Particle Swarm Optimization Iteration Research Articles

Terahertz spectroscopy and effective medium theory for thickness measurement of adhesive bonds

This study deals with the characterization of multilayer adhesive structures via terahertz waves, particularly focusing on inversion analysis of the adhesive layer thickness. Terahertz-time-of-flight (THz-TOF), sensitive to the dielectric properties of materials, serves as an excellent device for non-destructive evaluation. Further, the transfer matrix method is introduced to simulate THz propagation through layered materials with various optical properties. An improved model iteration approach and particle swarm optimization algorithm are employed to effectively determine the adhesive layer thickness. This methodology utilizes effective medium theory (EMT), specifically for adhesive penetration into the cushion, thereby enhancing the thickness measurement accuracy. Among the various EMT models, the Lorentz-Lorenz (L-L) model is commonly regarded as the most effective one. An objective function based on combining the Pearson correlation coefficient and the root mean square error is proposed to refine the iterative inversion process. The results of this approach have been benchmarked with traditional THz-TOF calculations and computed tomography imaging, revealing that the proposed methodology is consistent with CT findings and outperforms conventional THz-TOF calculations.

Blood CO Status Classification Using UV-VIS Spectroscopy and PSO-optimized 1D-CNN Model

Rapid and effective blood carbon monoxide (CO) assessment is of great importance, especially in estimating CO-related morbidity and instituting effective preventive measures. The conventional detection methods using CO breath analysis lack sensitivity, while collecting biological fluid samples for CO level measurement is prone to external contamination and expensive for frequent use. This study proposes a one-dimensional convolutional neural network (1D-CNN) consisting of three stacked biconvolutional layers for binary classification of blood CO status using the diffuse reflectance spectroscopy technique. Iterative particle swarm optimization (PSO) has efficiently found the best network parameters to learn important features from the reflectance spectroscopy data. The findings showed good testing accuracy, specificity, and precision of 92.9%, 90%, and 89.7%, respectively, and a high sensitivity of 96.3% in determining abnormal blood CO among smokers using the proposed CNN network. Comparisons with eight existing machine learning and deep learning models revealed the proposed method’s effectiveness in classifying blood CO status while reducing computing time by 8–13 folds. The findings of this work provide new insights that are valuable for researchers in neural network design automation, healthcare management, and skin-related research, specifically for application in nondestructive evaluation and clinical decision-making.

Optimum Design of Reinforced Concrete Beams with Large Opening Using Neural Network Algorithm

It is quite rare to find researches in the literature connecting neural networks to design challenges in civil engineering, particularly with regard to beam with large openings. Therefore, to ascertain the best cost design for such a beam, an optimization technique is developed, written using MATLAB functions, and verified in this study. Because of the efficiency and dependability of the iterative Particle Swarm Optimization (PSO) process, the ACI 318-19 code method is adopted via this algorithm. Using four beam design (decision) variables, which are: beam width b, longitudinal tension reinforcement area As, longitudinal compression reinforcement area A’s, and vertical shear reinforcement area Av; the objective is to minimize the overall beam cost function. The constraints for each of the 60 randomly chosen particles are handled using the self-adaptive penalty function technique, which is applied in all the 100 iterations for each of the four proposed beam design case studies with identical openings. For the developed method, completing all iterations serves as a stopping criterion. Comparative studies are conducted to demonstrate how the compressive strength of concrete and the live load scheme affect the optimal overall beam cost and the associated design variables. The relations between cost of beams and the four design variables, which are presented through graphs and tables, indicate that the beam’s cost is significantly impacted by the loading condition. This is shown by increasing the cost of beam B1, with no sustained live loads, from 722 US dollars to 984 US dollars for beam B4, which has concentrated and distributed sustained live loads. Future research is advised to include modern strengthening techniques for beams with openings, and apply this research to a PSO algorithm with several objectives.

Remaining useful life prediction of lithium battery based on CEEMD-SE-IPSO-LSSVM hybrid model

Abstract In order to prevent accidents caused by battery aging, accurately predicting the remaining useful life (RUL) is a critical and highly challenging task in battery management systems. This article describes a lithium-ion battery RUL prediction method based on a hybrid model of CEEMD-SE-IPSO-LSSVM. This method integrates various technologies and algorithms, enhancing the accuracy and practicality of predictions. Initially, the complete ensemble empirical mode decomposition (CEEMD) is utilized to decompose the raw data into multiple intrinsic mode functions, aiding in denoising and feature extraction. Subsequently, the sample entropy (SE) is used to assess the complexity and irregularity of the data, merging intrinsic mode function components with similar SE values into a new component. Building upon this, the advanced iterative particle swarm optimization (IPSO) algorithm refines the parameters of the least squares support vector machine (LSSVM) model, improving the predictive performance of the model. Finally, through iterative training and refinement of the LSSVM model, accurate prediction of the remaining life of lithium-ion batteries is achieved. This hybrid model approach integrates data processing, feature extraction, and model refinement, resulting in a significant improvement over the baseline model with a 69.5% increase in mean absolute percentage error and a 49.4% decrease in root mean squared error, providing a robust solution for predicting the remaining life of lithium-ion batteries.

