411 characteristic factors were extracted from Gaofen-2 (GF-2) sub-meter optical satellite imagery, Gaofen-3 (GF-3) multi-polarization synthetic aperture radar satellite imagery, and digital elevation model (DEM) data. Consequently, 17 characteristic parameters were selected based on their correlation with canopy density data to construct an estimation model. Three distinct models were developed, including a multiple stepwise regression model (a linear approach), a Back Propagation (BP) neural network model (a neural network-based method), and a Cubist model (a decision tree-based technique). The results indicate that combining optical and radar image characteristics significantly enhances accuracy, with an Average Absolute Percentage Precision (AAPP) value improvement in estimation accuracy from 76.50% (with optical image) and 78.50% (with radar image) to 78.66% (with both). Of the three models, the BP neural network model achieved the highest overall accuracy (79.19%). At the sub-component scale, the BP neural network model demonstrated superior accuracy in low canopy density estimation (75.37%), whereas the Cubist model, leveraging radar image characteristics, excelled in medium density estimations (87.46%). Notably, the integrated Cubist model combining optical and radar data achieved the highest accuracy for high canopy density estimation (89.17%). This study highlights the effectiveness of integrating optical and radar data for precise canopy density assessment, contributing significantly to ecological resource monitoring methodologies and environmental assessments. [ABSTRACT FROM AUTHOR]

The terminal settling velocity of microplastics plays a vital role in the physical behavior of microplastics, and is related to the migration and fate of these microplastics in the ocean. At present, the terminal settling velocity is mostly calculated by formulae, which also leads to a fewer studies on the use of machine-learning models to predict its settling velocity in this field. This study fills this gap by studying the prediction of the settling velocity by machine-learning models and compares it with the traditional formula calculation method. This study evaluates three machine-learning models, namely, random forest, linear regression, and the back propagation neural network. The results of this study show that the prediction results of the three machine-learning models are more accurate than those of traditional formula calculations, with an accuracy increase of 12.79% (random forest), 9.3% (linear regression), and 13.92% (back propagation neural network), respectively. At the same time, according to the results of this study, random forest is better than the other models in the mean absolute error and root mean square error evaluation indicators, which are only 0.0036 and 0.0047. This paper proposes three machine-learning methods to prove that the prediction effect of machine learning is much better than traditional formula calculations, thereby improving the shortcomings in this field. At the same time, it also provides reliable data support for studying the migration behavior of microplastics in water bodies. [ABSTRACT FROM AUTHOR]

To address the issue of supplier evaluation and selection in the current intelligent manufacturing industry, a supplier intelligent evaluation and selection model combining bilevel programming is proposed. Firstly, based on existing research results and the current situation of the manufacturing market, a manufacturing supplier evaluation index system is constructed, and the principal component analysis method is used to simplify it; Then, the simplified indicator system is input into the back propagation neural network to build a supplier intelligent evaluation model and achieve supplier intelligent evaluation; A mathematical model for supplier selection problem is constructed based on bilevel programming, and an optimized sparrow search algorithm is put forward to solve it. A supplier intelligent selection model is constructed to achieve intelligent supplier selection. The lab findings denote that the F1 value of the supplier intelligent evaluation model is 0.953, the accuracy is 0.985, the area under the curve is 0.962, and the error of the output comprehensive evaluation value is 0.003; The output accuracy of the supplier intelligent selection model is 0.8. The above results indicate that the model raised in the article has good application effects in supplier evaluation and selection, and has a certain promoting effect on the development of the intelligent manufacturing industry. [ABSTRACT FROM AUTHOR]

To characterize the complex creep behavior of steel slag–asphalt mixture influenced by both stress and temperature, predictive models employing Back Propagation (BP) and Long Short-Term Memory (LSTM) neural networks are described and compared in this paper. Multiple stress repeated creep recovery tests on AC-13 grade steel slag–asphalt mix samples were conducted at different temperatures. The experimental results were processed into a group of independent creep recovery test results, then divided into training and testing datasets. The K-fold cross-validation was applied to the training datasets to fine-tune the hyperparameters of the neural networks effectively. Compared with the experimental curves, both the effects of BP and LSTM models were investigated, and the broad applicability of the models was proven. The performance of the trained LSTM model was observed by a 95% confidence interval around the fit errors, thereby the creep strain intervals for the testing dataset were obtained. The results suggest that the LSTM model had enhanced prediction compared the BP model for creep deformation trends of steel slag–asphalt mixture at various temperatures. Due to the potent generalization strength of artificial intelligence technology, the LSTM model can be further expanded for forecasting road rutting deformations. [ABSTRACT FROM AUTHOR]

Multispectral imaging (MSI) is an emerging technique that ranges from light spectrum of UV–NIR (405–970 nm) used for rapid determination of phenolic contents. The current study focuses on the determination of total phenolic content (TPC) in a fast and non‐invasive way in hot‐air dehydrated Lentinus edodes. The spectral information of MSI has been combined with various chemometrics like partial least squares (PLS), back propagation neural networks (BPNN), and least squares‐support vector machines (LS‐SVM) for the quantitative prediction of phenolic contents. Fresh mushrooms possessed 1.49 and 1.73 GAE g kg−1 TPC in processed samples at 10% moisture content. Comparing models, PLS acquired significant Rp with a lower RMSEP of 0.9980 and 0.1039 and was considered an outstanding algorithm, followed by LS‐SVM of 0.9777 and 0.1417 coefficient of determination and error of prediction data set, respectively. It is concluded that MSI spectroscopy produces highly significant results to determine functional food properties in a rapid and non‐destructive way. The present study will provide a strong platform for the fast online determination of phenolic profile of agricultural commodities. Practical applications: Utilization of spectroscopy for non‐destructive detection of phenolic contents is defined through the combination of chemometrics with spectra of multispectral imaging system and emphasizes control processing of postharvest horticulture produces. Prediction results R2 above 80% describe that these chemometrics are successfully capable of estimating total phenolic contents in control processing for quality preservation. [ABSTRACT FROM AUTHOR]

