Search

This paper studies a common planning problem encountered in steelmaking production. The problem is to group different customer orders into a set of batches to accommodate the mass production mode of steelmaking furnaces. We formulate the problem as a novel mixed-integer programming model by considering the practical technological requirements. To solve the problem, we propose a cooperative coevolution framework in which an effective decomposition scheme based on graph partitioning is developed. The decomposition scheme first explores the problem structure by considering the production process rules and then exploits the batch information of the best-so-far solution to identify the potentially better decompositions. To solve each decomposed subcomponent, we propose a new differential evolution algorithm which incorporates a subpopulation-based classification mechanism and local search with an external archive strategy to balance the abilities of exploration and exploitation. Computational tests on a set of real production data as well as on a more diverse set of randomly generated problem instances show that our method is effective and efficient in practical application and outperforms other benchmark algorithms. [ABSTRACT FROM AUTHOR]

Seismic inversion, such as velocity and impedance, is an ill-posed problem. To solve this problem, swarm intelligence (SI) algorithms have been increasingly applied as the global optimization approach, such as differential evolution (DE) and particle swarm optimization (PSO). Based on the well logs, the sparse probability distribution (PD) of the reflectivity distribution is spatial stationarity. Therefore, we proposed a general SI scheme with constrained by a priori sparse distribution of the reflectivity, which helps to provide more accurate potential solutions for the seismic inversion. In the proposed scheme, as two key operations, the creating of probability density function library and probability transformation are inserted into standard SI algorithms. In particular, two targeted DE-PD and PSO-PD algorithms are implemented. Numerical example of Marmousi2 model and field example of gas hydrates show that the DE-PD and PSO-PD estimate better inversion solutions than the results of the original DE and PSO. In particular, the DE-PD is the best performer both in terms of mean error and fitness value of velocity and impendence inversion. Overall, the proposed SI with sparse distribution scheme is feasible and effective for seismic inversion.

This paper deals with a design method of adaptive output feedback control system with parallel feedforward compensator (PFC) for a multiple input multiple output system. An adaptive output feedback control system, which is based on almost strictly positive real (ASPR), requires PFC in most cases. The design method is using one-shot input/output data of the controlled system, and parameter optimization will be done by differential evolution (DE). By using DE, the designer can optimize the design parameters without trial and error, which can be a more practical method.

Differential evolution (DE) is a robust optimizer designed for solving complex domain research problems in the computational intelligence community. In the present work, a multi-hybrid DE (MHDE) is proposed for improving the overall working capability of the algorithm without compromising the solution quality. Adaptive parameters, enhanced mutation, enhanced crossover, reducing population, iterative division and Gaussian random sampling are some of the major characteristics of the proposed MHDE algorithm. Firstly, an iterative division for improved exploration and exploitation is used, then an adaptive proportional population size reduction mechanism is followed for reducing the computational complexity. It also incorporated Weibull distribution and Gaussian random sampling to mitigate premature convergence. The proposed framework is validated by using IEEE CEC benchmark suites (CEC 2005, CEC 2014 and CEC 2017). The algorithm is applied to four engineering design problems and for the weight minimization of three frame design problems. Experimental results are analysed and compared with recent hybrid algorithms such as laplacian biogeography based optimization, adaptive differential evolution with archive (JADE), success history based DE, self adaptive DE, LSHADE, MVMO, fractional-order calculus-based flower pollination algorithm, sine cosine crow search algorithm and others. Statistically, the Friedman and Wilcoxon rank sum tests prove that the proposed algorithm fares better than others. [ABSTRACT FROM AUTHOR]

In this research work, a differential evolution-based method has been used to compress the deep neural network architectures. The compression is achieved by selecting the most dominant filters/nodes during the training of the model. The usefulness of filters is established by the test accuracy of the model. The experimental results demonstrate that the performance of proposed model compares fairly well with other state of art model compression techniques. Moreover, the compression achieved with VGG16 on MNIST, CIFAR-10 and CIFAR-100 datasets was 98.32, 98.5 and 93.54%, respectively. The corresponding compression achieved on ResNet50 was 85.24, 85.38 and 79.37%, while SqueezeNet which is already compressed model could also be compressed by 72.94, 73.77 and 44.59%, respectively. MobileNet, which is already a compact model developed for mobile applications, could also be compressed by 93.04, 93.74 and 76.37% on MNIST, CIFAR-10 and CIFAR-100 datasets. The loss in accuracy in compressed models turns out to be less than 2%. Further, the compressed models report acceleration in inference time being 80.79% on VGG16, 74.14% on ResNet50, 42.96% on MobileNet and 11.79% on SqueezeNet. [ABSTRACT FROM AUTHOR]

Data hiding along with the security plays a vital role in the wireless network. In recent years, object detection, Feature Extraction (FE) and correlation based technique are developed to predict the pixels that is suitable for embedding. However, traditional data hiding method still suffered in the selection of less distortion pixels. To overcome these shortcomings, this method proposes (i) Moving Object (MO) detection, (ii) FE, (iii) data hiding, and (iv) data encryption. Initially, the MO in the video are detected using optical flow and tracking the weight of MO. Next, data hiding features are extracted using Optimization Fruit fly Algorithm (OFA) with Differential Evolution (DE) and Opposition Based Learning (OBL) algorithm. Then, a modified data hiding techniques based on correlation, regression and pixel segmentation with Prediction Error (PE) techniques are used for improving the performance. Finally, the hidden data is encrypted using Binary Tree Structure (BTS) to provide additional layer of security. Experimental result shows the proposed method is superior to other state-of-the-art methods in terms of Peak-Signal-to-Noise Ratio (PSNR), embedding capacity and security analysis. [ABSTRACT FROM AUTHOR]

These days image processing technology has evolved a lot. Digital media can be distorted easily due to the availability of editing tools. So, watermarking can be used to protect digital media. This article proposes the Discrete Wavelet Transform (DWT) and Singular Value Decomposition (SVD) based effective and optimal robust image watermarking technique using the differential evolution (DE) algorithm, which is invariant against rotation, scaling, and translation (RST). The watermark embedding process uses a 3-level DWT and SVD. Here, speeded-up robust features (SURF) are also used to extract feature points and match features. The RST-attacked image can be restored using these features by applying the M-estimator SAmple Consensus(M-SAC) algorithm. DE algorithm is used to estimate the optimized value of the scaling factor. The optimized scaling factor is assisted in acquiring the highest possible robustness with enhanced imperceptibility. After experimentation, results illustrate the proposed technique is invariant to the RST, and signal processing attacks. [ABSTRACT FROM AUTHOR]