A Novel Skeleton-Based Human Activity Discovery Using Particle Swarm Optimization With Gaussian Mutation

Human activity discovery aims to cluster human activities without any prior knowledge of what defines each activity. However, most existing methods for human activity recognition are supervised, relying on labeled inputs for training. In reality, it is challenging to label human activity data due to its large volume and the diversity of human activities. To address this issue, this article proposes an unsupervised framework for human activity discovery in 3-D skeleton sequences. The framework includes a data preprocessing step that selects important frames based on kinetic energy and extracts relevant features, such as joint displacement, statistical displacement, angles, and orientation. To reduce the dimensionality of the extracted features, the framework uses principle component analysis. Unlike many other methods for human activity discovery, the proposed framework is fully unsupervised and does not rely on presegmented videos. To segment the time series of activities, the framework uses a sliding time window with some overlapping. The hybrid particle swarm optimization (PSO) with Gaussian mutation and K-means algorithm is then proposed to discover the activities. PSO is chosen for its powerful global search capability and simple implementation. To further improve the convergence rate of PSO, K-means is applied to the outcome centroids from each iteration of PSO. The experimental results on five datasets demonstrate that the proposed framework has superior performance in discovering activities compared to other state-of-the-art methods. The framework achieves an average increase in accuracy of at least 4%.

Increasing Unbalanced Distribution Network’s Hosting Capacity for Distributed Energy Resources by Voltage Regulators

Renewable energy sources (RESs) are becoming promising nowadays. Deploying distributed energy resources (DERs) is thus increasing accordingly, but the intermittent nature associated with RES remains a significant challenge. Also, the distribution grid faces technological difficulties with the large-scale integration of PV generations and EV charging loads. This paper suggests a strategy for a novel optimal voltage regulator (VR) placement model associated with the optimal placement of electric vehicle chargers (EVCs) and photovoltaic (PV) cells. Also, VRs’ reference voltages are determined optimally to enhance the hosting capacity (HC) of three-phase unbalanced distribution feeders for large-scale residential/public EVCs and PVs. Two scenarios are defined to study the impact of VRs on the network’s HC for DERs, namely with and without VRs in the real three-phase unbalanced feeder. The developed model determines the optimal placement of VRs and their reference voltages linked with EVCs and PVs penetration. Time-Varying Acceleration Coefficients Iteration Particle Swarm Optimization (TVAC-IPSO) algorithm has been implemented to improve the real three-phase unbalanced distribution feeder’s HC for DERs. The efficiency of the proposed model and the solution method is validated on the modified IEEE 37-node medium voltage (MV) radial distribution network connected to 35 low voltage (LV) distribution feeders. The obtained results represent that installing a proper number of VRs will increase the network’s HC for DERs.

Multilevel Thresholding Image Segmentation Based-Logarithm Decreasing Inertia Weight Particle Swarm Optimization

Abstract The image segmentatation technique that is often used is thresholding. Image segmentation is a process of dividing the image into different regions according to their similar characteristics. This research proposes a multilevel thresholding algorithm using modified particle swarm optimization to solve a segmentation problem. The threshold optimal values are determined by maximizing Otsu’s objective function using optimization technique namely particle swarm optimization based on the logarithmic decreasing inertia weight (LogDIWPSO). The proposed method reduces the computational time to find the optimum thresholds of multilevel thresholding which evaluated on several grayscale images. A detailed comparison analysis with other multilevel thresholding based techniques namely particle swarm optimization (PSO), iterative particle swarm optimization (IPSO), and genetic algorithms (GA), From the experiments, Modified particle swarm optimization (MoPSO) produces better performance compared to the other methods in terms of fitness value, robustness and convergence. Therefore, it can be concluded that MoPSO is a good approach in finding the optimal threshold value. Keywords: grayscale image, inertia weight, image segmentation, particle swarm optimization.

