Search

A multi-sensor information fusion algorithm based on fault-tolerant Kalman filter is proposed in phase-sensitive optical time-domain reflectometer (Φ-OTDR) system, for achieving fading-free distributed vibration sensing. Firstly, a fault-tolerant dual-core complementary array model is designed. The Rayleigh scattering signal denoising, and vibration existence judgment of localization points are carried out to obtain the differentiated frequency demodulation results of the sensing points of the dual-core fiber array. Then a fault-tolerant control strategy is used to determine the sensor weight coefficients and vibration judgment coefficients during data fusion processing, and array data fusion is carried out based on time series data using Kalman filter to realize error value identification and filling. The advantage of this method is the combination of redundant data in a complementary way to improve the system stability. The frequency response ranges from 10 Hz to 2400 Hz and the localization accuracy is 98.33%. The influence of key parameters on the frequency demodulation performance of fault-tolerant Kalman filter is discussed, and a standard deviation of 14.6 Hz and an average error of 7.6 Hz are obtained. The demodulation frequency data matrix obtained by the classical demodulation method has a demodulation error probability of 89.18%, which proves the widespread existence of demodulation errors in vibration signals. The fusion error of demodulation frequency is reduced to 0.25 Hz, the frequency demodulation accuracy reaches 100%, and the demodulation error caused by interference attenuation can be completely eliminated. This system based on fault-tolerant Kalman filter has the characteristics of simple multiplexing structure, interference fading resistance and stable demodulation performance. • A sensor information fusion algorithm based on fault-tolerant Kalman filter is proposed to achieve fading-free sensing. [ABSTRACT FROM AUTHOR]

Digital transformation in manufacturing necessitates the integration of advanced fields such as robotics, the Internet of Things (IoT), and data science. The latter two, in particular, introduce a unique set of challenges related to data acquisition and processing, especially for emerging manufacturing sectors that lack substantial data volumes for analytical software. This paper specifically addresses the challenges associated with limited data within the context of precision poultry farming. We begin by identifying the specific challenges encountered during the development of a smart poultry management software system. Subsequently, we propose potential solutions for each identified challenge. The outcome of this study is a comprehensive algorithm for data processing that encapsulates these proposed solutions, offering a practical approach to overcoming data limitations in precision poultry farming. [ABSTRACT FROM AUTHOR]

Based on the rapid development of Internet new generation information technology, campus culture has become an important force to promote educational innovation. However, the uneven quality of campus culture teaching is also increasingly exposed. Therefore, we propose a set of campus culture intelligent quality assessment feedback system driven by data and knowledge. Based on real multi-modal campus culture data, the evaluation system covers different online teaching forms, and can comprehensively consider and measure the core elements of quality evaluation from the perspective of teachers and students, so as to achieve accurate evaluation and feedback of campus culture. This paper proposes to combine emotion computing with big data analysis, construct a theoretical framework of multi-modal emotion analysis from the perspective of multi-spatial fusion, and explore the students' learning emotion recognition and intelligent assessment based on multi-modal data fusion. By studying the key issues, logical path and implementation path of multimodal learning emotion analysis in campus culture, this paper reveals the mechanism of student emotion development from the perspective of campus culture, expands the intervention practice of student group emotion, and provides technical support and action guidance for after-school teaching and research on campus culture. [ABSTRACT FROM AUTHOR]

The encoder-decoder structure is the basic structure of most semantic segmentation models and is adopted by a large number of segmentation models. How to effectively extract image features and achieve high-precision mapping through the optimal design of encoder and decoder is the key issue of current research. SegFormer designs an encoder with excellent performance, which fully extracts the feature information of different semantic granularity in the image with a large receptive field. Even if a simple fully connected layer decoder is used, excellent segmentation results can also be achieved. However, this simplified decoder does not make full use of the advantages of the SegFormer encoder. Therefore, a decoder structure with dual-path multi-scale feature fusion is designed in this paper, and the decoder is redesigned according to the characteristics of the SegFormer encoder. The decoder adopts a dual-path structure, one path passes the abstract global information layer by layer to the local detail information through the layer-by-layer upsampling fusion module (LFM), and gradually upsamples the feature maps obtained from the encoder, and then use the channel fusion module to learn the importance of different channels in the deep abstract semantic feature map and the shallow local detail feature map, and perform dynamic fusion to obtain a feature map containing both abstract semantic information and local details. The other path takes advantage of the large receptive field of the feature map output by the SegFormer encoder, and uses the weighted hybrid multi-scale feature extraction module (WMF) to extract multi-scale features containing global semantics from the deep semantic feature map finally output by the encoder. Finally, the Deep Feature Fusion Module (DFM) is used to fuse the outputs of the first two modules, fully mining the multi-scale global information in the encoder, and obtaine the feature maps with rich semantic information, which effectively improves the algorithm model performance. [ABSTRACT FROM AUTHOR]

This paper introduces FusionDiff, a novel unified end-to-end image fusion network based on diffusion probabilistic models. The proposed method addresses the data dependency issue present in current unified image fusion approaches. Initially, the method pre-fuses the source images after diffusion. Subsequently, a noise prediction network forecasts the noise applied to the pre-fused image. During this stage, the Spatially-Adaptive Constraint layer is employed to restrict the fusion image, thereby maximizing the preservation of respective source image features. Finally, in the reverse process, skip-sampling is employed to effectively overcome the computational time drawback of traditional diffusion probabilistic models while maintaining high-quality image generation. Compared to other diffusion model-based image fusion methods, our approach stands out for its structural simplicity and ease of training, serving as a unified image fusion method. Furthermore, compared to other unified image fusion methods, our proposed network fully leverages the generalization of the diffusion probabilistic model, achieving adaptive feature extraction by constraining the inverse process of source images. Qualitative and quantitative experimental results across four classic image fusion tasks demonstrate the superiority of our method over state-of-the-art unified image fusion methods in recent years. • We propose a unified image fusion method based on the Diffusion Probabilistic Models. • We address the issue of data dependency, requiring training on only one multi-focus dataset to be applicable to other image fusion tasks. • We designed a spatially-adaptive constraint layer to impose conditional constraints on the diffusion model. [ABSTRACT FROM AUTHOR]