Suppressing the spontaneous emission in quantum emitters ensembles (atoms) is one of the topical problems in quantum optics and quantum technology. While many approaches are based on utilizing the subradiance effect in ordered quantum emitters arrays, the ensemble configurations providing the minimal spontaneous emission rate are yet unknown. In this work, we employ the differential evolution algorithm to identify the optimal configurations of a few atomic ensembles that support quantum states with maximal radiative lifetime. We demonstrate that atoms tend to assemble mostly in quasi-regular structures with specific geometry, which strongly depends on the minimally allowed interatomic distance r min . While the discovered specific non-radiative realizations of small ensembles cannot be immediately predicted, there is particular correspondence to the non-radiative states in the atomic lattices. In particular, we have found that states inheriting their properties either from the bound states in the continuum or band edge states of infinite lattices dominate across a wide range of r min values. Additionally, we show that for small interatomic distances, the linear arrays with modulated spacing have the smallest radiative losses exponentially decreasing as the size of the ensemble increases. [ABSTRACT FROM AUTHOR]

Differential evolution is a popular heuristic black-box numerical optimization algorithm which is often used due to its simplicity and efficiency. Parameter adaptation is one of the main directions of study regarding the differential evolution algorithm. The main reason for this is that differential evolution is highly sensitive to the scaling factor and crossover rate parameters. In this study, a novel adaptation technique is proposed which uses the success rate to replace the popular success history-based adaptation for scaling factor tuning. In particular, the scaling factor is sampled with a Cauchy distribution, whose location parameter is set as an nth order root of the current success rate, i.e., the ratio of improved solutions to the current population size. The proposed technique is universal and can be applied to any differential evolution variant. Here it is tested with several state-of-the-art variants of differential evolution, and on two benchmark sets, CEC 2017 and CEC 2022. The performed experiments, which include modifications of algorithms developed by other authors, show that in many cases using the success rate to determine the scaling factor can be beneficial, especially with relatively small computational resource. [ABSTRACT FROM AUTHOR]

This paper explores the energy-intensive cement industry, focusing on a plant in Greece and its mill and kiln unit. The data utilized include manipulated, non-manipulated, and uncontrolled variables. The non-manipulated variables are computed based on the machine learning (ML) models and selected by the minimum value of the normalized root mean square error (NRMSE) across nine (9) methods. In case the distribution of the data displayed in the user interface changes, the user should trigger the retrain of the AI models to ensure their accuracy and robustness. To form the objective function, the expert user should define the desired weight for each manipulated or non-manipulated variable through the user interface (UI), along with its corresponding constraints or target value. The user selects the variables involved in the objective function based on the optimization strategy, and the evaluation is based on the comparison of the optimized and the active value of the objective function. The differential evolution (DE) method optimizes the objective function that is formed by the linear combination of the selected variables. The results indicate that using DE improves the operation of both the cement mill and kiln, yielding a lower objective function value compared to the current values. [ABSTRACT FROM AUTHOR]

X-ray inspections of contraband are widely used to maintain public transportation safety and protect life and property when people travel. To improve detection accuracy and reduce the probability of missed and false detection, a contraband detection algorithm YOLOv8s-DCN-EMA-IPIO* based on YOLOv8s is proposed. Firstly, the super-resolution reconstruction method based on the SRGAN network enhances the original data set, which is more conducive to model training. Secondly, DCNv2 (deformable convolution net v2) is introduced in the backbone network and merged with the C2f layer to improve the ability of the feature extraction and robustness of the model. Then, an EMA (efficient multi-scale attention) mechanism is proposed to suppress the interference of complex background noise and occlusion overlap in the detection process. Finally, the IPIO (improved pigeon-inspired optimization), which is based on the cross-mutation strategy, is employed to maximize the convolutional neural network's learning rate to derive the optimal group's weight information and ultimately improve the model's detection and recognition accuracy. The experimental results show that on the self-built data set, the mAP (mean average precision) of the improved model YOLOv8s-DCN-EMA-IPIO* is 73.43%, 3.98% higher than that of the original model YOLOv8s, and the FPS is 95, meeting the deployment requirements of both high precision and real-time. [ABSTRACT FROM AUTHOR]

The Time-Slotted Channel Hopping (TSCH) protocol is known for its suitability in highly reliable applications within industrial wireless sensor networks. One of the most significant challenges in TSCH is determining a schedule with a minimal slotframe size that can meet the required throughput for a heterogeneous network. We proposed a Priority-based Customized Differential Evolution (PCDE) algorithm based on the determination of a collision- and interference-free transmission graph. Our schedule can encompass sensors with different data rates in the given slotframe size. This study presents a comprehensive performance evaluation of our proposed algorithm and compares the results to the Traffic-Aware Scheduling Algorithm (TASA). Sufficient simulations were performed to evaluate different metrics such as the slotframe size, throughput, delay, time complexity, and Packet Delivery Ratio (PDR) to prove that our approach achieves a significant result compared with this method. [ABSTRACT FROM AUTHOR]

Purpose: Graph coloring approach has emerged as a valuable problem-solving tool for both theoretical and practical aspects across various scientific disciplines, including biology. In this study, we demonstrate the graph coloring's effectiveness in computational network biology, more precisely in analyzing protein–protein interaction (PPI) networks to gain insights about the viral infections and its consequences on human health. Accordingly, we propose a generic model that can highlight important hub proteins of virus-associated disease manifestations, changes in disease-associated biological pathways, potential drug targets and respective drugs. We test our model on SARS-CoV-2 infection, a highly transmissible virus responsible for the COVID-19 pandemic. The pandemic took significant human lives, causing severe respiratory illnesses and exhibiting various symptoms ranging from fever and cough to gastrointestinal, cardiac, renal, neurological, and other manifestations. Methods: To investigate the underlying mechanisms of SARS-CoV-2 infection-induced dysregulation of human pathobiology, we construct a two-level PPI network and employed a differential evolution-based graph coloring (DEGCP) algorithm to identify critical hub proteins that might serve as potential targets for resolving the associated issues. Initially, we concentrate on the direct human interactors of SARS-CoV-2 proteins to construct the first-level PPI network and subsequently applied the DEGCP algorithm to identify essential hub proteins within this network. We then build a second-level PPI network by incorporating the next-level human interactors of the first-level hub proteins and use the DEGCP algorithm to predict the second level of hub proteins. Results: We first identify the potential crucial hub proteins associated with SARS-CoV-2 infection at different levels. Through comprehensive analysis, we then investigate the cellular localization, interactions with other viral families, involvement in biological pathways and processes, functional attributes, gene regulation capabilities as transcription factors, and their associations with disease-associated symptoms of these identified hub proteins. Our findings highlight the significance of these hub proteins and their intricate connections with disease pathophysiology. Furthermore, we predict potential drug targets among the hub proteins and identify specific drugs that hold promise in preventing or treating SARS-CoV-2 infection and its consequences. Conclusion: Our generic model demonstrates the effectiveness of DEGCP algorithm in analyzing biological PPI networks, provides valuable insights into disease biology, and offers a basis for developing novel therapeutic strategies for other viral infections that may cause future pandemic. [ABSTRACT FROM AUTHOR]

