Search

This paper introduces a methodology designed to augment the inverse design optimization process in scenarios constrained by limited compute, through the strategic synergy of multi-fidelity evaluations, machine learning models, and optimization algorithms. The proposed methodology is analyzed on two distinct engineering inverse design problems: airfoil inverse design and the scalar field reconstruction problem. It leverages a machine learning model trained with low-fidelity simulation data, in each optimization cycle, thereby proficiently predicting a target variable and discerning whether a high-fidelity simulation is necessitated, which notably conserves computational resources. Additionally, the machine learning model is strategically deployed prior to optimization to compress the design space boundaries, thereby further accelerating convergence toward the optimal solution. The methodology has been employed to enhance two optimization algorithms, namely Differential Evolution and Particle Swarm Optimization. Comparative analyses illustrate performance improvements across both algorithms. Notably, this method is adaptable across any inverse design application, facilitating a synergy between a representative low-fidelity ML model, and high-fidelity simulation, and can be seamlessly applied across any variety of population-based optimization algorithms.

Effective and well-structured customer segmentation enables organizations to accurately identify and comprehend the distinct characteristics and needs of various customer groups, thereby facilitating the development of more targeted marketing strategies. Contemporary artificial intelligence technologies have emerged as the predominant tools for customer segmentation, owing to their robust capabilities in analyzing complex datasets and extracting profound customer insights. This paper proposes a customer segmentation framework within the realm of digital marketing, which integrates a reinforcement learning-based differential evolution algorithm with K-means clustering using dimensionality reduction techniques to address challenges in the customer segmentation process. Initially, a correlation matrix is used to identify redundant noise and multicollinear features within customer feature groups, and Principal Component Analysis is applied for denoising and dimensionality reduction to enhance the ability of the model to identify potential features. Subsequently, a parameter adaptive adjustment method based on Q-learning is proposed, which significantly augments the clustering performance of K-means. Ultimately, the effectiveness of the proposed method is validated using a Kaggle dataset, and the elbow method is employed to ascertain the optimal number of clusters. Based on the cluster category centers, the typical characteristics of different customer types are analyzed. Furthermore, four widely recognized machine learning methods are employed to classify the clustering results, achieving over 95% classification accuracy on the test set. The experimental results demonstrate that the proposed model exhibits a high degree of customer characteristic identification and segmentation, which not only enhances marketing efficiency and customer satisfaction but also fosters corporate profit growth through the strategic formulation of various marketing initiatives. [ABSTRACT FROM AUTHOR]

Stochastic simulations are often used to determine the crossover rates and step size of evolutionary algorithms to avoid the tuning process, but they cannot fully utilize information from the evolutionary process. A two-stage adaptive differential evolution algorithm (APDE) is proposed in this article based on an accompanying population, and it has unique mutation strategies and adaptive parameters that conform to the search characteristics. The global exploration capability can be enhanced by the accompanying population to achieve a balance between global exploration and local search. This algorithm has proven to be convergent with a probability of 1 using the theory of Markov chains. In numerical experiments, the APDE is statistically compared with nine comparison algorithms using the CEC2005 and CEC2017 standard set of test functions, and the results show that the APDE is statistically superior to the comparison methods. [ABSTRACT FROM AUTHOR]

This paper presents a bi-objective solution approach to address the production scheduling challenge encountered by manufacturers in a shared manufacturing environment. In such scenarios, manufacturers are required to manage orders received through a cloud platform (referred to as cloud orders) while simultaneously fulfilling orders from their long-term and regular clients (local orders). The problem is to efficiently coordinate the production of both types of orders within shared manufacturing facilities. We formulate the problem into a bi-objective mixed integer programming model aimed at simultaneously minimizing the delivery time of cloud orders and mitigating the disruptions to local order production caused by cloud orders. This solution approach comprises three key components: computation of cloud orders' starting times, construction of available time intervals of manufacturing facilities, and a bi-objective heuristic. This heuristic combines an enhanced hybrid discrete differential evolution with a modified forward–backward earliest starting time algorithm. We introduce an advanced population initialization technique, a novel individual update strategy, and an adaptive local search mechanism based on Pareto-dominance principles to improve the search capabilities of the algorithm towards discovering Pareto non-dominated solutions. Computational results show that the proposed approach outperforms the existing algorithm in most test instances in terms of five common metrics. Insights are discussed, highlighting the practical implications and potential benefits of the proposed approach for shared manufacturing scheduling. [ABSTRACT FROM AUTHOR]