This study assesses the risk of coal and gas outburst using the theory of rock engineering system, selects seven prediction indices that are significantly related to outburst, and improves on the Back Propagation (BP) neural network technique by applying the Sparrow Search Algorithm (SSA). The improved SSA-BP network is expected to yield discoveries with higher calculation accuracy. Simultaneously, quantitative and qualitative indicators are obtained, and a coal and gas outburst prediction model is constructed based on the cloud drop chart's subordination degree of outburst indicators. The created cloud model predicts the hazard of coal and gas outburst in analyzed coal seams. Eight of the previous nine measurement locations are at risk of medium coal and gas outburst, while only one is at risk of a super big coal and gas outburst. [ABSTRACT FROM AUTHOR]

Because of the improved physical, mechanical and crack–resistant properties, smart cementitious materials have garnered significant attention in civil engineering. However, the method of predicting performance of smart cementitious materials remains a formidable task. To address this issue, this study develops a neural network optimized by particle swarm algorithm, specifically designed for predicting the strength of smart cementitious materials. Particle swarm optimization is used to determine the initial weights and biases of the neural network in this algorithm. Two types of smart cementitious materials, namely 3D printed fiber reinforced concrete and graphene nanoparticles–reinforced cementitious composites, are studied as examples. Utilizing the PSO–BPNN method and data gathered from the existing articles, the predictive models for the mechanical properties of these materials are developed. Five commonly used statistical metrics are applied to evaluate the predictive performance. The results indicate suggest the PSO–BPNN outperforms the traditional back propagation neural network. Thus, a reliable and robust performance predictive model can be built for smart cementitious materials using the proposed approach. [ABSTRACT FROM AUTHOR]

Reducing the influence of significant noise signal components on the obtained raw ECG signal is essential for precise identification of cardiac arrhythmias (CA), which frequently present as irregularities in heart rate or rhythm. Preprocessing is used to remove noise signals and baseline drift from the ECG wave that is recorded using the internet of things (IoT). After that, the denoised signal is subjected to dimensionality reduction and feature extraction. In order to determine whether classification method is more effective in detecting cardiac arrhythmias, this study compares two methods: an adaptive neuro-fuzzy inference system and artificial feed-forward neural networks trained with the back-propagation learning algorithm. An Adaptive Neuro Fuzzy Inference System analyses ICA features obtained by non-parametric power spectral estimates, and an Artificial Neural Network (ANN) classifier uses the ECG signal's morphological and statistical aspects to identify patterns. The creation of artificial feed-forward neural networks provides a rich framework for studying the Back Propagation Algorithm. Sensitivity, specificity, accuracy, and positive predictiveivity are some of the performance characteristics that are thoroughly examined. An overall accuracy of 97.79%, sensitivity of 99.82%, specificity of 99.68%, and positive predictivity of 98.58% were seen in the results of the Artificial Neural Feed Forward Network (ANFFN). The Adaptive Neuro Fuzzy Inference System (ANFIS) outperforms these metrics with an astounding overall accuracy of 99.62%, specificity of 98.63%, and positive predictivity of 99.46%. With a classification accuracy of 99.82%, ANFIS demonstrates to be the most effective classifier for identifying cardiac arrhythmias. [ABSTRACT FROM AUTHOR]

Wireless sensor networks (WSN) have emerged in all digital and smart applications for providing the finest communication range. However, the main problem that often occurs in the WSN is the wide range of energy consumption. So, the present study has focused on developing a novel chimp-based back propagation recurrent neural model for optimizing energy usage. Here, the application that has been considered in this study is multi-input multi-output (MIMO) vehicle communication. Moreover, the planned procedure is executed in the NS2 environment and the performance has been validated in the dual phases that are before applying the chimp fitness and after applying the chimp optimal solution. Finally, the gained metrics have been compared with other conventional models and have earned the finest packet delivery rate and less energy consumption. This finest performance has proved the need for the designed model in the MIMO vehicle application for managing the energy resources by proper cluster-head selection and shortest route finding. [ABSTRACT FROM AUTHOR]

Waste water treatment plants (WWTPs) are considered as energy-intensive industries. A comprehensive assessment of energy efficiency in sewage treatment reveals issues of energy waste, offers insights into the energy consumption structure, fosters optimization of energy management, and enhances overall energy utilization. However, the mathematical modeling for energy efficiency evaluation becomes challenging due to the ambiguity and uncertainty of the parameters regarding energy consumption and various indicators. In this paper, the newly constructed evaluation method was used to quantify the energy efficiency. This method employs dimensional reduction, projecting the numerical values of unit energy consumption for various pollutants onto a lowerdimensional space, and then computing the projected eigenvalues. Larger eigenvalues indicate higher energy efficiency. Based on the evaluation results, the back propagation (BP) neural network is used to simulate the sewage treatment process. The results demonstrated that when the inflow load is small and the concentration of pollutants is high, the energy efficiency of sewage treatment plants performs well. When the inflow load is high and the concentration of pollutants is low, the energy efficiency of sewage treatment plants is poor. Meanwhile, the energy consumption could be decreased by 11.61 %, 14 %, and 15.26 % respectively in different scenarios. [ABSTRACT FROM AUTHOR]

In the low-carbon era, photovoltaic power generation has emerged as a pivotal focal point. The inherent volatility of photovoltaic power generation poses a substantial challenge to the stability of the power grid, making accurate prediction imperative. Based on the integration of a backpropagation (BP) neural network and a genetic algorithm (GA), a prediction model was developed that contained two sub-models: no-rain and no-snow scenarios, and rain and snow scenarios. Through correlation analysis, the primary meteorological factors were identified which were subsequently utilized as inputs alongside historical power generation data. In the sub-model dedicated to rain and snow scenarios, variables such as rainfall and snowfall amounts were incorporated as additional input parameters. The hourly photovoltaic power generation output was served as the model's output. The results indicated that the proposed model effectively ensured accurate forecasts. During no-rain and no-snow weather conditions, the prediction error metrics showcased superior performance: the mean absolute percentage error (MAPE) consistently remained below 13%, meeting the stringent requirement of the power grid's tolerance level below 20%. Moreover, the normalized root mean square error (NRMSE) ranged between 6% and 9%, while the coefficient of determination (R2) exceeded 0.9. These underscored the remarkable prediction accuracy achieved by the model. Under rainy and snowy weather conditions, although MAPE slightly increased to the range of 14% to 20% compared to that of scenarios without rain and snow, it still adhered to the stringent requirement. NRMSE varied between 4.5% and 8%, and R2 remained consistently above 0.9, indicative of satisfactory model performance even in adverse weather conditions. The successful application of the proposed model in predicting hourly photovoltaic power generation output during winter in Henan Province bears significant practical implications for the advancement and integration of renewable energy technologies. [ABSTRACT FROM AUTHOR]