• Shanghai's urban building energy model was established using diverse open data. • The reliability of open data was enhanced through multi-source data fusion. • 95 % building classification accuracy was achieved with machine learning. • Urban building energy simulation for 539,119 buildings was conducted. Urban building energy modeling is crucial for guiding carbon reduction policies, but acquiring reliable data at the urban scale remains challenging. This study develops a model for Shanghai City, China, by integrating multi-source open data. Eight data sources were collected, including maps, satellite imagery, and GIS data, covering 609,763 building footprints and 539,119 buildings (1.57 billion m2). Spatial analysis, supervised learning, and unsupervised machine learning methods were used to categorize buildings into 63 prototypes, and classification accuracy reached 95 %. Historical satellite data and community boundaries determined the year built for over 95 % of buildings. Prototypes were modeled in AutoBPS using local energy saving standards and simulated in EnergyPlus to derive energy use intensities aligning with government ranges. This work demonstrates a practical data fusion approach to develop large-scale, reliable urban building energy models. Integrating heterogeneous open data sources expands the coverage of open data and improves accuracy. The framework and insights provide a valuable foundation to leverage open data for advancing city-scale energy modeling and sustainability planning. [ABSTRACT FROM AUTHOR]

Microelectronic technology and communication technology are developed in deep manner; the computing mode has been transferred from traditional computer-centered to human centered pervasive. So, the concept of Internet of things (IoT) is gradually put forward, which allows people to access information about their surroundings on demand through different terminals. The library is the major public space for human to read and learn. How to provide a more comfortable library environment to better meet people's learning requirements is a place where the Internet of things plays its role. The purpose of this paper is to solve the difference between the data fusion of library environment and the data fusion of other environments by the method of data fusion oriented to library. This paper presents a general technical framework of situational awareness for smart library system which includes a data fusion middleware. It can process data and inform the upper module of the changed library environment after deploying the smart library system in a library, including data collection and processing, how to judge whether events are triggered, how the system reacts, and the acquisition and update of user preferences. This paper presents a situational awareness recommendation method based on an effective data fusion model and algorithm for library after conducting experimental in service of library, which give more accurate of book recommendation than traditional method and good learning service environment of library for readers. [ABSTRACT FROM AUTHOR]

The recommendation algorithm based on collaborative filtering is currently the most successful recommendation method. It recommends items to the user based on the known historical interaction data of the target user. Furthermore, the combination of the recommended algorithm based on collaborative filtration and other auxiliary knowledge base is an effective way to improve the performance of the recommended system, of which the Co-Factorization Model (CoFM) is one representative research. CoFM, a fusion recommendation model combining the collaborative filtering model FM and the graph embedding model TransE, introduces the information of many entities and their relations in the knowledge graph into the recommendation system as effective auxiliary information. It can effectively improve the accuracy of recommendations and alleviate the problem of sparse user historical interaction data. Unfortunately, the graph-embedded model TransE used in the CoFM model cannot solve the 1-N, N-1, and N-N problems well. To tackle this problem, a novel fusion recommendation model Joint Factorization Machines and TransH Model (JFMH) is proposed, which improves CoFM by replacing the TransE model with TransH model. A large number of experiments on two widely used benchmark data sets show that compared with CoFM, JFMH has improved performance in terms of item recommendation and knowledge graph completion, and is more competitive than multiple baseline methods. [ABSTRACT FROM AUTHOR]

With the progress of new technologies of information and communication, more andmore producers of data exist. On the other hand, the web forms a huge support of all these kinds of data. Unfortunately, existing data is not proper due to the existence of the same information in different sources, aswell as erroneous and incomplete data. The aim of data integration systems is to offer to a user a unique interface to query a number of sources. A key challenge of suchsystems is todealwith conflicting information fromthe samesource or fromdifferent sources. We present, in this paper, the resolution of conflict at the instance level into two stages: references reconciliation and data fusion. The reference reconciliationmethods seek to decide if two data descriptions are references to the same entity in reality. We define the principles of reconciliation method then we distinguish the methods of reference reconciliation, first on howto use the descriptions of references, then theway to acquire knowledge. We finish this sectionby discussing some current data reconciliation issues that are the subject of current research. Data fusion in turn, has the objective to merge duplicates into a single representation while resolving conflicts between the data. We define first the conflicts classification, the strategies for dealing with conflicts and the implementing conflict management strategies. We present then, the relational operators and data fusion techniques. Likewise, we finish this section by discussing some current data fusion issues that are the subject of current research. [ABSTRACT FROM AUTHOR]

Models that predict the future state of certain observations are commonly developed by utilizing heterogeneous data. Most traditional prediction models tend to ignore inconsistencies and imperfections in heterogeneous data, and they are also limited in their ability to consider spatial correlations among monitoring points, as well as predict for the entire study area simultaneously. As a solution, this paper proposes a deep learning method to fuse heterogeneous data collected from multiple monitoring points by exploiting graph convolutional networks (GCNs), thus perform prediction for certain observations. The effectiveness of this approach was evaluated by applying it to a prediction scenario for air quality. As the fundamental idea behind the proposed method, (1) the collected heterogeneous data is fused according to the coordinates of the monitoring points considering its spatial correlations, and (2) the prediction considers global information as opposed to local information. According to an air quality prediction scenario, (1) the fused data obtained by the RBF-based fusion method are reasonable and reliable; (2) the fusion significantly increases predictive model performance; and (3) the STGCN model enhanced by the fusion achieves the highest performance. This approach can be similarly applied in scenarios involving continuous heterogeneous data collected from scattered multiple monitoring points in the study area. [ABSTRACT FROM AUTHOR]

Chaotic time series data is extensively applied in financial stocks, climate monitoring, and sea clutter, in which data fusion from various sources and multi-sensor information make accurate predictions of chaotic time series challenging under complex nonlinear conditions. Previous works focus on designing different model frameworks to capture the temporal dependence and extract richer nonlinear features to improve the accuracy of univariate chaotic time series prediction, which ignores the spatial dependence of multivariable. However, in this paper, we argue that spatial correlation among multiple variables is essential to improve the prediction accuracy of chaotic time series. To fill the gap, we innovatively propose a D ynamic A daptive G raph C onvolutional T ransformer with a B road L earning S ystem (DAGCT-BLS), a GCN and Transformer-based model utilizing multivariate spatial dependence for multi-dimensional chaotic time series forecasting. In DAGCT-BLS, the multivariate chaotic time series are reconstructed into the phase space, and the reconstructed data are rapidly feature-extracted using a cascade network BLS with frozen weights to maximize the retention of chaotic properties and nonlinear relationships. Then, the Dynamic Adaptive Graph Convolutional Network (DAGCN) is proposed to capture the spatial correlation among the multiple variables. Finally, improved multi-head attention of the Transformer Encoder is used to capture the temporal dependence of the phase point sequence. Experiments of our proposed model on three datasets (Lorenz, Rossler, and Sea clutter) show that DAGCT-BLS can achieve the best prediction performance and have strong interpretability, and multivariate-based joint modeling of chaotic time series helps to improve the prediction performance. • A novel DAGCT-BLS model is proposed to sovel multi-dimensional chaotic time series forecasting problem, unreported previously. • The multivariate chaotic time series are reconstructed into the phase space, which are rapidly feature-extracted using a cascade network BLS with frozen weights. • Improved multi-head attention is proposed to capture the temporal dependence of the phase point sequence. • DAGCN is proposed to model multivariate spatial relationships dynamically. [ABSTRACT FROM AUTHOR]