Featured Application: Larger grounding systems are often designed using the finite element method. A finite element method model can only be made if the data of the soil structure are known in advance. A problem arises when the soil is not homogeneous and the horizontally layered soil model poorly represents the actual conditions in the soil. This paper introduces an approach that describes the use of an analytical horizontally layered soil model for inhomogeneous soil. In this approach, deeper layers of the soil are also considered, which are enabled using higher distances d in Wenner's method, as in the case of simply dividing the area into smaller parts. Our approach is unique. A new approach using a horizontally layered analytical soil model for inhomogeneous soil is presented. The presented approach also considers deeper soil layers, which is not the case when simply dividing the area of interest into smaller subareas. The finite element method model was used to prepare test data because, in such a case, the soil parameters are known. Six lines simulating Wenner's method were used, and their results were combined appropriately to determine the soil parameters of nine subareas. To determine the soil parameters in the scope of each subarea, different optimization methods were used and compared to each other. The results were analyzed, and Artificial Bee Colony was selected as the most appropriate method among those tested. Additionally, the convergence of the methods was analyzed, and Memory Assistance is presented, with the aim of shortening the calculation time. In this study, three-, four-, five-, and six-layered soil models were tested, and it is concluded that the three-layered model is most appropriate. A finite element method model based on the soil determination results was constructed to verify the results. The results of the Wenner's method simulation in the cases of the test data and final model were compared to confirm the accuracy of the presented method. [ABSTRACT FROM AUTHOR]

As the global population grows, vertical farming offers a promising solution by using vertically stacked shelves in controlled environments to grow crops efficiently within urban areas. However, the shading effects of farm structures make artificial lighting a significant cost, accounting for approximately 67 % of total operational expenses. This study presents a novel approach to optimizing the deployment of laser illumination in rotating vertical farms by incorporating structural design, light modeling, and photosynthesis. By theoretically analyzing the beam pattern of laser diodes and the dynamics in the coverage area of rotating farm layers, we accurately characterize the light conditions on each vertical layer. Based on these insights, we introduce a new criterion, cumulative coverage, which accounts for both light intensity and coverage area. Then, an optimization framework is formulated, and a swarm intelligence algorithm, Differential Evolution (DE) is used to solve the optimization while considering the structural and operational constraints. It is found that tilting lights and placing them slightly off-center are more effective than traditional vertically aligned and center-aligned deployment. Our results show that the proposed strategy improves light coverage by 4 % compared to the intensity-only optimization approach, and by 10 % compared to empirical methods. This study establishes the first theoretical framework for designing energy-efficient artificial lighting deployment strategies, providing insights into enhancing the efficiency of vertical farming systems. [ABSTRACT FROM AUTHOR]

Proton exchange membrane fuel cells (PEMFCs) are complex, nonlinear systems whose performance depends on several interrelated parameters. Accurate estimation of these parameters is crucial for enhancing the efficiency and reliability of PEMFCs. In this paper, we used Fitness Deviation-based Differential Evolution (FD-DE) algorithm to optimally identify the unknown parameters of PEMFC models. An adaptive parameter control with wavelet basis function and Gaussian distribution, a hybrid trial vector generation strategy using t-distribution based perturbation, and a dimensional replacement mechanism for maintaining population diversity are introduced in the FD-DE algorithm. The innovations in these algorithms tackle the common problems in differential evolution algorithms, including premature convergence and loss of diversity. The proposed FD-DE algorithm is validated on twelve different PEMFC case studies under different operating conditions and compared with several state-of-the-art algorithms, including other DE variants and non-DE algorithms. For that purpose, the optimization targets seven parameters ξ 1 , ξ 2 , ξ 3 , ξ 4 , λ , R c and B to closely match the polarization curves to those specified in the manufacturer datasheet, focusing initially on six primary stacks: BCS 500-W PEM, STD 250-W PEM, Nedstack PS6 PEM, 500W SR-12PEM, H-12 PEM, and HORIZON 500W PWM. The optimization approach is to minimize the sum of squared errors (SSE) between the predicted stack voltages of the model and the experimentally measured results. Absolute error (AE), relative error (RE), and mean bias error (MBE) are assessed across different datasheets. A comparative statistical analysis is made among several DE variants FD-DE, TDE, MadDE, LSHADE, LSHADE-cnEpSin, jSO, PaDE, and non-DE algorithms ACS, SSA, and TEO. The results show that the FD-DE algorithm performs consistently better than existing strategies in terms of accuracy, convergence speed and stability, and provides better parameter estimation with lower errors. The effectiveness of the FD-DE algorithm in solving complex optimization problems and its potential to enhance PEMFC modeling and analysis is shown in this work. [ABSTRACT FROM AUTHOR]