Dual polarization quadrature phase shifting keying (DP-QPSK) modulation format along with coherent receiver helps in increasing the data carrying capability of existing optical networks without any major change in existing transmission infrastructure. The various linear and nonlinear fiber effects, frequency and phase errors are corrected in electrical domain at the receiver end with digital back propagation algorithms (DBP) instead of in-line compensation. In such a case the selection of optimum values of system parameters make the task easier for DBP algorithms. This paper highlights the importance of finding optimum operating point of continuous modulus algorithm (CMA) for better adaptive equalization (AE). The paper also discusses the optical pulse shaping using Gaussian optical band-pass filter to improve the spectral characteristics of DP-QPSK signal. [ABSTRACT FROM AUTHOR]

Introduction: With the development of artificial intelligence and autonomous driving technology, the application of motion recognition in automotive autonomous driving is becoming more and more important. The traditional feature extraction method uses adaptive search hybrid learning and needs to design the feature extraction process manually, which is difficult to meet the recognition requirements in complex environments. Methods: In this paper, a fusion algorithm is proposed to classify the driving characteristics through time-frequency analysis, and perform backpropagation operation in artificial neural network to improve the convergence speed of the algorithm. The performance analysis experiments of the study were carried out on Autov data sets, and the results were compared with those of the other three algorithms. Results: When the vehicle action coefficient is 227, the judgment accuracy of the four algorithms is 0.98, 0.94, 0.93 and 0.95, respectively, indicating that the fusion algorithm is stable. When the road sample is 547, the vehicle driving ability of the fusion algorithm is 4.7, which is the best performance among the four algorithms, indicating that the fusion algorithm has strong adaptability. Discussion: The results show that the fusion algorithm has practical significance in improving the autonomous operation ability of autonomous vehicles, reducing the frequency of vehicle accidents during driving, and contributing to the development of production, life and society. [ABSTRACT FROM AUTHOR]

Multiple microgrids interconnect to form a microgrid cluster. To fully exploit the comprehensive benefits of the microgrid cluster, it is imperative to optimize dispatch based on the matching degree between the sources and loads of each microgrid. The power of distributed energy sources such as wind and photovoltaic systems and the sensitive loads in microgrids is related to the regional weather characteristics. Given the relatively small geographical scope of microgrid areas and the fact that distributed energy sources and loads within the grid share the same weather characteristics, simultaneous ultra-short-term forecasting of power for both sources and loads is essential in the same environmental context. Firstly, the introduction of the multi-variable uniform information coefficient (MV-UIC) is proposed for extracting the correlation between weather characteristics and the sequences of source and load power. Subsequently, the application of factor analysis (FA) is introduced to reduce the dimensionality of input feature variables. Drawing inspiration from the concept of combination forecasting models, a combined forecasting model based on Error Back Propagation Training (BP), Long Short-Term Memory (LSTM), and Bidirectional Long Short-Term Memory Neural Network (BiLSTM) is constructed. This model is established on the MV-UIC-FA foundation for the joint ultra-short-term forecasting of source and load power in microgrids. Simulation is conducted using the DTU 7K 47-bus system as an example to analyze the accuracy, applicability, and effectiveness of the proposed joint forecasting method for sources and loads. [ABSTRACT FROM AUTHOR]

Traditional finite element analysis methods have the problem of expensive and unstable band structure diagram calculation when generating phononic crystals. Therefore, this study combines back propagation neural networks with ocean predator algorithms to optimize the geometric structure of phononic crystals. The results show that the coefficient of determination for the predicted bandgap width and lower bound of Model 3 is 1.00, which is better than the comparison model. However, Models 2 and 5 have poor predictive performance for bandgap width due to overfitting during actual training. Therefore, after using the ocean predator algorithm for hyperparameter adjustment, it is found that the maximum number of failures in the validation set of Model 5 is 30, the number of hidden layer nodes is 30, and after 500 experiments, the average error of the bandgap width is 5.26%, and the average error of the lower bound of the bandgap is 1.33%. The average error of the bandgap width after hyperparameter adjustment is 4.89%, and the average error of the lower bound of the bandgap is 1.21%, both of which are effectively reduced and better than the comparison model. Overall, the combination model has high computational efficiency and stability in phononic crystal optimization and can be practically applied in the design and optimization of phononic crystal geometric structures. [ABSTRACT FROM AUTHOR]

This paper proposes a new method to improve the training efficiency of deep convolutional neural networks. During training, the method evaluates scores to measure how much each layer's parameters change and whether the layer will continue learning or not. Based on these scores, the network is scaled down such that the number of parameters to be learned is reduced, yielding a speed-up in training. Unlike state-of-the-art methods that try to compress the network to be used in the inference phase or to limit the number of operations performed in the back-propagation phase, the proposed method is novel in that it focuses on reducing the number of operations performed by the network in the forward propagation during training. The proposed training strategy has been validated on two widely used architecture families: VGG and ResNet. Experiments on MNIST, CIFAR-10 and Imagenette show that, with the proposed method, the training time of the models is more than halved without significantly impacting accuracy. The FLOPs reduction in the forward propagation during training ranges from 17.83% for VGG-11 to 83.74% for ResNet-152. As for the accuracy, the impact depends on the depth of the model and the decrease is between 0.26% and 2.38% for VGGs and between 0.4 and 3.2% for ResNets. These results demonstrate the effectiveness of the proposed technique in speeding up learning of CNNs. The technique will be especially useful in applications where fine-tuning or online training of convolutional models is required, for instance because data arrive sequentially. [ABSTRACT FROM AUTHOR]