Adaptive management of large herbivores requires an understanding of how spatial‐temporal fluctuations in forage biomass and quality influence animal performance. Advances in remote sensing have yielded information about the spatial‐temporal dynamics of forage biomass, which in turn have informed rangeland management decisions such as stocking rate and paddock selection for free‐ranging cattle. However, less is known about the spatial‐temporal patterns of diet quality and their influence on large herbivore performance. This is due to infrequent concurrent ground observations of forage conditions with performance (e.g., mass gain), and previously limited satellite data at fine spatial and temporal scales. We combined multi‐temporal field observations of diet quality (weekly) and mass gain (monthly) with satellite‐derived phenological metrics (pseudo‐daily, using data fusion and interpolation) to model daily mass gains of free‐ranging yearling cattle in shortgrass steppe. We used this model to predict grazing season (mid‐May to October) mass gains, a key management indicator, across 40 different paddocks grazed over a 10‐year period (n = 138). We found strong relationships between diet quality and the satellite‐derived phenological metrics, especially metrics related to the timing and rate of green‐up and senescence. Satellite‐derived diet quality estimates were strong predictors of monthly mass gains (R2 = 0.68) across a wide range of aboveground net herbaceous production. Season‐long predictions of average daily gain and cattle off‐mass had mean absolute errors of 8.9% and 2.9%, respectively. The model performed better temporally (across repeated observations in the same paddock) than spatially (across all paddocks within a given year), highlighting the need for accurate vegetation maps and robust field data collection across both space and time. This study demonstrates that free‐ranging cattle performance in rangelands is strongly affected by diet quality, which is related to the timing of vegetation green‐up and senescence. Senescing vegetation suppressed mass gains, even if adequate forage was available. The satellite‐based pseudo‐daily approach presented here offers new opportunities for adaptive management of large herbivores, such as identifying within‐season triggers to move livestock among paddocks, predicting wildlife herd health, or timing the grazing season to better match earlier spring green‐up caused by climate change and plant species invasion. [ABSTRACT FROM AUTHOR]

Multivariable machine learning (ML) models are increasingly used for time series predictions. However, avoiding the overfitting and underfitting in ML-based time series prediction requires special consideration depending on the size and characteristics of the available training dataset. Predictive error compensating wavelet neural network (PEC-WNN) improves the time series prediction accuracy by enhancing the orthogonal features within a data fusion scheme. In this study, time series prediction performance of the PEC-WNNs have been evaluated on two different problems in comparison to conventional machine learning methods including the long short-term memory (LSTM) network. The results have shown that PECNET provides significantly more accurate predictions. RMSPE error is reduced by more than 60% with respect to other compared ML methods for Lorenz Attractor and wind speed prediction problems. [ABSTRACT FROM AUTHOR]

Multi-source incomplete mixed data abound in real life, like medical data, biological data, remote sensing data, military data, etc. However, some of these sources are of less importance than others, and some are essentially worthless. Therefore, their information fusion and attribute reduction face many challenges. This paper studies information fusion for multi-source incomplete mixed data via conditional information entropy (CIE) and considers its application to attribute reduction based on D-S evidence theory. First of all, a new distance function is defined to measure the difference between nominal attribute values with missing information, and the neighborhood rough set model is used to establish the granularity structure of multi-source incomplete mixed data. Then, a source selection method is given via CIE, which is used to fuse multi-source incomplete mixed data into single-source incomplete mixed data. Based on the maximization of CIE, this method allows worthy and reliable information sources to be chosen. Next, the connection between neighborhood rough set model and D-S evidence theory is established. Moreover, two attribute reduction algorithms for the fused incomplete mixed data are proposed based on the belief and plausibility. Finally, experiments are done to verify the effectiveness of the proposed fusion and reduction algorithms. The results of experiment and statistical test on 12 datasets show that the proposed algorithms exceed other advanced algorithms in classification performance. • The neighborhood rough set model is used to establish the granularity structure. • A source selection method is given via conditional information entropy. • The connection between the neighborhood rough set model and D-S evidence theory is established. • Two attribute reduction algorithms for the fused incomplete mixed data are proposed. • Experiments are done to verify the effectiveness of the proposed fusion and reduction algorithms. [ABSTRACT FROM AUTHOR]

The intelligent agriculture monitoring is based on the perception and analysis of environmental data, which enables the monitoring of the production environment and the control of environmental regulation equipment. As the scale of the application continues to expand, a large amount of data will be generated from the perception layer and uploaded to the cloud service, which will bring challenges of insufficient bandwidth and processing capacity. A fog-based offline and real-time hybrid data analysis architecture was proposed in this paper, which combines offline and real-time analysis to enable real-time data processing on resource-constrained IoT devices. Furthermore, we propose a data process-ing algorithm based on the incremental principal component analysis, which can achieve data dimensionality reduction and update of principal components. We also introduce the concept of Squared Prediction Error (SPE) value and realize the abnormal detection of data through the combination of SPE value and data fusion algorithm. To ensure the accuracy and effectiveness of the algorithm, we design a regular-SPE hybrid model update strategy, which enables the principal component to be updated on demand when data anomalies are found. In addition, this strategy can significantly reduce resource consumption growth due to the data analysis architectures. Practical datasets-based simulations have confirmed that the proposed algorithm can perform data fusion and exception processing in real-time on resource-constrained devices; Our model update strategy can reduce the overall system resource consumption while ensuring the accuracy of the algorithm. [ABSTRACT FROM AUTHOR]

This research develops a novel lightweight sport activity detection system using time series data of the accelerometer sensors embedded in two modalities, the smartphone and smartwatch. This research focuses on the lightweight exercises, more specifically, jumping jack, push up, sit up, and squat jump. Such activities are chosen for two reasons; the smartphone is now accessible for many persons, and the lightweight exercises are deemed to be easily completed in daily basis for everyone. Our proposed algorithm includes two novel feature extraction methods, i.e. sliding variance and data range , combined with a digital filter and data clipping methods. The results of feature extraction processes were classified using a combination of k-NN and DTW algorithms. The classification results are subsequently compared with the-state-of-the-art algorithms, i.e. LMNN and Naïve Bayes algorithms. The final results imply the merge of k-NN and DTW algorithms (k = 1) with data range method achieves the highest accuracy. The average accuracy for this method is 97.4%, with the processing time of 0.86 s. Thus, the result of counting activity method was acceptable with average values 80% for the whole movement by using two sensor accelerometers. [ABSTRACT FROM AUTHOR]