Accurate Design of Microwave Filter Based on Surrogate Model-Assisted Evolutionary Algorithm

Filter optimization problems involve time-consuming simulations and many variables in the design. These problems require a large amount of computation. This paper proposes an adaptive online updating 1D convolutional autoencoders (AOU-1D-CAE) surrogate model for solving this computationally expensive problem. In the optimization process, an adaptive update surrogate mapping between input variables and output objectives is constructed within the surrogate model AOU-1D-CAE framework. AOU-1D-CAE can replace electromagnetic (EM) simulation software for data collection, and select and automatically use the accumulated data as training samples to train the AOU-1D-CAE surrogate model. With more and more training samples, the learning ability of the surrogate model is also becoming stronger and stronger. The experimental results show that the data collection efficiency of AOU-1D-CAE is greatly improved, and the automatic update of the sample set improves the prediction performance of the surrogate model. In this paper, the optimization framework is AOU-1D-CAE-assisted particle swarm optimization (PSO), and the surrogate model assists PSO to find the global optimal solution. In the PSO stage, PSO automatically updates and saves the optimal solution, and takes the optimal solution of each stage as the initial solution of the next optimization stage to avoid falling into the local optimal solution. The optimization time is greatly saved and the optimization efficiency is improved. The continuous iteration of PSO also improves the prediction accuracy of the surrogate model. The efficiency of the proposed surrogate model is demonstrated by using two cavity filters as examples.

Particle Swarm Optimization with Long and Short Term Memory in Feature Selection

<p>Taking each iteration of Particle swarm optimization (PSO) algorithm as a time node, the change of population in PSO algorithm can be regarded as a time series model. Particle population learns and evolves in multiple time nodes, which can be regarded as a dependent behavior on leader particles. In the traditional particle swarm optimization algorithm, this dependence behavior is independent of time, and its consideration standard is only the fitness value of particles. We deeply study the leadership mechanism of PSO algorithm in order to find a more robust leadership mechanism and improve the ability of PSO algorithm to explore the solution space, by extending the dependence behavior in the time dimension, we propose an improved PSO algorithm with long-term and short-term memory ability. In order to verify its performance, in the experimental part, we select 32 public data sets in UCI data to find the optimal feature subset. In a large number of feature selection experiments. The experimental results proofed that the performance of proposed algorithms is better than some state of the art algorithms. </p> <p> </p>

Research on strip crown by uncertain sampling strategy modified particle swarm optimization with RBF neural network

The strip crown directly affects the quality of strip. The prediction of strip crown by general machine learning models usually focuses on the production of a single category of strip steel, and the models lack good prediction ability for complex industrial data environments, which will lead to an increase in production costs and a decrease in product quality. Therefore, this paper proposes a new Radial Basis Function (RBF) neural network model, which is optimized by a new uncertain sampling strategy modified PSO algorithm (US-MPSO algorithm). The main feature of the algorithm is to add a new clustering dimension to the particle swarm algorithm (PSO). In this paper, clustering optimization is combined with uncertain sampling strategy to help particle swarm optimization iteration. On the one hand, the algorithm collects the population information of the strip crown as much as possible to help the neural network to converge. On the other hand, it adopts the uncertain sampling strategy to realize the optimal iteration of the samples. Finally, it realizes the improvement of the running speed and accuracy of the algorithm. This paper uses a specific model initialization strategy and combination strategy to combine the optimization algorithm with the RBF neural network. In this paper, four different types of data sets are tested to prove the adaptability of the model for complex industrial environments. The results show that the RBF neural network optimized by the US-MPSO algorithm has better prediction accuracy than the previous RBF neural network. The optimization effect of the algorithm is very significant.

Camera calibration based on improved differential evolution particle swarm

Aiming at the problem of camera calibration with multiple parameters, this paper proposes the fusion optimization algorithm of improved differential evolution and particle swarm in the calibration of the camera. Adaptive judgment factors is Introduced to control the improvement of differential evolution during each iteration (IDE) algorithm and particle swarm optimization (PSO) algorithm solve the multiple parameters of traditional camera calibration algorithm. The proposed algorithm can ensure the diversity and effectiveness of individual evolution of the population. Experimental results show that the algorithm has excellent global search capabilities and local optimizations ability. It can accurately calibrate the camera. The convergence precision, speed and robustness performance significantly is superior to PSO (Particle swarm optimization algorithm), DE (differential evolution algorithm), GA (Genetic algorithm) and Zhang’s method. It improves the precision and speed of the proposed calibration method. The root mean square error of the calibration algorithm proposed in this paper is only 0.182, the calibration error result is smaller than other several algorithms. The reprojection error of our method is 0.05938 (Ex/pixel) and 0.02988 (Ey/pixel). It is smaller than GA, PSO, DE, and Zhang’s method. So, the algorithm performance is excellent.