The explosion of wireless network services has led to a need for new technologies and resource allocation schemes. Device-to-device (D2D) can reduce the access pressure of base station (BS) equipment, and non-orthogonal multiple access (NOMA) can multiplex network spectrum resources. The integration of them can further improve the performance of multi-cell networks. However, users still face more interference in multi-cell networks. This paper analyzes the relationship between transmission power and serial interference cancellation decoding order. Aiming at maximizing the sum rate of D2Ds, we propose three feasible frameworks based on the differential evolution (DE) algorithm. First, we invoke a coevolution DE-based resource allocation (CDRA) framework, which encodes the RB assignment and power allocation into the same individual for evolution. Second, we propose an iteration-combining DE-based resource allocation (IDRA) framework, which adopts a two-step DE algorithm to solve optimal power allocation and RB assignment scheme iteratively. Lastly, we invoke the power-repairing DE-based resource allocation (PDRA) framework, which can perform power repair on individuals with failed evolution. Simulation results demonstrate that: (1) the integration of D2D and NOMA techniques is capable of enhancing the achievable sum rate of D2Ds; (2) the three proposed frameworks for multi-cell NOMA networks can effectively optimize the objective function. Compared with CDRA and IDRA, PDRA performs better in terms of convergence speed and maximum D2D rate. [ABSTRACT FROM AUTHOR]

In the real market, the item's demand is substantially affected by the item's selling price and frequency of advertising. This study focuses on an optimal ordering policy followed by advertising, pricing, and preservation policies. The present study incorporates the quantity-dependent ordering cost and time-dependent holding cost. A partial prepayment scheme is developed for inventory purchase decisions. The spoilage impact can be effectively reduced by an optimal investment in refrigeration. The optimal decision policy has been proposed by using three metaheuristic algorithms, namely particle swarm optimization, real-coded genetic algorithm, and differential evolution algorithm. A comparison is made for these metaheuristic schemes based on the numerical illustrations. The parameter sensitivity is performed to get insights of the variability in the indicators of the inventory model. [ABSTRACT FROM AUTHOR]

Vortex search (VS) algorithm is a recently proposed swarm intelligence or evolutionary algorithm for solving continuous optimization problems inspired by the behavior of whirlpool. In this study, an approach based on VS algorithm is proposed to deal with uncapacitated facility location problem (UFLP) which is a pure problem in binary domain. The update mechanism of VS algorithm is not sufficiently useful for solving the binary optimization problems; therefore, a binary form of VS method called modified binary vortex search (for short MBVS) is proposed for solving UFLPs. Three important changes have been carried out on basic VS algorithm such as (1) converting continuous values to binary values; (2) using genetic mutation operators for enhancing the exploration ability and (3) a local search mechanism for extending the exploitation ability. Based on these changes, MBVS has been tested on fifteen different UFLP instances. The UFLPs dealt with in this study is one of the famous binary optimization problems. It is widely used for comparing the performance of superior algorithms. Once an analysis of 10 different transfer functions, a genetic mutation operator, a local search parameter and population size have been made on proposed method; then, it has been compared with some binary metaheuristic methods and their variants: genetic algorithm (GA)-based approaches such as GA-SP, GA-TP and GA-UP; binary particle swarm optimization algorithm (BPSO); binary versions of artificial bee colony (ABC) algorithm such as binABC, DisABC and ABCbin; binary versions of differential evolution (DE) algorithm such as DisDe/rand and binDE and binary variants of the artificial algae algorithm (AAA) such as AAA-Tanh, AAA-Sig and binAAA methods. The experimental results and comparisons reveal that MBVS algorithm is highly competitive and robust optimizer for the problem addressed in this study. [ABSTRACT FROM AUTHOR]

In positioning systems in wireless sensor networks, the accuracy of localization is often affected by signal distortion or attenuation caused by environmental factors, especially in indoor environments. Although using a combination of K-Nearest Neighbor (KNN) algorithm and fingerprinting matching can reduce positioning errors due to poor signal quality, the improvement in accuracy by increasing the number of reference points and K values is not significant. This paper proposes a Differential Evolution-based KNN (DE-KNN) method to overcome the performance limitations of the KNN algorithm and enhance indoor area positioning accuracy in WSNs. The DE-KNN method aims to improve the accuracy and stability of indoor positioning in wireless sensor networks. According to the simulation results, in a simple indoor environment with four reference points, when the sensors are deployed in both fixed and random arrangements, the positioning accuracy was improved by 29.09% and 30.20%, respectively, compared to using the KNN algorithm alone. In a complex indoor environment with four reference points, the positioning accuracy was increased by 32.24% and 33.72%, respectively. When the number of reference points increased to five, in a simple environment, the accuracy improvement for both fixed and random deployment was 20.70% and 26.01%, respectively. In a complex environment, the accuracy improvement was 23.88% and 27.99% for fixed and random deployment, respectively. [ABSTRACT FROM AUTHOR]

This paper presents the development of a web-based course timetabling system based on an enhanced genetic algorithm. The enhanced method utilizes a heuristic mutation which concentrates on mutating the infeasible genes to improve the algorithms' exploration and exploitation capability. The method was implemented using a free and open-source application and can be accessed online. Based on the actual datasets from Caraga State University, the enhanced method optimized the use of classroom resources by using a smaller number of rooms. The generated timetable is more efficient as it satisfies not just hard constraints, which are conflicting schedules, but also soft constraints. [ABSTRACT FROM AUTHOR]

The unequal area facility layout problem (UAFLP) is a problem in industrial engineering that aims to minimize material handling costs between a group of unequal-shaped facilities. In this study, Differential Evolution (DE) with continuous variables and Particle Swarm Optimization (PSO) are employed to solve the UAFLP. This study also compares the performance of the DE method (never used to solve UAFLP) and the PSO method (frequently used to solve facility layout problems). The experimental results demonstrate that PSO is effective in solving small-sized datasets, while DE works well in solving datasets with larger areas. [ABSTRACT FROM AUTHOR]

We turn our attention on evaluating the most recent Hubble parameter data measured via the differential evolution of cosmic-chronometers from a deep learning perspective. To achieve this goal, we start our investigation by introducing the selected theoretical setup and compiling the most recent statistical data obtained in cosmology experiments. Then we implement a tuned version of the long-short term memory (LSTM) architecture and run it to predict possible values of the Cosmic Hubble parameter for different red-shift states. Since we observe a good correlation between the observed and predicted datasets of the Hubble parameter, we conclude that the machine learning approaches can play important roles in the future cosmology investigations. [ABSTRACT FROM AUTHOR]