In order to reduce the position errors of the Global Positioning System/Strapdown Inertial Navigation System (GPS/SINS) integrated navigation system during GPS denial, this paper proposes a method based on the Particle Swarm Optimization–Back Propagation Neural Network (PSO-BPNN) to replace the GPS for positioning. The model relates the position information, velocity information, attitude information output by the SINS, and the navigation time to the position errors between the position information output by the SINS and the actual position information. The performance of the model is compared with the BPNN through an actual ship experiment. The results show that the PSO-BPNN can obviously reduce the position errors in the case of GPS signal denial. [ABSTRACT FROM AUTHOR]

A stochastic model updating framework is proposed in this work to address the problem of uncertain model calibration. This framework includes an effective uncertainty quantification metric of sub-interval similarity to measure the discrepancy between model predictions and experimental observations. A back propagation neural network is employed as a surrogate model for finite element method models, and a sparrow search algorithm is introduced as an optimization operator. Two typical numerical examples of a 3-degree-of-freedom mass-spring system and a satellite finite element model have been presented to demonstrate the feasibility and the effectiveness of the proposed stochastic model updating algorithm. [ABSTRACT FROM AUTHOR]

Interior design is a complex system that includes many elements, such as user needs, spatial structure, material selection, color matching, lighting layout, etc. The relationships between them are complex and intricate. How to design space according to user needs is one of the most challenging issues in interior design. A distributed 3D interior design system based on artificial intelligence algorithms was adopted to address this issue. The system divided indoor spatial information into several subsets. It assigned subsets to various professional designers based on user needs, and generated multiple design schemes through artificial intelligence algorithms, which were then screened, modified, and optimized by each professional designer. In this way, designers from different professional backgrounds can personalized design the space according to their own understanding, thereby avoiding issues such as information fragmentation and low efficiency in traditional interior design. The results showed that the distributed 3D interior design system based on artificial intelligence algorithms can improve the overall performance from an average of 82.68% to 94.94%, effectively avoiding the problem of information fragmentation. [ABSTRACT FROM AUTHOR]

Geological landslide disasters significantly threaten the safety of people's lives and property. Landslides are a significant threat in Zayu County, Tibet, resulting in numerous geological disasters, including the 1950 earthquake that caused significant casualties and river blockages. More recent landslides have caused substantial economic losses and infrastructure damage, posing ongoing risks to the local population and their property. Landslide hazard assessment is a critical task in geological disaster prevention and mitigation. This study applied the Inception model to assess landslide hazard in the Zayu area. The Inception model excels at capturing multi-scale features efficiently through its architecture. Fifteen disaster-causing factors were selected as the primary indicators for landslide susceptibility assessment. On this basis, the Inception model was used for landslide susceptibility assessment. Combined with daily precipitation data in the Zayu area, the landslide hazard assessment of the "25 April 2010, heavy rainstorm in Zayu, Tibet" was completed. Back Propagation Neural Network (BPNN), Residual Neural Network (ResNet), Convolutional Neural Network (CNN), and Visual Geometry Group-16 (VGG-16) were introduced for comparison of the fitting effects, and SHapley Additive exPlanations (SHAP) was used for interpretability analysis. The comparative experimental results show that the Inception model performed best in landslide susceptibility assessment and is feasible in practical use. The results also show that the most critical factors in the model were topographic wetness index (TWI), normalized difference water index (NDWI), and road density. This study is significant for assessing landslide hazard in geological landslide disaster prevention and mitigation. It provides a reference for further research and response to similar disasters. [ABSTRACT FROM AUTHOR]

Cosmic radiation exposure is one of the important health concerns for aircrews. In this work, we constructed a back propagation neural network model for the real-time and rapid assessment of cosmic radiation exposure to the public in aviation. The multi-dimensional dataset for this neural network was created from modeling the process of cosmic ray transportation in magnetic field by geomagnetic cutoff rigidity method and air shower simulation by a Monte Carlo based Geant4 code. The dataset was characterized by parameters including cosmic ray energy spectrum, K p-index, coordinated universal time, altitude, latitude, and longitude. The effective dose and dose rate was finally converted from the particle fluxes at flight position by the neural network. This work shows a good agreement with other models from International Civil Aviation Organization. It is also illustrated that the effective dose rate by galactic cosmic ray is <10 μSv h−1 and the value during ground level enhancement (GLE) 42 is 4 ~ 10 times larger on the routes calculated in this work. In GLE 69, the effective dose rate reaches several mSv h−1 in the polar region. Based on this model, a real-time warning system is achieved. [ABSTRACT FROM AUTHOR]

Using the high-rate phase and amplitude scintillation data from FORMOSA7/COSMIC two mission and back-propagation method, we geolocate plasma irregularities that cause scintillations. The results of geolocation are compared with the NASA GOLD UV image data of plasma bubbles. The root mean square of the zonal difference between estimated locations of plasma irregularities and plasma bubbles are about 1.5° and for single intersection cases 0.5° in the magnetic longitude. The geolocation data provide more accurate scintillation location around the globe compared to assigning to the tangent point and is valuable space weather product, which will be routinely available for public use. [ABSTRACT FROM AUTHOR]

The spiking convolutional neural network (SCNN) is a kind of spiking neural network (SNN) with high accuracy for visual tasks and power efficiency on neuromorphic hardware, which is attractive for edge applications. However, it is challenging to implement SCNNs on resource-constrained edge devices because of the large number of convolutional operations and membrane potential (Vm) storage needed. Previous works have focused on timestep reduction, network pruning, and network quantization to realize SCNN implementation on edge devices. However, they overlooked similarities between spiking feature maps (SFmaps), which contain significant redundancy and cause unnecessary computation and storage. This work proposes a dual-threshold spiking convolutional neural network (DT-SCNN) to decrease the number of operations and memory access by utilizing similarities between SFmaps. The DT-SCNN employs dual firing thresholds to derive two similar SFmaps from one Vm map, reducing the number of convolutional operations and decreasing the volume of Vms and convolutional weights by half. We propose a variant spatio-temporal back propagation (STBP) training method with a two-stage strategy to train DT-SCNNs to decrease the inference timestep to 1. The experimental results show that the dual-thresholds mechanism achieves a 50% reduction in operations and data storage for the convolutional layers compared to conventional SCNNs while achieving not more than a 0.4% accuracy loss on the CIFAR10, MNIST, and FashionMNIST datasets. Due to the lightweight network and single timestep inference, the DT-SCNN has the least number of operations compared to previous works, paving the way for low-latency and power-efficient edge applications. [ABSTRACT FROM AUTHOR]