Sequential data is widely used in several fields, such as power payload prediction, traffic flow prediction, and stock trend prediction. Driven by the urgent needs, sequential forecasting based on deep learning methods has received lots of attention in recent years. However, the potential of deep learning methods in sequential data forecasting has not yet fully been exploited in terms of model architecture. In this study, a pre-trained nodal model with multiple fusion layers architecture (Sequence Prediction via Node Fusion, SPNF) was proposed, the model considered both connections of nodes and the temporal components of nodes to predict value of next node. Multiple fusion layers were adopted to capture spatial features and temporal dependencies from historical data. The proposed model also conpensated missing data via a masking mechanism. To validate the proposed model, experiments were also carried out using field-captured traffic data, the performance of proposed model were compared with classical and state-of-theart models, such as, ARIMA, SVR, LSTM, TGC-LSTM neural networks. The results showed that proposed model yields higher accuracy and robustness than others, especially in the case of large sequence changes occurred. [ABSTRACT FROM AUTHOR]

Sequential data is widely used in several fields, such as power payload prediction, traffic flow prediction, and stock trend prediction. Driven by the urgent needs, sequential forecasting based on deep learning methods has received lots of attention in recent years. However, the potential of deep learning methods in sequential data forecasting has not yet fully been exploited in terms of model architecture. In this study, a pre-trained nodal model with multiple fusion layers architecture (Sequence Prediction via Node Fusion, SPNF) was proposed, the model considered both connections of nodes and the temporal components of nodes to predict value of next node. Multiple fusion layers were adopted to capture spatial features and temporal dependencies from historical data. The proposed model also conpensated missing data via a masking mechanism. To validate the proposed model, experiments were also carried out using field-captured traffic data, the performance of proposed model were compared with classical and state-of-theart models, such as, ARIMA, SVR, LSTM, TGC-LSTM neural networks. The results showed that proposed model yields higher accuracy and robustness than others, especially in the case of large sequence changes occurred. [ABSTRACT FROM AUTHOR]

This study proposes a method of feature extraction based on Kernel Fisher Discriminant Analysis (KFDA) to solve problems in the classification of underwater targets, specifically the large number of original characteristic parameters and significant nonlinearity. First, a large number of features are combined through serial feature fusion to establish a new feature vector space, and KFDA is used to extract the optimal nonlinear discriminant features. Second, a test bed for an underwater experiment featuring a data processing system, echo signal acquisition, and feature extraction is described. Finally, underwater acoustic experiments are carried out, and the results of the measurement data indicate that the proposed method is superior to currently used techniques in the area. [ABSTRACT FROM AUTHOR]

In view that the MEMS gyroscope has low precision, the virtual gyro is used to improve the output precision of the MEMS gyro. By analyzing the factors affecting the fusion precision of virtual gyro, the essential relationship between gyro noise correlation and the fusion precision is obtained. In static, the influence of the noise characteristics and noise correlation coefficient on the output precision of MEMS gyro array is great; while in dynamic, the influence of the noise characteristics is small due to system modeling error and non-linear error, the influence of the angular rate modeling accuracy is great. Therefore, the Taylor model is used to extend the motion model described by classical Newtonian mechanics, and the angular motion model of different modeling series is established. This paper introduces the improved current statistical model, uses the modified Rayleigh function to describe the angular motion variable, establishes the zero-mean colored driving noise model, realizes the adaptive adjustment of the frequency and the driving noise intensity, and establishes the high-precision dynamic angular rate model of different modeling series. Finally, the improved interactive multi-model Kalman filter is used to fuse the output of the gyro in the MEMS gyro array. The simulation results show that the proposed method can improve the output precision of MEMS gyroscopes effectively and has certain engineering application value. • The influence of noise characteristics on the virtual gyro output accuracy is clarified. In dynamic state, the effect of noise characteristics on the virtual gyro output accuracy is small, while the effect of angular rate modeling is large. • The current statistical model is improved to adjust the standard deviation of driving noise adaptively. • An improved interactive multi-model Kalman filter based on improved current statistical model (ICS-IMM) is proposed, which improves the output accuracy and adaptability of the system. [ABSTRACT FROM AUTHOR]

This study proposes a method of feature extraction based on Kernel Fisher Discriminant Analysis (KFDA) to solve problems in the classification of underwater targets, specifically the large number of original characteristic parameters and significant nonlinearity. First, a large number of features are combined through serial feature fusion to establish a new feature vector space, and KFDA is used to extract the optimal nonlinear discriminant features. Second, a test bed for an underwater experiment featuring a data processing system, echo signal acquisition, and feature extraction is described. Finally, underwater acoustic experiments are carried out, and the results of the measurement data indicate that the proposed method is superior to currently used techniques in the area. [ABSTRACT FROM AUTHOR]

Big data is developing rapidly these years, leading people gradually to realize the enormous value and various utilities hidden behind data. Therefore, it is vital to form a deep and detailed knowledge of big data field. In this paper, we mainly collect data from different sources, cite the conclusions and discoveries from other scholars in order to get more accurate answers and predictions. Meanwhile, we analyze some fields that may be closely related to big data and may obviously influence it, such as cloud computing and AI, and hope to find the inner connection between their development and big data development. Plus, we analyze the history and current situation of big data, and make reasonable predictions towards the development of its future. Based on data, thesis from other scholars and conclusions from reports, we speculate our prediction towards the future of big data, and it can strongly support the judgement of big data development tendency. [ABSTRACT FROM AUTHOR]

There is a quality benefit error in the block collaborative supply chain that is affected by the environmental factors of the supply market. In order to improve the ability of collaborative quality evaluation, a supply chain collaborative quality evaluation method based using structural characteristics is proposed. The sampling model of collaborative quality characteristic information of block collaborative supply chain is constructed. According to the quantitative recurrent analysis results of the collaborative quality sample data, block chain information fusion is carried out, and the association rules fusion feature distribution parameter set of block collaborative supply chain collaborative quality panel data is extracted. The collaborative quality data fusion of block collaborative supply chain is carried out by using the method of random probability density feature detection. Combined with the method of piecewise linear estimation, the statistical feature quantity of supply chain collaborative quality evaluation is constructed. According to the prior sample quantitative recurrent analysis results of supply chain collaborative quality evaluation, the collaborative quality feature quantity is analyzed, and the cooperative quality feature quantity of block collaborative supply chain is extracted. The structure feature extraction and fusion clustering method are used for information clustering. According to the distributed fusion results of characteristics, the collaborative quality of supply chain is evaluated. The simulation results show that the proposed method has high accuracy and good confidence in the evaluation of supply chain collaborative quality, and it improves the ability of collaborative quality control of block collaborative supply chain. [ABSTRACT FROM AUTHOR]