The wind farm layout optimization problem (WFLOP) aims to maximize wind energy utilization efficiency and mitigate energy losses caused by wake effects by optimizing the spatial layout of wind turbines. Although Genetic Algorithms (GAs) and Particle Swarm Optimization (PSO) have been widely used in WFLOP due to their discrete optimization characteristics, they still have limitations in global exploration capability and optimization depth. Meanwhile, the Differential Evolution algorithm (DE), known for its strong global optimization ability and excellent performance in handling complex nonlinear problems, is well recognized in continuous optimization issues. However, since DE was originally designed for continuous optimization scenarios, it shows insufficient adaptability under the discrete nature of WFLOP, limiting its potential advantages. In this paper, we propose a Fractional-Order Difference-driven DE Optimization Algorithm called FODE. By introducing the memory and non-local properties of fractional-order differences, FODE effectively overcomes the adaptability issues of advanced DE variants in WFLOP's discreteness while organically applying their global optimization capabilities for complex nonlinear problems to WFLOP to achieve more efficient overall optimization performance. Experimental results show that under 10 complex wind farm conditions, FODE significantly outperforms various current state-of-the-art WFLOP algorithms including GA, PSO, and DE variants in terms of optimization performance, robustness, and applicability. Incorporating more realistic wind speed distribution and wind condition data into modeling and experiments, further enhancing the realism of WFLOP studies presented here, provides a new technical pathway for optimizing wind farm layouts. [ABSTRACT FROM AUTHOR]

Injection molding (IM) is one complex manufacturing process characterized by nonlinear behavior. Unlike classic linear modeling techniques like simple regression, many machine learning (ML) models have the ability to adjust to the nonlinear behaviors and interactions between input and output parameters. Artificial neural networks (ANNs) specifically have demonstrated exceptional performance in problems involving nonlinear modeling. This work will employ complete factorial design of experiments (DoE) to acquire a dataset which is both resilient and suitable for training, validation, and testing purposes. The predictive model demonstrated outstanding performance throughout the training, validation, and test sets. The aggregate R2 values for the training, validation, and tests datasets were 97.58%, 93.76%, and 91.31%, respectively, demonstrating a strong ability to accurately foresee outcomes. Differential evolution (DE) successfully achieved a 2% decrease in weight and a notable 14% decrease in energy consumption. The results indicate that combining an ANN with DE is a viable approach for enhancing injection molding parameters, especially in scenarios with multiple objectives. [ABSTRACT FROM AUTHOR]

Cesium is an open-source 3D virtual earth engine based on WebGL and one of the most widely used 3D GIS engines. The 3DTiles rendering parameters in Cesium improve rendering efficiency but involve numerous data types and complex interdependencies, making simultaneous optimization challenging. In this paper, we proposed a multi-strategy probabilistic discrete differential evolution algorithm (MSPDDE) for finding the optimal values of the rendering parameters of Cesium 3DTiles model, which increases the search space and improves the convergence speed by introducing multiple mutation strategies. These strategies effectively reduce the probability of falling into local optimality due to too many parameters and deal with discrete variables in the rendering parameters by applying a probabilistic discretization strategy to the discrete variables so that MSPDDE is able to find all the rendering parameter optima of Cesium 3DTiles. Under three different sizes of 3DTiles model cases, the rendering model time using the optimal parameter configurations found by the algorithm is reduced by 28.84%, 27.89%, and 13.32%, respectively, compared with the default parameter configurations of Cesium, which shortens the time of rendering 3DTiles models by Cesium. [ABSTRACT FROM AUTHOR]

This study employs four metaheuristic optimization methods to optimize the dimensional synthesis of Ackermann steering mechanisms for three-axle, six-wheeled vehicles with front-axle steering mode and reverse-phase steering mode. The employed optimization methods include Particle Swarm Optimization (PSO), Hybrid Particle Swarm Optimization (HPSO), Differential Evolution with golden ratio (DE-gr), and Linearly Ensemble of Parameters and Mutation Strategies in Differential Evolution (L-EPSDE). With a front-wheel steering angle range of 70 degrees, two hundred optimization experiments were conducted for each method, and statistical analyses revealed that DE-gr and L-EPSDE methods outperformed PSO and HPSO methods in terms of standard deviation, mean value, and minimum error. These two methods exhibited superior convergence stability, faster convergence, and higher accuracy compared to PSO and HPSO. Reverse-phase (K = 1) steering mode outperformed front-axle steering mode, delivering reduced steering errors and turning radii. Considering the transmission ratio of front to rear axle (K) as a design variable in reverse-phase steering mode increased design flexibility and significantly lowered steering errors for the front and rear axle steering mechanisms. However, this comes with a slight increase in the turning radius of the vehicle's front part compared to when K = 1. The optimized mechanism, designed using the DE-gr method, was validated through kinematic simulations and steering analyses using MSC-ADAMS v2015 software, further confirming the effectiveness and reliability of the proposed design. [ABSTRACT FROM AUTHOR]