The hot spot temperature of oil-immersed transformer winding is an important factor affecting the aging of material insulation. In this paper, a magnetic field simulation model is established based on the electrical and structural parameters of the oil-immersed transformer, and the loss distribution characteristics of each wall of the transformer core, winding and fuel tank are accurately calculated by using the finite element simulation software. The simulation model of transformer fluid-thermal field is established, the simulation results of transformer thermal field are obtained, and the temperature distribution of oil-immersed transformer core and winding and the flow velocity around it are obtained. According to the simulation results of thermal field, the characteristic temperature measuring points with strong correlation between tank wall and winding temperature were determined. The inversion models of tank wall and winding hot spot temperature were established by using the support vector regression and back propagation neural network algorithm, respectively by central composite design method. The results show that the correlation coefficient of support vector regression algorithm in predicting winding hot spot temperature reaches 0.98, and the relative error between the model predicted value and the real value is less than 8%, which is more accurate than back propagation neural network. The aforementioned research provides the theoretical basis and technical support for real-time monitoring of oil-immersed transformer winding hot spot temperature. [ABSTRACT FROM AUTHOR]

Geophysical logging plays a very important role in reservoir evaluation. In the actual production process, some logging data are often missing due to well wall collapse and instrument failure. Therefore, this paper proposes a logging reconstruction method based on improved sand cat swarm optimization (ISCSO) and a temporal convolutional network (TCN) and bidirectional gated recurrent unit network with attention mechanism (BiGRU-AM). The ISCSO-TCN-BiGRU-AM can process both past and future states efficiently, thereby extracting valuable deterioration information from logging data. Firstly, the sand cat swarm optimization (SCSO) improved by the variable spiral strategy and sparrow warning mechanism is introduced. Secondly, the ISCSO's performance is evaluated using the CEC–2022 functions and the Wilcoxon test, and the findings demonstrate that the ISCSO outperforms the rival algorithms. Finally, the logging reconstruction method based on the ISCSO-TCN-BiGRU-AM is obtained. The results are compared with the competing models, including the back propagation neural network (BPNN), GRU, and BiGRU-AM. The results show that the ISCSO-TCN-BiGRU-AM has the best performance, which verifies its high accuracy and feasibility for the missing logging reconstruction. [ABSTRACT FROM AUTHOR]

The high-speed railway subgrade compaction quality is controlled by the compaction degree (K), with the maximum dry density (ρdmax) serving as a crucial indicator for its calculation. The current mechanisms and methods for determining the ρdmax still suffer from uncertainties, inefficiencies, and lack of intelligence. These deficiencies can lead to insufficient assessments for the high-speed railway subgrade compaction quality, further impacting the operational safety of high-speed railways. In this paper, a novel method for full-section assessment of high-speed railway subgrade compaction quality based on ML-interval prediction theory is proposed. Firstly, based on indoor vibration compaction tests, a method for determining the ρdmax based on the dynamic stiffness Krb turning point is proposed. Secondly, the Pso-OptimalML-Adaboost (POA) model for predicting ρdmax is determined based on three typical machine learning (ML) algorithms, which are back propagation neural network (BPNN), support vector regression (SVR), and random forest (RF). Thirdly, the interval prediction theory is introduced to quantify the uncertainty in ρdmax prediction. Finally, based on the Bootstrap-POA-ANN interval prediction model and spatial interpolation algorithms, the interval distribution of ρdmax across the full-section can be determined, and a model for full-section assessment of compaction quality is developed based on the compaction standard (95%). Moreover, the proposed method is applied to determine the optimal compaction thicknesses (H0), within the station subgrade test section in the southwest region. The results indicate that: (1) The PSO-BPNN-AdaBoost model performs better in the accuracy and error metrics, which is selected as the POA model for predicting ρdmax. (2) The Bootstrap-POA-ANN interval prediction model for ρdmax can construct clear and reliable prediction intervals. (3) The model for full-section assessment of compaction quality can provide the full-section distribution interval for K. Comparing the H0 of 50~60 cm and 60~70 cm, the compaction quality is better with the H0 of 40~50 cm. The research findings can provide effective techniques for assessing the compaction quality of high-speed railway subgrades. [ABSTRACT FROM AUTHOR]

Permanent engravings on contact lenses provide information about the manufacturing process and lens positioning when they are placed on the eye. The inspection of their morphological characteristics is important, since they can affect the user's comfort and deposit adhesion. Therefore, an inverted wavefront holoscope (a lensless microscope based on Gabor's principle of in-line digital holography) is explored for the characterization of the permanent marks of soft contact lenses. The device, based on an in-line transmission configuration, uses a partially coherent laser source to illuminate the soft contact lens placed in a cuvette filled with a saline solution for lens preservation. Holograms were recorded on a digital sensor and reconstructed by back propagation to the image plane based on the angular spectrum method. In addition, a phase-retrieval algorithm was used to enhance the quality of the recovered images. The instrument was experimentally validated through a calibration process in terms of spatial resolution and thickness estimation, showing values that perfectly agree with those that were theoretically expected. Finally, phase maps of different engravings for three commercial soft contact lenses were successfully reconstructed, validating the inverted wavefront holoscope as a potential instrument for the characterization of the permanent marks of soft contact lenses. To improve the final image quality of reconstructions, the geometry of lenses should be considered to avoid induced aberration effects. [ABSTRACT FROM AUTHOR]