The main purpose of this paper is to study an abstract system which consists of a parabolic hemivariational inequality with a Hilfer fractional evolution equation involving history-dependent operators, which is called a Hilfer fractional differential hemivariational inequality. We first show existence and a priori estimates for the parabolic hemivariational inequality. Then, by using the well-known Bohnenblust-Karlin fixed point theorem, we prove an existence theorem for the Hilfer fractional differential hemivariational inequality. [ABSTRACT FROM AUTHOR]

An accurate method is proposed to measure the electrical parameters describing the experimental charge‐discharge curves of Li‐ion batteries and a pack through an equivalent electric dynamic model based on heuristic optimization algorithms (eg, differential evolution approach). The accuracy of these parameters is critical for on‐board state of charge and Health (SOH) estimations of numerous applications and devices. The procedure is constructed, based on the simple Thévenin electrical model. Charge‐discharge experimental curves collected for different chemistry and capacity batteries are described with the model, revealing better estimation results with respect to other algorithms such as impedance‐based methodologies (decreasing the error until 64.19%) and particle swarm optimization based approaches (decreasing the error until 94.02%). The proposal was assessed with different chemistries, capacities and with a 28 battery pack (LiFeMnPO4, 100 Ah) used in an electric vehicle. [ABSTRACT FROM AUTHOR]

As a typical intelligent device, magnetorheological (MR) dampers have been widely applied in vibration control and mitigation. However, the inherent hysteresis characteristics of magnetic materials can cause significant time delays and fluctuations, affecting the controllability and damping performance of MR dampers. Most existing mathematical models have not considered the adverse effects of magnetic hysteresis characteristics, and this study aims to consider such effects in MR damper models. Based on the magnetic circuit analysis of MR dampers, the Jiles–Atherton (J-A) model is adopted to characterize the magnetic hysteresis properties. Then, a weight adaptive particle swarm optimization algorithm (PSO) is introduced to the J-A model for efficient parameter identifications of this model, in which the differential evolution and the Cauchy variation are combined to improve the diversity of the population and the ability to jump out of the local optimal solution. The results obtained from the improved J-A model are compared with the experimental data under different working conditions, and it shows that the proposed J-A model can accurately predict the damping performance of MR dampers with magnetic hysteresis characteristics. [ABSTRACT FROM AUTHOR]

Benefiting from cloud computing's elasticity, scalability, and pay-per-use model, more and more scientific applications are deployed in or migrated to the cloud. Workflow scheduling still faces many challenges due to the growing scales of workflows and the diversified user QoS requirements. In this work, we propose a Mutation-driven and population Grouping Poor and Rich Optimization algorithm (MG-PRO) for scheduling workflows in the cloud to minimize makespan while satisfying the budget constraints. Specifically, we first adopt the middle-class sub-population into the original Poor and Rich Optimization algorithm (PRO), and develop the update strategies for rich and middle-class sub-populations to increase the randomness and search diversity. Secondly, the update mechanism for rich individuals is enriched, and the middle-class sub-population is guided by elite rich individuals, which enhances the information exchange and sharing among sub-populations. Finally, an evolution-aware mutation strategy is designed, where the mutation probability is adjusted adaptively as the dynamic monitoring of the population update process, and the two-point and triangular crossover-based mutations are used alternately to intervene the evolution trajectory according to the degree of objective optimization, resulting in a better balance between exploration and exploration. Extensive experiments are conducted on well-known scientific workflows with different types and scales through WorkflowSim. The experimental results show that, in most cases, MG-PRO outperforms existing algorithms in terms of constraint satisfiability, solution quality and stability. It can generate near-optimal solutions with the different budget constraints satisfied in a relatively short time, for example, the makespan resulting from MG-PRO is at most 59.95% shorter than other meta-heuristic algorithms, and at least 7.33% shorter than all its peers. [ABSTRACT FROM AUTHOR]

This document is a special issue of the journal Algorithms, featuring selected papers from the journal's editorial board members from 2022 and 2023. The issue includes 16 research papers covering a range of topics such as game theory, fault detection in cellular networks, optimization algorithms, machine learning, cryptocurrency trading, and more. Each paper presents its own unique research findings and methodologies. The issue aims to showcase the diverse research interests and expertise of the journal's editorial board members. [Extracted from the article]

The exploration of premium and new locations is regarded as a fundamental function of every evolutionary algorithm. This is achieved using the crossover and mutation stages of the differential evolution (DE) method. A best-and-worst position-guided novel exploration approach for the DE algorithm is provided in this study. The proposed version, known as "Improved DE with Best and Worst positions (IDEBW)", offers a more advantageous alternative for exploring new locations, either proceeding directly towards the best location or evacuating the worst location. The performance of the proposed IDEBW is investigated and compared with other DE variants and meta-heuristics algorithms based on 42 benchmark functions, including 13 classical and 29 non-traditional IEEE CEC-2017 test functions and 3 real-life applications of the IEEE CEC-2011 test suite. The results prove that the proposed approach successfully completes its task and makes the DE algorithm more efficient. [ABSTRACT FROM AUTHOR]

Building on previous radar‐based human activity recognition (HAR), we expand the micro‐Doppler signature to 6 domains and exploit each domain with a set of handcrafted features derived from the literature and our patents. An adaptive thresholding method to isolate the region of interest is employed, which is then applied in other domains. To reduce the computational burden and accelerate the convergence to an optimal solution for classification accuracy, a holistic approach to HAR optimisation is proposed using a surrogate model‐assisted differential evolutionary algorithm (SADEA‐I) to jointly optimise signal processing, adaptive thresholding and classification parameters for HAR. Two distinct classification models are evaluated with holistic optimisation: SADEA‐I with support vector machine classifiers (SVM) and SADEA‐I with AlexNet. They achieve an accuracy of 89.41% and 93.54%, respectively. This is an improvement of ∼11.3% for SVM and ∼2.7% for AlexNet when compared to the performance without SADEA‐I. The effectiveness of our holistic approach is validated using the University of Glasgow human radar signatures dataset. This proof of concept is significant for dimensionality reduction and computational efficiency when facing a multiplication of radar representation domains/feature spaces and transmitting/receiving channels that could be individually tuned in modern radar systems. [ABSTRACT FROM AUTHOR]

The fast growth of under developing mobile applications in recent years has emerged a diversity of delay-sensitive applications such as multimedia streaming, virtual reality, augmented reality, and online gaming applications to facilitate daily activities in different aspects of human life. Edge computing has been raised as an Internet-based distributed computing model to enable mobile devices to offload tasks to nearby edge servers rather than transfer them to remote cloud servers. A joint auto-scaling and task offloading approach in edge/cloud computing is proposed in this paper. Due to dynamic changes in usage and access to mobile applications over time, it requires addressing their workload fluctuations as challenging issues. The future workload is predicted using long short-term memory (LSTM) network, supported with the differential evolution (DE) algorithm for selecting the LSTM hyperparameters. A fuzzy Q-learning technique is also utilized to make scaling decisions at runtime, and a learning automata-based technique is used to make decisions on offloading tasks of mobile devices to edge/cloud layers. The proposed approach is validated using the iFogSim simulator under synthetic and real-world patterns. The results show that it achieves better performance in terms of execution time, energy consumption, and delay violation compared to the baseline approaches. [ABSTRACT FROM AUTHOR]