Multiple Mobile Robots System (MMRS) has shown many attractive features in lots of real-world applications that motivate their rapid and wide diffusion. In MMRS, the Cooperative Localization (CL) is the basis and premise of its high-performance task. However, the statistical characteristics of the system noise should be already known in traditional CL algorithms, which is difficult to satisfy in actual MMRS because of the numerous of disturbances form the complex external environment. So the CL accuracy will be reduced. To solve this problem, an improved Adaptive Active Cooperative Localization (A2CL) algorithm based on information optimization strategy for MMRS is proposed in this manuscript. In this manuscript, an adaptive information fusion algorithm based on the variance component estimation under Extended Kalman filter (VCEKF) method for MMRS is introduced firstly to enhance the robustness and accuracy of information fusion by estimating the covariance matrix of the system noise or observation noise in real time. Besides, to decrease the effect of observation uncertainty on CL accuracy further, an observation optimization strategy based on information theory, the Model Predictive Control (MPC) strategy, is used here to maximize the information amount from observations. And semi-physical simulation experiments were carried out to verity the A2CL algorithm's performance finally. Results proved that the presented A2CL algorithm based on information optimization strategy for MMRS cannot only enhance the CL accuracy effectively but also have good robustness. [ABSTRACT FROM AUTHOR]

In order to improve the big data analysis and traffic data management ability of the Jiangxi high-speed traffic condition monitoring platform, a multi-dimensional multi-scale analysis model of Jiangxi high-speed traffic condition characteristics based on vehicle networking and traffic big data information fusion is proposed. On the basis of building a good knowledge base, model base, and method base, the model base feature analysis and dynamic detection method are used to effectively mine big data of the Jiangxi high-speed traffic status monitoring platform. Combined with information fusion theory, the multi-dimensional model of the Jiangxi high-speed traffic status monitoring platform based on text information, location information, picture, audio, video, and other Jiangxi high-speed traffic status data can be realized. The test results show that big data mining in the platform with this method has better clustering and achieves high multi-dimensional multi-scale fusion of high-speed traffic condition characteristics in Jiangxi. The efficiency of data scheduling and access are improved effectively, enhancing the multi-dimensional multi-scale analysis and resource scheduling ability of Jiangxi high-speed traffic condition characteristics. [ABSTRACT FROM AUTHOR]

In order to improve the decision-making level for public opinion responses and realize the semantic fusion of multi-level and multi-source heterogeneous public opinion information, an ontology-based public opinion information fusion method is proposed. Firstly, aiming at quick response decision-making, the situation assessment model of public opinion information fusion is studied, and the information fusion system is constructed. The multi-level evaluation model of situation recognition, situation understanding, and situation prediction is formed. Then, the multi-indicator ontology model and method for public opinion decision-making are constructed, and the public opinion data fusion model based on ontology semantics is proposed, which realizes the relevance analysis and semantic fusion of domain knowledge. Finally, a multi-level public opinion data fusion model is constructed, and the construction of the underlying emergency information knowledge base to support the above functions is deeply studied. The simulation results show that the feasibility and efficiency of the situation assessment problem are solved by this method, the time complexity and space complexity of attribute reduction and value reduction are reduced, and the matching efficiency of situation assessment rules is improved. [ABSTRACT FROM AUTHOR]

The goal of image fusion is to obtain large amount of information into a single image with more quantitatively and qualitatively. The image fusion can be applicable in various fields like multi-focus, multi-modal medical, satellite etc. This paper proposed statistical measurements of Multimodal MRI-PET medical image fusion using 2D Hartley transform (HT) in HSV color space. This proposed method was discussed with two different modalities of medical images like MRI (Magnetic Resonance Imaging) and PET (Positron Emission Tomography) and also discussed with five steps. Initially the PET color image is converted into HSV channels. Second step is MRI and V component of PET image are divided into 8*8 blocks and then apply 2D Hartley transform on each block of two input images. Third step is compute variance of each block of two images and then select best blocks. Fourth step is applying inverse 2D HT and all blocks are arranged into single image i.e. new V component. Finally concordination of New V component, H, S to get HSV image and then convert HSV to RGB to obtain final fused image with more accurate. The result shows the importance of the proposed image and this is superior to existing methods like DFT along with Smooth, Hartley along with Smooth, Hartley along with Mean, DFT along with Mean and DCT along with Smooth. The evaluation parameters such as Mean, Standard Deviation and Gradient plays a major role in image fusion for testing the quality. [ABSTRACT FROM AUTHOR]

Ambient ozone (O 3) predictions can be very challenging mainly due to the highly nonlinear photochemistry among its precursors, and meteorological conditions and regional transport can further complicate the O 3 formation processes. The emission-based chemical transport models (CTM) are broadly used to predict O 3 formation, but they may deviate from observations due to input uncertainties such as emissions and meteorological data, in addition to the treatment of O 3 nonlinear chemistry. In this study, an innovative recurrent spatiotemporal deep-learning (RSDL) method with model-monitor coupled convolutional recurrent neural networks (ConvRNN) has been developed to improve O 3 predictions of CTM. The RSDL method was first used to build the ConvRNN within a 24-h scale to characterize the spatiotemporal relationships between the monitored O 3 data and CTM simulations, and then incorporated the recurrent pattern to achieve 72-h multi-site forecasts based on a pilot study over the Pearl River Delta (PRD) region of China. The results showed that the RSDL method predicted O 3 with high accuracy over this case study, with an increase of 27.54% in the correlation coefficient (R) average for all sites as well as an increase in R of 0.14–0.21 for all cities compared to CTM. Moreover, the regional distribution of CTM was further improved by the RSDL predictions with the data fusion technique, which greatly reduced the underpredictions of O 3 concentrations, particularly in high O 3 -level areas (concentrations >160 μg/m3), with a 33.55% reduction in the mean absolute error (MAE). [Display omitted] • Recurrent spatiotemporal deep learning (RSDL) enhanced hourly O 3 forecast. • RSDL effectively reduced the concentration bias of CTM caused by initial conditions. • RSDL can improve the underprediction of O 3 peak concentrations in polluted days. • CTM fused with RSDL predictions had better O 3 regional distributions. [ABSTRACT FROM AUTHOR]