To investigate the pattern recognition of complex defect types in XLPE (cross-linked polyethylene) cable partial discharges and analyze the effectiveness of identifying partial discharge signal patterns, this study employs the variational mode decomposition (VMD) algorithm alongside entropy theories such as power spectrum entropy, fuzzy entropy, and permutation entropy for feature extraction from partial discharge signals of composite insulation defects. The mean power spectrum entropy (PS), mean fuzzy entropy (FU), mean permutation entropy (PE), as well as the permutation entropy values of IMF2 and IMF13 (Pe) are selected as the characteristic quantities for four categories of partial discharge signals associated with composite defects. Six hundred samples are selected from the partial discharge signals of each type of compound defect, amounting to a total of 2400 samples for the four types of compound defects combined. Each sample comprises five feature values, which are compiled into a dataset. A Snake Optimization Algorithm-optimized Support Vector Machine (SO-SVM) model is designed and trained, using the extracted features from cable partial discharge datasets as case examples for recognizing cable partial discharge signals. The identification outcomes from the SO-SVM model are then compared with those from conventional learning models. The results demonstrate that for partial discharge signals of XLPE cable composite insulation defects, the SO-SVM model yields better identification results than traditional learning models. In terms of recognition accuracy, for scratch and water ingress defects, SO-SVM improves by 14.00% over BP (Back Propagation) neural networks, by 5.66% over GA-BP (Genetic Algorithm–Back Propagation), and by 12.50% over SVM (support vector machine). For defects involving metal impurities and scratches, SO-SVM improves by 13.39% over BP, 9.34% over GA-BP, and 12.56% over SVM. For defects with metal impurities and water ingress, SO-SVM shows enhancements of 13.80% over BP, 9.47% over GA-BP, and 13.97% over SVM. Lastly, for defects combining metal impurities, water ingress, and scratches, SO-SVM registers increases of 11.90% over BP, 9.59% over GA-BP, and 12.05% over SVM. [ABSTRACT FROM AUTHOR]

This study presents a new intelligent computing method based on the Levenberg–Marquardt back-propagation-based supervised neural network (LMB-SNN) for the flow of Oldroyd 4-constant fluid in the forward roll coating process (OFC-FRCP). The mathematical formulations for OFC-FRCP are converted into ordinary differential equations using lubrication approximation theory (LAT) and suitable dimensionless parameters. The analytic expressions for velocity distribution, pressure and pressure gradients have been obtained using Adomian's Decomposition Method (ADM) for several situations. The fluid's separation point and coating thickness are then calculated using integration. The reference dataset for LMB-SNN has been obtained by ADM for multiple scenarios by adjusting various physical variables. The neural network's training, testing and validation are carried out in parallel to minimize the mean-squared error function (MSEF). The proposed LMB-SNN's effectiveness was confirmed by analyzing the MSEF, error histograms and regression plots. The precision of the method is verified by the close agreement between numerical outputs and dataset values. The intelligent numerical results computed using ADM and LMB-SNN are presented in various graphs and tables. The method's accuracy is demonstrated by the numerical outputs closest to the built and dataset values at levels between 1 0 − 9 and 1 0 − 1 2 . [ABSTRACT FROM AUTHOR]

The Huanglong Scenic and Historic Interest Area in China, a UNESCO World Heritage Site, is famous for its large-scale, diverse, intricately structured, and brightly colored surface travertine landscapes. However, severe degradation of the Huanglong travertine formations, such as blackening and algae erosion, has occurred in recent years, necessitating monitoring and identification. We collected hyperspectral reflectance data of the travertine formations in different states and bare ground using a ground-based hyperspectral radiometer (PSR-2500) from ASD company. After conducting a correlation analysis between the hyperspectral reflectance data and the travertine formations, we identified healthy travertine formations, blackened travertine formations, travertine formations affected by algae erosion, and bare ground. The Siamese network method was employed to generate data labels, and the spectral features of the travertine formations were extracted by combining the sensitive bands with pre-processed and reduced data. The PSO-BPNN classifier was developed by optimizing the back propagation neural network (BPNN) using the particle swarm optimization algorithm (PSO). To verify the effectiveness of PSOBPNN in accurately distinguishing different states of travertine formations, we compared its performance with that of BPNN using three performance indices. Finally, the proposed method was applied to the real-world hyperspectral image data collected by the Micro-Hyperspectral imaging instrument to classify the travertine formations in different states and bare ground. The test set demonstrated good overall performance, with an average overall accuracy (OA) of 0.93, F1-score of 0.92, and Kappa coefficient of 0.97. [ABSTRACT FROM AUTHOR]

China has resolved its overall regional poverty in 2020 by attaining moderate societal prosperity. The country has entered a new development stage designed to achieve its second centenary goal. However, ecological fragility and risk susceptibility have increased the risk of returning to ecological poverty. In this paper, the Liupan Mountain Region of China was used as a case study, and the counties were used as the scale to reveal the spatiotempora differentiation and influcing factors of the risk of returning to poverty in study area. The indicator data for returning to ecological poverty from 2011–2020 were collected and summarized in three dimensions: ecological, economic and social. The autoregressive integrated moving average model (ARIMA) time series and exponential smoothing method (ES) were used to predict the multidimensional indicators of returning to ecological poverty for 61 counties (districts) in the Liupan Mountain Region for 2021–2030. The back propagation neural network (BPNN) and geographic information system (GIS) were used to generate the spatial distribution and time variation for the index of the risk of returning to ecological poverty (RREP index). The results show that 1) ecological factors were the main factors in the risk of returning to ecological poverty in Liupan Mountain Region. 2) The RREP index for the 61 counties (districts) exhibited a downward trend from 2021–2030. The RREP index declined more in medium- and high-risk areas than in low-risk areas. From 2021 to 2025, the RREP index exhibited a slight downward trend. From 2026 to 2030, the RREP index was expected to decline faster, especially from 2029–2030. 3) Based on the RREP index, it can be roughly divided into three types, namely, the high-risk areas, the medium-risk areas, and the low-risk areas. The natural resource conditions in low-risk areas of returning to ecological poverty, were better than those in medium- and high-risk areas. [ABSTRACT FROM AUTHOR]

In order to improve the calculation efficiency and reduce the uncertainty in the reliability evaluation process of flat car underframe, a new method based on an improved genetic simulated annealing algorithm-back propagation neural network (GSA-BP) and probability box is proposed. Firstly, a certain type of flat car underframe is taken as the research object, genetic evolution algorithm (GA) and simulated annealing algorithm (SA) are utilized to optimize the weights and thresholds of the GSA-BP neural network and obtain the best initial parameter values. Applying the values, the BP neural network is trained, and the improved GSA-BP neural network is established which is verified by test functions. Secondly, the central plate sub-model of the flat car underframe is obtained by sub-model technology and the correctness of its boundary conditions is tested. On this basis, the static strength analysis of the sub-model is carried out. Finally, the improved GSA-BP neural network and probability box are utilized to evaluate the reliability of the sub-model of flat car underframe, and the Monte Carlo method is utilized to verify it. The results show that the proposed method not only improves the accuracy of reliability evaluation but also shortens the calculation time from 10 hours to 6 hours. The calculation efficiency is increased by 40%, which further verifies its superiority and feasibility in structural reliability evaluation. [ABSTRACT FROM AUTHOR]