The substantial challenge posed by inherent hydrological forecasting uncertainty has critical implications for the optimization of real-time reservoir operations. In response, this study introduces a stochastic framework explicitly devised to comprehensively quantify the ramifications of hydrological forecasting uncertainty, notably its temporal correlations, on the outcomes of real-time reservoir optimization and risk assessment. Furthermore, this framework seeks to delineate the pivotal influence of incorporating or neglecting these temporal dynamics on the eventual results, while concurrently elucidating the underlying mechanisms governing these discernible influences. The framework adopts a comprehensive approach to simulating hydrological forecast uncertainty through ensemble forecasts and scenario trees, employing three methods (two Monte Carlo sampling-based methods and one Gaussian copula method) to generate inflow forecast ensembles. To improve the adaptability to uncertainties in inflow forecasts, the framework incorporates a transformation of the generated ensembles into scenario trees, serving as input for a stochastic optimization model that derives the final optimal decision based on optimizing the expected value of the objective function for all scenarios. Additionally, a parallel differential evolution algorithm is proposed to solve the stochastic optimization model efficiently. Risk assessment is performed to capture the uncertainty and corresponding risk associated with the reservoir optimal decision. The proposed framework is demonstrated in a flood control reservoir system in China, where several numerical experiments are conducted to explore the effect of forecast uncertainty level and temporal correlation on real-time reservoir optimal operation. Results show that the temporal correlation of inflows must be considered in both inflow stochastic simulation and reservoir stochastic optimization to avoid overestimating or underestimating operational risk, potentially leading to operation failures. By examining the risk simulation surface, reservoir operators can evaluate the robustness of operational decisions and make more reliable final decisions. [ABSTRACT FROM AUTHOR]

The tire pattern structure has a serious influence on automobile performance. In order to improve the tire performance, the strain energy density of tire tread under different working conditions was analyzed. The finite element model of longitudinal groove pattern of automobile tire was established to analyze the strain energy density under different working conditions, and the influence of parameter changes on the strain energy density of tire on dry road and slippery road was analyzed respectively. The multi-objective optimization model was established based on the finite element analysis results, and the function constraint conditions were set. Differential evolution method (DE) and non-dominated sorting genetic optimization method (NSGA-II) are used to design a multi-objective optimization program, which can quickly determine the optimization scheme after the input data. The finite element model was modified according to the optimization scheme and the strain energy density was analyzed. The results show that the performance of the optimized tire is greatly improved. The combination of finite element analysis technology and multi-objective optimization method can effectively reduce a lot of manpower, material resources and financial resources in mechanical product structure optimization. [ABSTRACT FROM AUTHOR]

The Whale Optimization Algorithm (WOA) is an efficient meta-heuristic algorithm inspired by the feeding behavior of whales. Although it has successfully solved many optimization problems, it still suffers from premature convergence and poor accuracy when solving complex problems. To address these issues, this paper proposes an improved whale optimization algorithm called EIWOA to improve the search efficiency and accuracy. First, we introduce a new global search mechanism and encircling prey strategy in EIWOA, which utilizes differential evolution, and sine-cosine search strategy to improve the global search efficiency and avoid premature convergence. Second, we use a lévy flight-based spiral update position strategy to improve the local search efficiency of the algorithm, thus improving the convergence speed and accuracy. Again, we introduce a balancing factor with fluctuating decay properties into EIWOA to better balance exploration and exploitation. Finally, we introduce a dynamic opposite learning-based whale-fall strategy in EIWOA to equip the algorithm with the ability to jump out of the local optimum. The qualitative analysis of the algorithm shows that the EIWOA algorithm converges quickly, is highly accurate, and has the ability to jump out of the local optimum. In order to validate the performance of the proposed EIWOA algorithm, we evaluated the algorithm on CEC2017 benchmark function and four real world engineering problems. We also conduct a comparative study of the EIWOA algorithm with EWOA, an excellent variant of WOA, as well as some excellent meta-heuristics developed recently. The results of numerical experiments demonstrate the superiority of the proposed EIWOA, which is further confirmed statistically by the results of Friedman's test and Wilcoxon signed rank test. The proposed EIWOA algorithm is significantly better than WOA and other competing algorithms in terms of convergence speed, accuracy and optimization ability in dealing with complex optimization problems. [ABSTRACT FROM AUTHOR]

The ventilation resistance coefficients of a mine is vital to ventilation system safety management, diagnosis, and intelligentization. Airflow typically serves as the basis for inverting ventilation resistance coefficients. However, the issue of non-uniqueness in conventional nonlinear optimization methods affects the accuracy of inversion. Therefore, this study introduces a novel optimization approach based on deep reinforcement learning (DRL) to invert resistance coefficients. In this methodology, inversion is regarded as a Markov decision process, with the ventilation network solving model (VNSM) embedded within the DRL environment. We design an agent utilizing deep neural networks, which dynamically adjusts the resistance coefficient state variables by interacting with the VNSM to enhance the consistency between theoretical and measured airflow. The consistency corresponds to the agent's reward. The proximal policy optimization is employed to optimize the agent's policy. In field experiment, the MAE between the airflow calculated and the measured airflow is 0.354, with MSE of 0.287, RMSE of 0.536, and MRE of 0.013. Compared with the standard genetic algorithm, differential evolution algorithm, and evolution strategy algorithm, the DRL method shows lower MRE, MAE, RMSE, and MSE values by 23.5%, 15.3%, 14.1%, and 26.4%, respectively. Additionally, DRL method exhibits smaller sensitivity differences for different roadways than others algorithm. [ABSTRACT FROM AUTHOR]

Latin hypercube design (LHD) is one of the most frequently used sampling methods. However, most LHDs generate data samples in a manner that hinders computational efficiency and space-filling performance when high dimensions and large samples are involved. Therefore, a sequential recursive evolution Latin hypercube design (RELHD) is proposed in this article, which adopts a permutation inheritance algorithm to update and optimize the LHD. A recursive split algorithm is also proposed and used to enhance the computational efficiency by dividing the sample set into smaller subsets. Numerical experiments demonstrate that the space-filling quality of the RELHD compares well with the enhanced stochastic evolutionary algorithm (ESE) in complex problems with large samples and high dimensions, with RELHD having a significantly higher computational efficiency than ESE. Finally, the sequential approach of RELHD proves to be a more efficient strategy when dealing with sampling-based analysis problems. [ABSTRACT FROM AUTHOR]