The strategic optimization of a container terminal's quayside assets, including the berth and quay cranes, is crucial for maximizing their deployment and utilization. The interrelated and complex challenges of Berth Allocation (BAP) and Quay Crane Scheduling (QCSP) are fundamental to enhancing the resilience of container ports, as berths and quay cranes constitute essential infrastructure. Efficient berth allocation and quay crane scheduling can mitigate operational disruptions, even in the face of maintenance or failures, thereby improving both operational reliability and resilience. However, previous studies have often overlooked the uncertainty associated with quay crane maintenance when planning these operations. This paper aims to minimize vessel turnaround time by accounting for the uncertain in quay crane maintenance activities. To address this novel problem, we propose a proactive-reactive method that incorporates a reliability-based model into the Swarm Optimization with Differential Evolution (SWO-DE) algorithm. Computational results confirm the practical relevance and effectiveness of our proposed solution methods for container terminals. [ABSTRACT FROM AUTHOR]

Electrochemical energy conversion technologies include proton exchange membrane fuel cells (PEMFCs) where proton interchange is an alternative to diesel distributed generation, and PEMFCs are considered as a promising backup power source and a tool to regulate power consumption. Some of the major benefits of these PEMFCs especially in power system applications include low emission of carbon, fast load following capability, no noise and high start-up reliability. It is challenging to find the best PEMFC parameters because the model is complex and the problem is nonlinear; not all optimization algorithms can solve this problem. This paper presents a new approach that applies QUasi-Affine TRansformation Evolution algorithm with a new adaptation of Evolution Matrix and Selection operation (QUATRE-EMS) to determine optimal values of uncertain parameters in PEMFC stack references. The objective function of the optimization problem is defined as the sum of squared errors of the actual and predicted voltage data. The effectiveness of the proposed QUATRE-EMS algorithm is also checked through statistical analysis and the QUATRE-EMS variant is compared with other variants of DE optimization algorithms which are recently proposed in the state-of-the-art literature such as LSHADE, MadDE, CS-DE, LPalmDE, EDEV, jSO, SHADE, ISDE, and JADE. Results show that the QUATRE-EMS algorithm reduces SSE significantly, with an average SSE of 0.078492, which is 15% less than the best performing existing algorithms. QUATRE-EMS also achieved the lowest average values of absolute error, relative error and mean bias error among different PEMFC stack references, with accuracy improved by up to 20%. It was also computationally more efficient, cutting runtime in half compared to other methods. The results of these findings confirm the effectiveness and practicability of the QUATRE-EMS algorithm for improving the accuracy of BCS500W, NedStackPS6, SR12, H12, HORIZON, and Standard 250W PEMFC stack references. [ABSTRACT FROM AUTHOR]

With rapid technological advancements, videos are captured, stored, and shared in multiple formats, increasing the requirement for summarization techniques to enable shorter viewing durations. Key Frame Extraction (KFE) algorithms are crucial in video summarization, compression, and offline analysis. This study aims to develop an efficient KFE approach for generic videos. Existing methods include the Adaptive Key Frame Extraction Algorithm, which reduces redundancy while ensuring maximum content coverage; the Optimal Key Frame Extraction Algorithm, which utilizes a Genetic Algorithm (GA) to select key frames optimally; and the Rapid Key Frame Extraction Algorithm, which employs clustering techniques to identify typical key frames. However, a clear prerequisite remains for a more versatile KFE technique that can address generic applications rather than specific use cases. Evolutionary algorithms offer a powerful solution for achieving optimal KFE. This proposed method leverages an interactive GA with a well-designed Fitness Function and elitism-based survivor selection to enhance performance. This proposed algorithm has been tested on diverse datasets, including VSUMM, SumMe, Mall, user-generated videos, surveillance footage from Amrita Vishwa Vidyapeetham University (Coimbatore, India), and web-sourced videos. The results demonstrate that the proposed KFE approach adheres to benchmark data and captures additional significant frames. Compared to Differential Evolution (DE) techniques and Deep Learning (DL) models from the literature, this recommended algorithm demonstrates superior efficiency, as verified through quantitative and qualitative evaluation metrics. Furthermore, the computational complexity of the GA is intricately compared to that of DE and DL-based approaches, highlighting the distinct efficiencies and performance features. [ABSTRACT FROM AUTHOR]