This paper presents three major Visible Light Communication (VLC) transmitter LED detection techniques and proposes a new detection method based on the fusion of their outputs. Statistics of each detection technique are used to estimate their outputs' likelihood, which is then used for weighted averaging of their results and to provide more robust detection output than the state of the art. An analytical model of the entire detection system is provided and a practical simulation using Open Computer Vision (OpenCV) and Python language is performed. [ABSTRACT FROM AUTHOR]

Highlights • A novel spatio-temporal data fusion workflow based on a multi-sensor multi-resolution algorithm was developed and applied to generate TIR synthetic image data at high temporal and spatial resolution. • The workflow includes two steps: in the first step, an estimator is yielded to generate TIR TOA radiance data on any given date, provided a MODIS radiance image is available. • The next step applies constrained unmixing of the 30-m resolution TIR images using the information about sub-pixel land-cover obtained from co-registered images at higher spatial resolution in the VNIR (Visible Near InfraRed) spectrum. • Two geographically diverse experiments were carried out using the same procedure: one in The Netherlands to evaluate the procedure and other in Puglia (Italy). • The results confirm the feasibility of the proposed methodology, which yields a synthetic thermal band to integrate with the multi-spectral data provided by the S2/MSI at 10 m resolution. Abstract Thermal InfraRed (TIR) image data at high temporal and spatial resolution are required to monitor the rapid development of crops during the growing season, taking into account the fragmentation of most agricultural landscapes. Moreover, integrating high-resolution satellite TIR data to calibrate hydrological models is a powerful information to efficiently monitor crop water use. Conversely, no single sensor meets these combined requirements in the TIR spectral region. Data fusion approaches offer an alternative to exploit observations from multiple sensors, providing image data to meet the combined requirements on spatial and temporal resolution. A novel spatio-temporal data fusion workflow based on a multi-sensor multi-resolution algorithm was developed and applied to generate TIR synthetic image data at high temporal and spatial resolution. The workflow includes two steps: in the first step, synthetic daily radiance images at Top of Atmosphere (TOA) and 30-m spatial resolution (at the ground) are generated using TIR radiometric data at TOA collected by the Moderate Resolution Imaging Spectroradiometer (MODIS) daily 1-km and Landsat 8/TIRS 16-day 30-m. This procedure is applied to two image pairs on different dates. The workflow yields an estimator to generate TIR TOA radiance data on any given date, provided a MODIS radiance image is available. The next step applies constrained unmixing of the 30 m (now considered as low-resolution) TIR images using the information about sub-pixel land-cover obtained from co-registered images at higher spatial resolution in the VNIR (Visible Near InfraRed) spectrum. In our case study, the L8/TIRS synthetic image data were unmixed to the Sentinel 2/MSI with 10 m × 10 m spatial resolution. Two geographically diverse experiments were carried out using the same procedure: one in The Netherlands to evaluate the procedure and other in Puglia (Italy) to generate a time series of the 10-m × 10-m TIR image data product. The validation experiment, where an actual TIRS image was applied as a reference, gave a RMSE value of 35.3 W/(m2 μm sr), which corresponds to a relative value of 8.5% against the TIRS reference values. The results confirm the feasibility of the proposed methodology, which yields a synthetic thermal band to integrate with the multi-spectral data provided by the S2/MSI at 10 m resolution. [ABSTRACT FROM AUTHOR]

Although evidence theory has been applied in sensor data fusion, it will have unreasonable results when handling highly conflicting sensor reports. To address the issue, an improved fusing method with evidence distance and belief entropy is proposed. Generally, the goal is to obtain the appropriate weights assigning to different reports. Specifically, the distribution difference between two sensor reports is measured by belief entropy. The diversity degree is presented by the combination of evidence distance and the distribution difference. Then, the weight of each sensor report is determined based on the proposed diversity degree. Finally, we can use Dempster combination rule to make the decision. A real application in fault diagnosis and an example show the efficiency of the proposed method. Compared with the existing methods, the method not only has a better performance of convergence, but also less uncertainty. [ABSTRACT FROM AUTHOR]

Copyright of Transactions of Beijing Institute of Technology is the property of Beijing University of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The objectives of this study were to accurately model relationships between spectral reflectance and water-quality parameters, including blue-green algae phycocyanin, chlorophyll a, total suspended solids, turbidity, and total dissolved solids; evaluate feature-level fusion to spectral data for water-quality modeling; and evaluate the effectiveness of machine learning regression techniques and decision-level fusion for water-quality variable prediction. We introduce the application of canonical correlation analysis fusion as a method for water-based spectral analysis to overcome the low signal-to-noise ratio of the data. Water-quality variables and spectral reflectance were used to create predictive models via machine learning regression models, including multiple linear regression, partial least-squares regression, Gaussian process regression, support vector machine regression, and extreme learning machine regression. The models were then combined using decision-level fusion. Results indicate that canonical correlation analysis feature-level fusion and machine learning techniques are superior to traditional methods. [ABSTRACT FROM AUTHOR]

Insider threats within organizations can take a variety of forms including fraud, theft of classified or proprietary data, and workplace violence. Insiders can be overt individuals who commit deliberate acts against their organization, or inadvertent insiders who provide access to organizational IT systems. The damage from insider threats was estimated at an annualized average of $4.3M per company across a variety of industries in a 2016 study, which does not include the damage to an organization's reputation, loss of business, and decrease in company value. The Scientific advances to Continuous Insider Threat Evaluation (SCITE) program sponsored by the Intelligence Advanced Research Projects Activity (IARPA) was created to improve insider threat detection. In the first phase of this effort, three competing teams were provided aggregated individual performance data and asked to answer system performance questions on a range of complex challenge problems. The purpose was to be able to detect a subset of individuals from within the overall population of an organization that displayed a behavior or belonged to a group. Teams were scored on three metrics: mean squared error, calibrated certainty interval, and interval score. The team led by Innovative Decisions, Inc. (IDI) used a multiple model approach with data fusion to model these problems. This approach proved far superior to the methods used by the two other teams across all three metrics. The results of this research are applicable to any area where the objective is to identify a very small subset of a much larger group using incomplete, noisy data. Specific areas benefitting education and research include airline passenger screening, vetting of immigration or visa applications, and security clearance reviews. [ABSTRACT FROM AUTHOR]