Due to the mental stress of the human, the negative effects are known to be recent decades. Early detections of high level stresses are necessary to stop harmful consequences. Studies have proposed on wearable technologies which detect human stress. This study proposes stress detection systems which use physiological signals of people collected by wearable technologies and attached to human bodies. They can carry it during their daily routine. This work's proposed system includes removal of artifacts in bio signals and feature extractions from these cleaned signals. Since, DL (deep learning) based models are proven to be the best for these analyses, this article uses a random differential GWO (Grey wolf optimization) algorithm for feature extraction and a ML (machine learning) algorithm called RF (random forest) has been used for classification of the human body parameters like activities of the heart, conductance in skins and corresponding accelerometer signals. The proposed stress detection system is implemented with the real time data gathered from 21 participants. This approach can detect the stress of a human and prevent it from early stages with necessary lectures to avoid the negative effects. [ABSTRACT FROM AUTHOR]

In the era of e-commerce, accurately predicting user purchase intentions has become a crucial factor for enhancing sales efficiency and optimizing the customer experience. Addressing the limitations of traditional ensemble strategies, which often suffer from subjective biases during the model design phase, this paper introduced an adaptive evolutionary ensemble learning model to predict user purchase intentions. This model adaptively selected the optimal base learners and meta-learners, incorporating both the predictive information from the base learners and the differential information between features to expand the feature dimensions, enhancing prediction accuracy. Moreover, to further refine the predictive capabilities of the model, this paper designed a binary adaptive differential evolution algorithm for feature selection, aiming to identify features that significantly influence the prediction outcome. Research results show that the binary adaptive differential evolution algorithm outperforms traditional optimization algorithms in global searches and feature selection. Compared to six common ensemble models and the DeepForest model, the proposed evolutionary ensemble model achieves a 2.76% and 2.72% increase in AUC value, respectively, and effectively mitigates the impacts of data imbalance [ABSTRACT FROM AUTHOR]

The influence maximization problem that has caused great attention in social network analysis aims at selecting a small set of influential spreaders so that the information cascade triggered by the seed set is maximized. The majority of the existing works mainly focus on developing single-stage seeding strategies that would ignite all the seeds before the influence spread. However, it cannot depict the scenarios of the practical, where ones would like to make further decisions based on observed activation. In this paper, we investigate the policies for the intractable sequential influence maximization problem. A Q-learning-driven discrete differential evolution algorithm based on the reinforcement Q-learning model, which is treated as a parameter controller to adaptively adjust the parameters during the evolution of the algorithm, is proposed. The policy distributes the seeding actions over the spreading process by estimating the latest node status of the network dynamically. Extensive simulations are conducted on six social networks of the practical, and the findings demonstrate the superiority and effectiveness of the hybrid meta-heuristic algorithm compared with the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

The combined use of LiDAR (Light Detection And Ranging) scanning and field inventories can provide spatially continuous wall-to-wall information on forest characteristics. This information can be used in many ways in forest mapping, scenario analyses, and forest management planning. This study aimed to find the optimal way to obtain continuous forest data for Catalonia when using kNN imputation (kNN stands for “k nearest neighbors”). In this method, data are imputed to a certain location from k field-measured sample plots, which are the most similar to the location in terms of LiDAR metrics and topographic variables. Weighted multidimensional Euclidean distance was used as the similarity measure. The study tested two different methods to optimize the distance measure. The first method optimized, in the first step, the set of LiDAR and topographic variables used in the measure, as well as the transformations of these variables. The weights of the selected variables were optimized in the second step. The other method optimized the variable set as well as their transformations and weights in one single step. The two-step method that first finds the variables and their transformations and subsequently optimizes their weights resulted in the best imputation results. In the study area, the use of three to five nearest neighbors was recommended. Altitude and latitude turned out to be the most important variables when assessing the similarity of two locations of Catalan forests in the context of kNN data imputation. The optimal distance measure always included both LiDAR metrics and topographic variables. The study showed that the optimal similarity measure may be different for different regions. Therefore, it was suggested that kNN data imputation should always be started with the optimization of the measure that is used to select the k nearest neighbors. [ABSTRACT FROM AUTHOR]

Dynamic differential annealed optimization (DDAO) is a recently developed physics-based metaheuristic technique that mimics the classical simulated annealing mechanism. However, DDAO has limited search abilities, especially when solving complicated and complex problems. A unique variation of DDAO, dubbed as mDDAO, is developed in this study, in which opposition-based learning technique and a novel updating equation are combined with DDAO. mDDAO is tested on 10 different functions from CEC2020 and compared with the original DDAO and nine other algorithms. The proposed mDDAO algorithm performance is evaluated using 10 numerical constrained functions from the recently released CEC 2020 benchmark suite, which includes a variety of dimensionally challenging optimisation tasks. Furthermore, to measure its viability, mDDAO is employed to solve feature selection problems using fourteen UCI datasets and a real-life Lymphoma diagnosis problem. Results prove that mDDAO has a superior performance and consistently outperforms counterparts across benchmarks, achieving fitness improvements ranging from 1% to 99.99%. In feature selection, mDDAO excels by reducing feature count by 23% to 79% compared to other methods, enhancing computational efficiency and maintaining classification accuracy. Moreover, in lymphoma diagnosis, mDDAO demonstrates up to 54% higher average fitness, 18% accuracy improvement, and 86% faster computation times. [ABSTRACT FROM AUTHOR]

Unstable processes are challenging to control because they have one or more positive poles that produce unrestrained dynamic activity. Controlling such unstable plants becomes more challenging with the occurrence of the delay. This article presents a novel direct synthesis based sliding mode controller design for unstable second order plus dead-time processes. A sliding surface with three parameters has been considered. The continuous control law, which is responsible for maintaining the system mode to the desired sliding surface mode, has been obtained using the direct synthesis approach. The discontinuous control law parameters have been obtained using the differential evolution optimization technique. A desired reference model is considered for the direct synthesis method, and an objective function is constituted in terms of performance measure (integral absolute error) and control effort measure (total variation of controller output) for the optimization approach. Illustrative examples show the superiority of the proposed controller design method over recently reported literature, especially in terms of load rejection. The proposed controller approach is further extended to control the temperature of a nonlinear chemical reactor. Furthermore, the robustness of the proposed controller is also investigated for plant parametric uncertainty. [ABSTRACT FROM AUTHOR]