Accurate prediction of lithium ion (li-ion) battery capacity is of great significance to battery health status management. In this paper, the different discharge time corresponding to the equal voltage interval is taken as the health factor. Three highly correlated health factors are extracted from the battery discharge curve, and the Back Propagation neural network (BPNN) optimized by a genetic algorithm (GA) is used to estimate the battery capacity accurately and quickly. Firstly, health factors related highly to battery capacity from the battery cycle life test are extracted, and the selected health factors are analyzed using the Spearman correlation coefficient and Pearson correlation coefficient. Secondly, this paper analyzes the prediction effect of different combinations of selected health factors using BPNN optimized by a genetic algorithm. Then, to verify the superiority of the proposed optimization algorithm, different optimization algorithms are used to adjust and optimize the parameters of BPNN automatically, and the experimental data are used for comparative analysis. Finally, the GA-BP is compared with other common battery capacity prediction methods. The results show that the GA-BP neural network prediction model can accurately and effectively predict the capacity of li-ion battery when using selected health factors. [ABSTRACT FROM AUTHOR]

The study area was a derelict electroplating factory in Huzhou of Zhejiang Province, and the health risk of soil heavy metal pollution in the region by means of the Nemero index method and human health risk assessment models was undertaken to analyze the characteristics of the spatial distribution of soil heavy metal concentration, in order to provide theoretical support for environmental pollution recovery of electroplating plant demolition. A BP (Back Propagation) neural network prediction model was constructed by MATLAB to forecast the human health risk of the heavy metal contamination in deep soil. At the same time, the inverse distance weight interpolation method in ArcGIS software was used to draw the distribution map of the predicted health risk. The results revealed that the studied area was largely contaminated by Cr and Ni. The carcinogenic risk of Cr in buried underground storage tanks was 0.35, which was a high risk for cancer. The excessive heavy metals in the surface soil were mainly concentrated in underground storage tank burial, galvanizing and nickel-plating workshops, and temporary waste accumulation areas, which may be caused by insufficient anti-seepage measures on the electroplating operation ground and improper disposal of electroplating waste. The constructed BP neural network prediction model showed that Cr, As, and Ni in the plot still have a carcinogenic risk in the deep soil at a depth of 2.0 ~2.5 m; Ni, Cr, As, Zn, Cu, and other elements no longer have non carcinogenic risks in the soil at a depth of 2.0 ~2.5 m. [ABSTRACT FROM AUTHOR]

Corn is easily infected by Aspergillus ochraceus, which is severely hazardous to human health. As the traditional methods for Aspergillus ochraceus detection are time-consuming and destructive, it is necessary to develop a rapid and non-destructive method for monitoring the growth of Aspergillus ochraceus in corn during storage. In this study, multispectral imaging technology combined with chemometric methods was used to obtain the optimal model for predicting the count of Aspergillus ochraceus quantitively and classifying the infection degree qualitatively. The results showed that compared with partial least square (PLS) and least square-support vector machine (LS-SVM), back propagation neural network (BPNN) showed the best prediction performance with correlation coefficient of prediction (Ip) value of 0.9494, and the lowest root-mean-square error of calibration (RMSEC) and root-mean-square error of prediction (RMSEP) values of 2.6693 and 2.2743 CFU/g, respectively. In addition, for the classification experiment of the infective degree, BPNN was also the best prediction model with the accuracy of calibration (A c) and the accuracy of prediction ( A p) both reached 100%. The results indicated that multispectral imaging combined with chemometric methods provided a promising technique to evaluate the infection of Aspergillus ochraceus in corn. [ABSTRACT FROM AUTHOR]

Efficient acquisition of crop leaf moisture information holds significant importance for agricultural production. This information provides farmers with accurate data foundations, enabling them to implement timely and effective irrigation management strategies, thereby maximizing crop growth efficiency and yield. In this study, unmanned aerial vehicle (UAV) multispectral technology was employed. Through two consecutive years of field experiments (2021–2022), soybean (Glycine max L.) leaf moisture data and corresponding UAV multispectral images were collected. Vegetation indices, canopy texture features, and randomly extracted texture indices in combination, which exhibited strong correlations with previous studies and crop parameters, were established. By analyzing the correlation between these parameters and soybean leaf moisture, parameters with significantly correlated coefficients (p < 0.05) were selected as input variables for the model (combination 1: vegetation indices; combination 2: texture features; combination 3: randomly extracted texture indices in combination; combination 4: combination of vegetation indices, texture features, and randomly extracted texture indices). Subsequently, extreme learning machine (ELM), extreme gradient boosting (XGBoost), and back propagation neural network (BPNN) were utilized to model the leaf moisture content. The results indicated that most vegetation indices exhibited higher correlation coefficients with soybean leaf moisture compared with texture features, while randomly extracted texture indices could enhance the correlation with soybean leaf moisture to some extent. RDTI, the random combination texture index, showed the highest correlation coefficient with leaf moisture at 0.683, with the texture combination being Variance1 and Correlation5. When combination 4 (combination of vegetation indices, texture features, and randomly extracted texture indices) was utilized as the input and the XGBoost model was employed for soybean leaf moisture monitoring, the highest level was achieved in this study. The coefficient of determination (R2) of the estimation model validation set reached 0.816, with a root-mean-square error (RMSE) of 1.404 and a mean relative error (MRE) of 1.934%. This study provides a foundation for UAV multispectral monitoring of soybean leaf moisture, offering valuable insights for rapid assessment of crop growth. [ABSTRACT FROM AUTHOR]