Recently, continuous dimensional emotion recognition from audiovisual clues has attracted increasing attention in both theory and in practice. The large amount of data involved in the recognition processing decreases the efficiency of most bimodal information fusion algorithms. A novel algorithm, namely the incomplete Cholesky decomposition based kernel cross factor analysis (ICDKCFA), is presented and employed for continuous dimensional audiovisual emotion recognition, in this paper. After the ICDKCFA feature transformation, two basic fusion strategies, namely feature-level fusion and decision-level fusion, are explored to combine the transformed visual and audio features for emotion recognition. Finally, extensive experiments are conducted to evaluate the ICDKCFA approach on the AVEC 2016 Multimodal Affect Recognition Sub-Challenge dataset. The experimental results show that the ICDKCFA method has a higher speed than the original kernel cross factor analysis with the comparable performance. Moreover, the ICDKCFA method achieves a better performance than other common information fusion methods, such as the Canonical correlation analysis, kernel canonical correlation analysis and cross-modal factor analysis based fusion methods. [ABSTRACT FROM AUTHOR]

The degradation analysis of products has been demonstrated as a significant toolkit for reliability analysis. Data from the same batch of products in different working environments cannot be directly used to analyze product reliability. In this paper, motivated by this circumstance, we first assume that degradation data sets from different working environments are subject to different inverse Gaussian process models, and maximum likelihood estimation is used to obtain multiple model parameters. Secondly, we construct evidence by quantifying different information of products, apply the evidence theory to fuse model parameters, and then analyze the reliability of products from the same batch. Finally, we use performance degradation data of the laser to illustrate the method. [ABSTRACT FROM AUTHOR]

Abstract Cognitive load condition is of great significance for judging learners' learning state and improving the learning and teaching effects. This paper proposed a feature fusion based processing framework for high cognitive load detection, which includes heart rate variability (HRV) and pulse rate variability (PRV) acquisition, data preprocessing, feature extraction, feature selection, feature fusion by linear feature dependency modeling (LFDM) and high cognitive load detection by XGBoost classifier. This paper experiment on simulated learning paradigm, and the experimental results show that the proposed framework for detection of high cognitive load outperforms conventional processing approaches that uses HRV or PRV only. This paper compared the effects of using different feature fusion algorithms (PCT, SKRRR, ADMM, LFDM) and different classification algorithms (KNN, SVM, DT, RF, XGBoost), and the final proposed framework outperforms other schemes. The proposed framework achieves approximately 97.2% accuracy of high cognitive load detection. [ABSTRACT FROM AUTHOR]

Highlights • Data fusion and Gaussian mixture model are proposed to improve the performance of travel time estimation on highways. • The model provides advantages such as being robust to noisy data, reducing biases, and producing a more precise estimation. • The results from the case studies show improvement in the accuracy of travel time estimation in a range of 3.46%–16.3% (MAPE). Abstract This paper presents a framework for the development of the travel time estimation model using multiple sources of data with consideration of spurious data and traffic conditions. A modified Bayesian data fusion approach, combined with the Gaussian mixture model, is used to fuse the travel time data, which are estimated from different types of sensors to improve accuracy, precision, as well as completeness of data, in terms of spatial and temporal distribution. Two additional features are added into existing models including the difference of traffic conditions classified by the Gaussian mixture model and the bias estimation from individual sensor by introducing a non-zero mean Gaussian distribution which learned from the training dataset. The methodology and computational procedure are presented. The Gaussian mixture model is used to classify states of traffic into predefined number of traffic regimes. Once a traffic condition is classified, the modified Bayesian data fusion approach is used to estimate travel time. The proposed model provides explicit advantages over the basic Bayesian approach, such as being robust to noisy data, reducing biases of an individual estimation, and producing a more precise estimation of travel time. Two different real-world datasets and one simulated dataset are used to evaluate the performance of the proposed model under three different traffic regimes: free flow, transitional flow and congested flow regimes. The results when compared with the results from benchmark models show significant improvement in the accuracy of travel time estimation in terms of mean absolute percentage errors (MAPE) in the range of 3.46% to 16.3%. [ABSTRACT FROM AUTHOR]

Target recognition in uncertain environments is a hot issue, especially in extremely uncertain situation where both the target attribution and the sensor report are not clearly represented. To address this issue, a model which combines fractal theory, Dempster-Shafer evidence theory and analytic hierarchy process (AHP) to classify objects with incomplete information is proposed. The basic probability assignment (BPA), or belief function, can be modelled by conductivity function. The weight of each BPA is determined by AHP. Finally, the collected data are discounted with the weights. The feasibility and validness of proposed model is verified by an evidential classifier case in which sensory data are incomplete and collected from multiple level of granularity. The proposed fusion algorithm takes the advantage of not only efficient modelling of uncertain information, but also efficient combination of uncertain information. [ABSTRACT FROM AUTHOR]

Supplier selection is a common and important problem, which is usually modeled in a multi-criterion decision-making framework. In this framework, multiple criteria need to be determined and an expert team needs to be constructed. Generally, different criteria and experts always hold different weights. Due to the complexity of the practical problems, sometimes lots of fuzzy and uncertain information exists inevitably and the weights are tough to determine accurately. Interval number is a simple but effective method to handle the uncertainty. However, most current papers transform an interval number to a crisp number to make decisions, which causes loss of information more or less. To address this issue, a new evidential method based on interval data fusion is proposed in this paper. Dempster-Shafer evidence theory is a widely used method in a supplier selection problem due to its powerful ability in handling uncertainty. In our method, the criteria and decision makers are weighted in the form of interval numbers to generate interval basic probability probabilities, which represent the supporting degree to an event in Dempster-Shafer evidence theory. The obtained interval basic probabilities assignments are fused to make a final decision. Due to the properties of interval data, our method has its own specific advantages, like losing less information, providing different strategies for decision making and so on. The new method is proposed aiming at solving multi-criterion decision-making problems in which the reliability of decision makers or weights of criteria are described in the form of fuzzy data like linguistic terms or interval data. A numerical example of supplier selection is used to illustrate our method. Also, the results and comparison show the correctness and effectiveness of the new interval data fusing evidential method. [ABSTRACT FROM AUTHOR]

Text summarization is an application of information retrieval where short and non-redundant version of comparatively large text is presented to the end user. In this paper a hybrid approach is presented to generate abstract summary in which sentences are clustered using sentence level relationships among sentences in association with Markov clustering principle. Then sentence ranking is done in each cluster and if possible the top weighted sentence of each cluster is fused using some linguistic rules with its best-fit sentence(if found) within that cluster to generate a new sentence. Then top ranked sentences from each cluster are compressed using classification technique to generate the abstract summary. The proposed system is evaluated with DUC 2002 data set and the result is performing better than other existing systems. [ABSTRACT FROM AUTHOR]