Purpose: A control scheme is presented to attenuate the vibration of the large flexible space structures (LFSS), which includes the optimization of actuator positions and the design of the robust vibration control algorithm. In contrast to the existing literature, the optimization criterion does not depend on the controller parameters, and robust adaptive dynamic programming (RADP) can effectively suppress the vibration of the LFSS under mixed continuous disturbance and model uncertainties. Method: First, the optimization criteria of optimal placements of the actuator, which can maximize the actuating efficiency, decoupling input voltage and modal contribution factor, are proposed. On this basis, the adaptive parameter differential evolution (APDE) algorithm is employed to speed up the optimization process. Subsequently, the RADP is designed to enhance the control effect and robustness of the system under external disturbances and model uncertainties. Finally, the RADP-based vibration control performance under optimal and random placement layouts is compared. Results and Conclusion: The effectiveness of the proposed optimization criterion, APDE method and RADP-based vibration control algorithm is verified by simulation. The proposed optimization criteria save the actuator energy consumption and enhance the adaptability of the optimization. The optimal actuator positions can be swiftly identified by the APDE algorithm. In addition, the RADP-based vibration control algorithm improves the control effect, reduces decay time and enhances the robustness of the LFSS under mixed continuous excitation and model uncertainties. [ABSTRACT FROM AUTHOR]

Due to the unreasonable connectivity of wireless sensor networks, there are errors in calculating the distance between the location-unaware nodes and the location-aware nodes. Therefore, a novel method combining improved particle swarm optimization and improved differential evolution (IPSO-IDE) is proposed. Firstly, the learning factor of the particle swarm optimization method is modified, and the disturbance term is added to prevent premature, thus strengthening its exploration ability. Then, the adaptive mutation and crossover mechanism are applied to the differential evolution method so that each chromosome can mutate and cross adaptively with individual fitness and average fitness, to improve its exploitation ability. Finally, the improved particle swarm optimization (IPSO) and differential evolution method (IDE) are run in parallel in the node positioning model. After fixed number of iterations, the optimal individual information of the two methods is shared to obtain the IPSO-IDE method. Therefore, IPSO-IDE has the advantages of both. The results show that the IPSO-IDE algorithm not only has some advantages in time complexity, but also improves the positioning accuracy by 27.2%, 8.7%, 4.5% and 5.7% respectively, compared with particle swarm optimization with a variable neighborhood algorithm, modified particle swarm optimization algorithm, hybrid particle swarm optimization and differential evolution algorithm and artificial bee colony algorithm. [ABSTRACT FROM AUTHOR]

The cracking behavior of lignite during drying–wetting cycles impacts the efficiency of coal mining, underground coal gasification, and coalbed methane development. Lignite is known to have high water content, and it exhibits strong hydrophilic properties, resulting in a weak and rigid gel structure caused by water interactions with coal matrix molecules. Even at ambient temperatures, dehydration causes violent shrinkage of the lignite matrix and numerous fractures. Due to the shallow burial depth of lignite, it is often mined in open pits, and it often undergoes cycles of dehydration and rewetting in the open-air environment, resulting in significant changes to its physical structure and mechanical properties. Thus, a detailed characterization of this evolution process is necessary. In this study, lignite samples with different compositional properties were collected from the Shengli Coalfield in the Erlian Basin, and their physical structure and mechanical properties were characterized using nondestructive testing methods such as nuclear magnetic resonance, environmental scanning electron microscopy, X-ray computed tomography, and atomic force microscopy. The results provide insights into the impact of material composition on the syneresis behavior and destruction process of lignite. Upon removal of water from the coal matrix, rapid shrinkage occurs, resulting in the simultaneous formation of numerous fractures. Significant differences were observed in the contraction characteristics and fracture propagation patterns among the various lignite lithotypes. Xylite lignite exhibited the highest degree of contraction following dehydration, with the macropore reduction rate being the most significant. Matrix lignite showed a lower degree of syneresis compared to xylite lignite, while fusain-rich lignite demonstrated the weakest syneresis ability. The homogeneous xylite lignite shrinks as a whole after dehydration, and its fractures are long and straight, with good orientation. Fractures formed in the detrital humic groundmass of matrix lignite are short, convoluted, and poorly oriented. Fusain-rich lignite fractures are thin, and straight, with the best orientation. The expansion of fractures can cause geological disasters such as land subsidence, collapse, and landslides. Fusain hinders fractures' expansion and acts as a "skeleton" to support the lignite structure. The presence of fusain is advantageous to maintaining the steadiness of the lignite. The differential evolution of the physical structure of lignite during the drying–wetting cycles, which aids to avert and mitigate disasters linked to dehydration cracking in different lignite-related engineering geological fields, is discussed in this paper. [ABSTRACT FROM AUTHOR]

In this article, a new hybrid improved differential evolution and Nelder-Mead (IDE-NM) is introduced for optimizing the multi-objective machining process during the turning operation under three modes of lubrication conditions. The fitness functions are the tangential cutting force, the surface roughness, and the cutting power. Five mixed design variables are considered in the optimization procedure including the cutting speed, feed rate, and depth of cut, mode of lubrication, and the type of cutting material. The mathematical expressions of the three objectives are created based on experimental results and modeled using the artificial neural network (ANN). In the first step, the proposed method is examined by solving seven mechanical engineering design problems. The comparison results illustrate that the IDE-NM algorithm outperforms other state-of-the-art optimization methods considered in the literature. Moreover, for the turning operation problem, the results of IDE-NM are compared with those of four recent metaheuristics. The results show that the proposed method outperforms the four compared algorithms in terms of robustness, high success rate, and can provide effective solutions. [ABSTRACT FROM AUTHOR]

The optimization of the rule base of a fuzzy logic system (FLS) based on evolutionary algorithm has achieved notable results. However, due to the diversity of the deep structure in the hierarchical fuzzy system (HFS) and the correlation of each sub fuzzy system, the uncertainty of the HFS’s deep structure increases. For the HFS, a large number of studies mainly use fixed structures, which cannot be selected automatically. To solve this problem, this paper proposes a novel approach for constructing the incremental HFS. During system design, the deep structure and the rule base of the HFS are encoded separately. Subsequently, the deep structure is adaptively mutated based on the fitness value, so as to realize the diversity of deep structures while ensuring reasonable competition among the structures. Finally, the differential evolution (DE) is used to optimize the deep structure of HFS and the parameters of antecedent and consequent simultaneously. The simulation results confirm the effectiveness of the model. Specifically, the root mean square errors in the Laser dataset and Friedman dataset are 0.0395 and 0.0725, respectively with rule counts of rules is 8 and 12, respectively. When compared to alternative methods, the results indicate that the proposed method offers improvements in accuracy and rule counts. [ABSTRACT FROM AUTHOR]

The selection of critical features from microarray data as biomarkers holds significant importance in disease diagnosis and drug development. It is essential to reduce the number of biomarkers while maintaining their performance to effectively minimize subsequent validation costs. However, the processing of microarray data often encounters the challenge of the "curse of dimensionality". Existing feature-selection methods face difficulties in effectively reducing feature dimensionality while ensuring classification accuracy, algorithm efficiency, and optimal search space exploration. This paper proposes a hybrid feature-selection algorithm based on an enhanced version of the Max Relevance and Min Redundancy (mRMR) method, coupled with differential evolution. The proposed method improves the quantization functions of mRMR to accommodate the continuous nature of microarray data attributes, utilizing them as the initial step in feature selection. Subsequently, an enhanced differential evolution algorithm is employed to further filter the features. Two adaptive mechanisms are introduced to enhance early search efficiency and late population diversity, thus reducing the number of features and balancing the algorithm's exploration and exploitation. The results highlight the improved performance and efficiency of the hybrid algorithm in feature selection for microarray data analysis. [ABSTRACT FROM AUTHOR]