Optimization algorithms play a crucial role in solving complex problems across various fields, including global optimization and feature selection (FS). This paper presents the enhanced polar lights optimization with cryptobiosis and differential evolution (CPLODE), a novel improvement upon the original polar lights optimization (PLO) algorithm. CPLODE integrates a cryptobiosis mechanism and differential evolution (DE) operators to enhance PLO's search capabilities. The original PLO's particle collision strategy is replaced with DE's mutation and crossover operators, enabling a more effective global exploration and using a dynamic crossover rate to improve convergence. Furthermore, a cryptobiosis mechanism records and reuses historically successful solutions, thereby improving the greedy selection process. The experimental results on 29 CEC 2017 benchmark functions demonstrate CPLODE's superior performance compared to eight classical optimization algorithms, with higher average ranks and faster convergence. Moreover, CPLODE achieved competitive results in feature selection on ten real-world datasets, outperforming several well-known binary metaheuristic algorithms in classification accuracy and feature reduction. These results highlight CPLODE's effectiveness for both global optimization and feature selection. [ABSTRACT FROM AUTHOR]

This study investigates the integrated multi-objective scheduling problems of job shops and material handling robots (MHR) with minimising the maximum completion time (makespan), earliness or tardiness, and total energy consumption. The collaborative scheduling of MHR and machines can enhance efficiency and reduce costs. First, a mathematical model is constructed to articulate the concerned problems. Second, three meta-heuristics, i.e., genetic algorithm (GA), differential evolution, and harmony search, are employed, and their variants with seven local search operators are devised to enhance solution quality. Then, reinforcement learning algorithms, i.e., Q-learning and state–action–reward–state–action (SARSA), are utilised to select suitable local search operators during iterations. Three reward setting strategies are designed for reinforcement learning algorithms. Finally, the proposed algorithms are examined by solving 82 benchmark instances. Based on the solutions and their analysis, we conclude that the proposed GA integrating SARSA with the first reward setting strategy is the most competitive one among 27 compared algorithms. [ABSTRACT FROM AUTHOR]

The document "Special Issue 'Algorithms for Feature Selection (2nd Edition)" focuses on research advancements in algorithms, particularly in feature selection techniques. It covers topics such as evolutionary search techniques, ensemble methods, high-dimensional data, time-series data, textual data, and deep feature selection. The document highlights various research papers that address feature selection in fields like bioinformatics, scheduling, and time-series analysis, showcasing the importance of feature selection in different domains. The accepted papers present innovative methods and models for feature extraction, classification, and optimization, demonstrating the diverse applications of algorithms in solving complex data challenges. [Extracted from the article]

The Special Issue on Scheduling Theory and Algorithms for Sustainable Manufacturing was initiated at the MIM 2022 conference in Nantes, France. The issue focuses on how scheduling theory and algorithms can address complex optimization problems in manufacturing from sustainability perspectives. Various papers in the issue explore topics such as reinforcement learning in production scheduling, remanufacturing through machine learning, and sustainable inventory management in multi-storage supply chains. The articles aim to provide efficient solutions for planning and scheduling issues in digital, sustainable, and human-centric manufacturing and logistics. [Extracted from the article]

Feedforward neural networks (FNNs) are among the most widely used artificial neural network models, which have consistently demonstrated excellent performance in various fields, such as classification and regression tasks. However, the performance of FNNs does mainly depend on their structure (i.e., the number of hidden layers and that of neurons for each hidden layer) and initial weights, and they are usually required to manually determine via extensive tuning experiments in advance, thereby seriously limiting their practical applicability. To address this issue, this paper proposes an integrated meta-heuristic intelligent algorithm using genetic algorithm (GA) and differential evolution (DE). To satisfy the dynamic adjustment of network structure during the search process, a novel mixed encoding strategy is devised to represent both the structure and initial parameters of network model. Moreover, a combined offspring generation strategy is designed to explore more promising networks using GA and DE. Specifically, GA is utilized to search the optimal network structure, while DE is employed to refine its associated parameters. Following them, the potential information of different network models can be effectively interacted and shared, thus improving the search effectiveness and efficiency of algorithm. Finally, the performance of the proposed algorithm is validated by comparing with eight existing algorithms on twelve commonly used classification datasets and a practical problem, named air quality index prediction. Experimental results demonstrate that the proposed algorithm exhibits superior performance. [ABSTRACT FROM AUTHOR]