A New method of fusion is presented to find each object present in the image. The proposed algorithm used USAN (univalue segment assimilating nucleus) technique to fuse the high and low intensity edges. The effective way of multi focus fusion is based on the fuzzy edge detection using modified fuzzy membership function. However, the proposed method computes the edge intensity region in an optimal way by using particle swarm optimization (PSO). We conduct experiments on several images and compare with different state of art methods. The effectiveness of the proposed approach outperforms than other methods in each case of performance measures. [ABSTRACT FROM AUTHOR]

The astronomical observation model, light curve model and solar radiation pressure force model are established to allow estimation of the positions and attitudes of space objects using a fusion of photometric and astronomical data. Existing algorithms for this purpose involve large errors. The proposed method firstly estimates the attitude of the object by using photometric and astronomical data fusion. Secondly, the position of the object is filtered by the estimated attitude. Theory and numerical simulations show the efficiency of the proposed method. [ABSTRACT FROM AUTHOR]

Energy consumption is an important performance indexe for Wireless Sensor Networks (WSNs) because the node energy is limited. Clustering structure facilitates data fusion and load balancing, thus reducing energy consumption. We propose an unequal multiple hops clustering algorithm (UMCA) for WSNs. UMCA is a multiple hops clustering algorithm. By changing the probabilities that nodes become cluster heads (CHs), UMCA selects CHs in the nodes with high residual energy. A node that has more around nodes, and is close to around nodes is more likely to become a CH. The possibility to becomes a CH also has relative with the distance to the base station (BS). There are more CHs near the BS, and less CHs away from the BS to balance the load of CHs. UMCA and other related algorithms are compared by simulation. The results show that UMCA can effectively prolong the network lifetime. [ABSTRACT FROM AUTHOR]

A system for monitoring and diagnosing the working conditions of equipment is developed in this study to ensure equipment safety and reliability. First, a perceptual model with four layers is designed. Original data are collected by sensors, and analyses are performed with intelligent algorithms. Decisions are then made and displayed on a screen in real time. Second, a method for monitoring equipment conditions is developed based on evidence weights. Basic probability assignment of evidence is corrected according to evidence and sensor weights, and an optimal fusion algorithm is selected by comparing an introduced threshold and a conflict factor. Third, the effectiveness and practicability of the algorithm are tested by simulating the monitoring and diagnosis of centrifugal pumps. Finally, the system is implemented to verify its validity. [ABSTRACT FROM AUTHOR]

Upscaling land cover maps is broadly employed to fill data gaps or match the spatial-resolution of preexisting projects. However, existing methods introduce systematic errors in the area information and the landscape pattern. We developed an upscaling method fusing the spatial structure information (i.e., class Membership probability) and the uncertainty information of the base map (e.g., Confidence level probability), called Fusing class Membership probability and Confidence level probability (FMC). The results showed that FMC obtained higher upscaling efficiency, and mitigated the negative influence of landscape heterogeneity and the negative influence of unequal proportions of land cover in the base maps, on the upscaling compared to Majority Rule Based (MRB) method. Additionally, FMC can reduce the uncertainty/error when these upscaled maps are used as input to Earth observation model (e.g., land cover change). [ABSTRACT FROM AUTHOR]

Quantifying change in urban land provides important information to create empirical models examining the effects of human land use. Maps of developed land from the National Land Cover Database (NLCD) of the conterminous United States include rural roads in the developed land class and therefore overestimate the amount of urban land. To better map the urban class and understand how urban lands change over time, we removed rural roads and small patches of rural development from the NLCD developed class and created four wall-to-wall maps (1992, 2001, 2006, and 2011) of urban land. Removing rural roads from the NLCD developed class involved a multi-step filtering process, data fusion using geospatial road and developed land data, and manual editing. Reference data classified as urban or not urban from a stratified random sample was used to assess the accuracy of the 2001 and 2006 urban and NLCD maps. The newly created urban maps had higher overall accuracy (98.7 percent) than the NLCD maps (96.2 percent). More importantly, the urban maps resulted in lower commission error of the urban class (23 percent versus 57 percent for the NLCD in 2006) with the trade-off of slightly inflated omission error (20 percent for the urban map, 16 percent for NLCD in 2006). The removal of approximately 230,000 km2 of rural roads from the NLCD developed class resulted in maps that better characterize the urban footprint. These urban maps are more suited to modeling applications and policy decisions that rely on quantitative and spatially explicit information regarding urban lands. [ABSTRACT FROM AUTHOR]

In this study, the reliability of nonlinear ultrasonic modulation based fatigue crack detection is improved using a feature-level data fusion approach. When two ultrasonic inputs at two distinct frequencies are applied to a specimen with a fatigue crack, modulation components at the summation and difference of these two input frequencies appear. First, the spectral amplitudes of the modulation components and their spectral correlations are defined as individual features. Then, a 2D feature space is constructed by combining these two features, and the presence of a fatigue crack is identified in the feature space. The effectiveness of the proposed fatigue crack detection technique is experimentally validated through cyclic loading tests of aluminum plates, full-scale steel girders and a rotating shaft component. Subsequently, the improved reliability of the proposed technique is quantitatively investigated using receiver operating characteristic analysis. The uniqueness of this study lies in (1) improvement of nonlinear ultrasonic modulation based fatigue crack detection reliability using feature-level data fusion, (2) reference-free fatigue crack diagnosis without using the baseline data obtained from the intact condition of the structure, (3) application to full-scale steel girders and shaft component, and (4) quantitative investigation of the improved reliability using receiver operating characteristic analysis. [ABSTRACT FROM AUTHOR]

The cowshed environment has significant impacts on the yield, diseases and behaviors of dairy cows. Heat stress, in particular, has a great impact on yield. The cowshed environment monitoring system based on wireless sensor network can accurately sense the temperature and other environmental parameters in real time and provide basis for manual environmental intervention and control. Energy constraint is one of the important problems that affect the long-term stable monitoring by the dairy cow wireless sensor network. So, the weighted Markov chain method is used to predict the time series of cowshed temperature. Replacing the actual values with the predicted values at the cluster head can effectively reduce data traffic in the cluster, thereby reducing network power consumption. Test data show that, the average variance of the cowshed environment temperature predicted by the method proposed in this paper is 0.185, and the average power consumption is reduced by about 40% when the compression ratio is 0.3, which effectively prolongs the network lifetime. In addition to that, the cowshed environment prediction can also help make prejudgments for environmental control, reduce or avoid the heat stress of dairy cows after the environmental parameters exceed the thresholds and provide the basis for the multi-source data fusion for dairy cow. [ABSTRACT FROM AUTHOR]