Battery-swapping is a mechanism that involves exchanging discharged batteries for charged ones. Battery-swapping and charging stations (BSCS) enhance operational flexibility and interact with electric vehicle (EVs) batteries. An optimal battery-swapping mechanism is proposed for electric vehicles using a hybrid approach. The proposed intelligent method is a wrapper of the radial basis function neural network (RBFNN) and the war strategy optimization (WSO) algorithm. Hence, it is known as the WSO-RBFNN method. The key objectives of the proposed method are to reduce the total cost by defining an enhanced charging schedule for EV batteries, the number of batteries pulled from inventory to fulfill all incoming EV swap orders, the risk of charging damage while using high-rate chargers, and the cost of electricity at various times of the day. The proposed method minimizes the net costs based on EV energy consumption and travel time. WSO is exploited to attain the optimum control parameters of RBFNN. The performance of the proposed method is measured in MATLAB and compared with existing methods. The simulation outcome shows that the proposed method-related cost is lower than the existing methods. The proposed method provides a low cost of 32.5 $ and a high efficiency of 90% compared with existing methods like differential evolution (DE), genetic algorithm (GA), and particle swarm optimization (PSO). [ABSTRACT FROM AUTHOR]

Differential evolution (DE) is a population-based optimization algorithm widely used to solve a variety of continuous optimization problems. The self-adaptive DE algorithm improves the DE by encoding individual parameters to produce and propagate better solutions. This paper proposes a self-adaptive differential evolution algorithm with dynamic fitness-ranking mutation and pheromone strategy (SDE-FMP). The algorithm introduces the dynamical mutation operation using the fitness rank of the individuals to divide the population into three groups and then select groups and their vectors with adaptive probabilities to create a mutant vector. Mutation and crossover operations use the encoded scaling factor and the crossover rate values in a target vector to generate the corresponding trial vector. The values are changed according to the pheromone when the trial vector is inferior in the selection, whereas the pheromone is increased when the trial vector is superior. In addition, the algorithm also employs the resetting operation to unlearn and relearn the dominant pheromone values in the progressing search. The proposed SDE-FMP algorithm using the suitable resetting periods is compared with the well-known adaptive DE algorithms on several test problems. The results show that SDE-FMP can give high-precision solutions and outperforms the compared methods. [ABSTRACT FROM AUTHOR]

Silicon content of molten iron is an important indicator in the blast furnace ironmaking process. Accurate prediction of silicon content is very important for monitoring the operating condition of the blast furnace and the quality of the molten iron. However, accurate and effective online prediction of silicon content is a challenging task due to the complex and high-dimensional nonlinear relationship between silicon content and process variables. Therefore, a two-stage multiobjective evolutionary ensemble learning algorithm is proposed to achieve a high-accuracy and low-complexity prediction model using support vector machine (SVR) as the base learner. In the first stage, a non-dominated sorting differential evolution algorithm with dynamic resource allocation (DRA-NSDE) is proposed to generate a set of non-dominated solutions (SVRs) with the objectives of accuracy and complexity. In the second stage, a stacking method based on clustering and differential evolution (CDE-Stacking) is proposed to select base learners with better diversity and construct the ensemble model. The effectiveness of the proposed DRA-NSDE and CDE-Stacking strategies is verified through a series of numerical experiments. The experimental results show that the proposed algorithm outperforms the rival methods on both the UCI benchmark data set and the actual blast furnace data set. In addition, the analysis of model complexity shows that the proposed model can achieve higher prediction accuracy with relatively low model complexity, which indicates that the algorithm is very suitable for the online prediction of silicon content in actual blast furnace ironmaking process. [ABSTRACT FROM AUTHOR]

In this paper, a novel Adaptive Parameter Strategy Differential Evolution (APSDE) algorithm is proposed to overcome the parameters dependence and avoid local optima. The Parameter Update Mechanism (PUM), which has three different strategies, is used to reduce the dependence on parameters of DE. The Adaptive Proportion Adjustment Mechanism (APAM) is used to balance the proportion of PUM strategies in different development terms of exploitation and exploration, and the Random Restart Mechanism (RRM) is used to improve population diversity when exploitation is in stagnation. The proposed algorithm is verified in the CEC2018 test functions and the results show that APSDE has good abilities of exploitation, exploration, convergence, and stability. Secondly, Midcourse Guidance Maneuver Decision-making (MGMD) in Beyond Visual Range (BVR) air combat is studied and transformed into a single objective variational optimization problem, a MGMD system based on APSDE is established. Finally, the simulation of MGMD is carried out. The APSDE ranks first in the typical MGMD scenario experiment. In the adaptive Midcourse guidance confrontation, the winning rate of APSDE is 54%, and the statistical results show that the APSDE has an excellent MGMD ability. [ABSTRACT FROM AUTHOR]

Differential evolution (DE) is one of the most efficient evolution algorithms (ES) for dealing with nonlinear, complicated and difficult global optimization problems. The main contribution of this paper can be summarized in three directions: Firstly, a novel crossover rate (CR) generation scheme based on the zscore value of fitness, named fcr, is introduced. For a minimization problem, the proposed CR generation strategy always assigns a smaller CR value to individual with smaller fitness value. Therefore, the parameters of individuals with better fitness are inherited by their offspring with high probability. In the second direction, the control parameters are adjusted by unused bimodal settings in which each parameter setting is selected according to the evolution status of individual. The third direction of our work is introducing the L1 norm distance as the weights for updating the mean value of crossover rate and scale factor. Theoretically, compared with L2 norm, L1-norm is more efficient to suppress outliers in the difference vector. These modifications are first integrated with the mutation strategy of JADE, then a modified version, named JADEfcr, is proposed. In addition, to improve the optimization ability further, another variant LJADEfcr by using a linear population reduction mechanism is considered. So as to confirm and examine the performance of JADEfcr and LJADEfcr, numerical experiments are conducted on 29 optimization problems defined by CEC2017 benchmark. For JADEfcr, its experimental results are made a comparison with twelve state-of-the-art algorithms. The comparative study demonstrates that in terms of robustness, stability and solution quality, JADEfcr are better and highly competitive with these well-known algorithms. For LJADEfcr, its results are compared with JADEfcr and other nine powerful algorithms including four recent algorithms and five top algorithms on CEC2017 competition. Experimental results indicate that LJADEfcr is superior and statistically competitive with these excellent algorithms in terms of robustness, stability and the quality of the obtained solutions. [ABSTRACT FROM AUTHOR]