Artificial neural networking (ANN) based models are being increasingly and effectively used in the prediction and estimation of response variables during various food processing applications. We demonstrated the feasibility of a feed forward back propagation ANN model to estimate ascorbic acid content in fresh amla juice that was ohmic heating assisted vacuum evaporated (OHVC) with/without ultrasonication (US) from color attributes (L*, a*, b* and ΔE) during 4 weeks of ambient storage. There was a positive effect of synergistic processing of OHVC and US on the quality of stored amla juice with the fresh untreated sample witnessing ascorbic acid and total phenolic degradation of up to 60% and 14% higher than the OHVC + US treated samples respectively. At the end of the storage period, the color change was lowest in OHVC samples stored in glass containers (ΔE = 5.00) in comparison to untreated samples stored in PET containers (ΔE = 9.61). The results further suggested that coefficient of determination (R2) for the fitted ANN model was approximately 0.93 with a mean square error of 0.15, which was 20% higher and 66% lower than that obtained using multiple linear regression model, respectively. Furthermore, a number of training algorithms were tested out of which Levenberg–Marquardt algorithm showed the highest efficiency in model prediction. Overall, the quality in OHVC and US processed juice was retained better in comparison to control and the fitted ANN model effectively predicted ascorbic acid content from input color parameters. The results provided can be helpful in the non-destructive evaluation of ascorbic acid content in stored amla juice using AI mediated modelling. [ABSTRACT FROM AUTHOR]

Given that the conventional back propagation neural network (BPNN) easily falls into the local optimal solutions, resulting in poor prediction accuracy, an improved BPNN based on the differential evolution and grey wolf optimizer (DEGWO) is proposed, the so-called DEGWO-BPNN. The prediction of the water-conducting fracture zone (WCFZ) height is significant for mine safety operations. A total of 104 sample data are trained and 25 sample data are tested to identify the optimal prediction model. Five evaluation indexes are selected to assess the prediction performance of the models quantitatively. Finally, the DEGWO-BPNN model is applied to a specific engineering case. The main conclusions are as follows: (1) Mining height, mining depth, coal seam dip, panel width, and ratio of hard rock as the main factors affecting the WCFZ height are selected. The topology structure of the model is defined as '5-12-1'; (2) the bias between the predicted value and the actual value of the training samples is smaller with an average error of 2.39. Test samples further validate the prediction precision through evaluation indexes. The values of MAE, RMSE, MAPE, and R2 are 2.3952, 3.4674, 5.3148%, and 0.99077, respectively. The prediction accuracy is 94.6852%; (3) 'Mining Code', MLR, BPNN, and GWO-BPNN models are treated as the comparison groups. The comparative analysis shows that the prediction performance of 'Mining Code' is the worst, while that of DEGWO-BPNN is the best, and it outperforms other algorithms and statistical approaches; (4) the prediction of WCFZ height in the 11601 panel is in line with the actual value. The prediction error of the DEGWO-BPNN model is lower than that of the comparison models. As such, the DEGWO-BPNN model can be well applied to the prediction of WCFZ height and is suitable for coal mines with different regional geological conditions. It can provide a valuable reference for mine safety operations. [ABSTRACT FROM AUTHOR]

Deep and ultra-deep shale gas resources have great potential, but well drilling faces many challenges. The Polycrystalline Diamond Compact (PDC) bit has become the primary rock-breaking instrument for oil and gas drilling. Reasonable bit structure designs can promote rock-breaking efficiency and extend service life. In this study, reverse modeling technology is used to analyze the structural characteristics of PDC bits collected in the field, and the influence of the structural characteristics of the bit at a specific interval on the rate of penetration (ROP) and drill footage is investigated using the Spearman rank correlation coefficient method. The number of blades, cutting angle of the cutters, crown rotation radius, internal cone angle, and diameter of the cutters are discovered to be the main structural characteristics that affect the ROP and footage of the bits, and the degree of influence varies depending on the formation conditions. The number of blades, crown rotation radius, inner cone angle, and cutting angle of the cutters have a significant impact on the ROP, whereas blade thickness, gauge length, gauge width, nozzle equivalent diameter have a significant impact on the bit footage. In addition, a back propagation (BP) neural network is utilized to build a prediction model of bit footage and ROP over a certain interval based on the structural characteristics of the bit. The goodness of fit of the model is greater than 85%, and its accuracy is high. Based on the usage of the bit, the evaluation and prediction of the bit can provide a reference for the structural design and optimization of the bit in a specific interval, guide the bit selection work, rationally plan the drilling operation, and reduce the drilling cost. [ABSTRACT FROM AUTHOR]

Centrifugal chillers have been widely used in medium- and large-scale air conditioning projects. However, equipment running with faults will result in additional energy consumption. Meanwhile, it is difficult to diagnose the minor faults of the equipment. Therefore, the Extreme Gradient Boost (XGBoost) algorithm was used to solve the above problem in this article. The ASHRAE RP-1043 dataset was employed for research, utilizing the feature splitting principle of XGBoost to reduce the data dimension to 23 dimensions. Subsequently, the five important parameters of the XGBoost algorithm were optimized using Multi-swarm Cooperative Particle Swarm Optimization (MSPSO). The minor fault diagnosis model, MSPSO-XGBoost, was established. The results show that the ability of the proposed MSPSO-XGBoost model to diagnose eight different states is uniform, and the diagnostic accuracy of the model reaches 99.67%. The accuracy rate is significantly improved compared to that of the support vector machine (SVM) and back propagation neural network (BPNN) diagnostic models. [ABSTRACT FROM AUTHOR]

Conventional methods for calculating tension currently suffer from an excessive simplification of boundary conditions and a vague definition of effective cable length, both of which cause inaccurate cable tension calculations. Therefore, this study utilizes bridge field data to establish a BP neural network for tension prediction, with design cable length, line density, and frequency as the input parameters and with cable tension as the output parameter. After disregarding the selection of effective cable length and innovatively integrating the particle swarm optimization–back propagation (PSO-BP) neural network for tension prediction, it is found that the MAPE between the predicted results of the BP neural network and the actual tension values is 7.93%. After optimization using the particle swarm optimization algorithm, the mean absolute percentage error (MAPE) of the neural network prediction is reduced to 2.78%. Both of these values significantly outperform those obtained from the theoretical equations of string vibration. Moreover, the MAPE of PSO-BP also surpasses that of the optimized calculation formulas in the literature. Utilizing the PSO-BP neural network for tension prediction avoids inaccuracies in tension calculation caused by an excessive simplification of boundary conditions and a vague definition of effective cable length; thus, it possesses certain engineering practical value. [ABSTRACT FROM AUTHOR]