Wind-field identification for parafoils based on deep Q-learning iterative inversion

Powered parafoils are flexible aircraft with high load capacities and low costs and are often used in long-term surveillance and airdrop missions. As a flexible aircraft, powered parafoil flights are easily affected by winds. However, wind-field identification is difficult for low-to-medium speed unmanned aerial vehicles (UAV), which makes it challenging for UAVs such as powered parafoils to perform targeted wind compensation. To observe and compensate for this, we must accurately identify the wind field for the powered parafoil. In this study, we designed a deep Q-learning iterative inversion (DQII) algorithm for wind-field identification to concretely analyze wind-field identification through model inversion. Using a deep Q-learning (DQ) inverting dynamic model, a powered parafoil can identify the wind field within 2 s. In addition, iterative chaotic particle swarm optimization was performed based on DQ inversion, which achieved a high-precision wind-field identification effect. Simulation experiments showed that, compared with the traditional dead reckoning (DR) algorithm, DQII can reduce the error by 80%–90%. In a powered parafoil homing experiment, DQII achieved a wind-field identification effect that was significantly better than that of the DR algorithm.

An innovative forecasting model to predict wind energy.

In recent years, the global wind power construction is accelerating. Although wind power is a clean energy without pollution, its volatility and irregularity have a great impact on wind power integration. Therefore, scholars pay more and more attention to the ultra-short-term prediction of wind speed. At present, the popular wind speed prediction model usually combines wind speed decomposition algorithm, machine learning algorithm, and intelligent optimization algorithm. The general wind speed decomposition algorithm cannot use the information contained in the factors affecting wind speed. Besides, the current popular optimization algorithms, such as gray wolf optimization algorithm, have strong convergence and better optimization effect, but their structure is complex and their operation complexity is large. And the PSO algorithm has simple structure and fast operation speed. To solve the above problems, a novel combination prediction model is proposed in this paper. This model uses VMD to decompose the wind speed into high-frequency signal and low-frequency signal and then uses principal component analysis and spectral clustering to extract and classify the influencing factors. In addition, aiming at the problem of slow convergence speed in the later stage of PSO iteration, an adaptive improved PSO is proposed. Finally, this paper also designs a rolling train method to adjust the size of training samples. Through four experiments of wind speed series in two periods, it is proved that the combined prediction model proposed in this paper has the following advantages: the model fully extracts the information of wind speed and influencing factors; the improved PSO algorithm has better optimization effect; rolling training method effectively improves the prediction ability of the model; the combined forecasting model has good adaptability and competitiveness.

A Particle Swarm Optimization Algorithm with Gradient Perturbation and Binary Tree Depth First Search Strategy

In this study, a particle swarm optimization (PSO) algorithm with a negative gradient perturbation and binary tree depth‐first strategy (GB‐PSO) is proposed. The negative gradient term accelerates particle optimization in the direction of decreasing the objective function value. To calculate the step size of this gradient term more easily, a method based on the ratio was proposed. In addition, a new PSO strategy is also proposed. Each iteration of PSO yields not only the current optimal solution of the group, but also the solution based on the 2‐norm maximum. Under the current iteration solution of PSO, these two solutions are the children nodes. In the sense of the binary tree concept, the three solutions constitute the father‐son relationship, and the solution generated throughout the entire search process constitutes the binary tree. PSO uses a traceable depth‐first strategy to determine the optimal solution. Compared with the linear search strategy adopted by several algorithms, it can fully utilize the useful information obtained during the iterative process, construct a variety of particle swarm search paths, and prevent premature and enhance global optimization. The experimental results show that the algorithm outperforms some state‐of‐the‐art PSO algorithms in terms of search performance.

A Recurrent Neural Network–Based Proxy Model for Well-Control Optimization with Nonlinear Output Constraints

Summary In well-control optimization problems, the goal is to determine the time-varying well settings that maximize an objective function, which is often the net present value (NPV). Various proxy models have been developed to predict NPV for a set of inputs such as time-varying well bottomhole pressures (BHPs). However, when nonlinear output constraints (e.g., maximum well/field water production rate or minimum well/field oil rate) are specified, the problem is more challenging because well rates as a function of time are required. In this work, we develop a recurrent neural network (RNN)–based proxy model to treat constrained production optimization problems. The network developed here accepts sequences of BHPs as inputs and predicts sequences of oil and water rates for each well. A long-short-term memory (LSTM) cell, which is capable of learning long-term dependencies, is used. The RNN is trained using well-rate results from 256 full-order simulation runs that involve different injection and production-well BHP schedules. After detailed validation against full-order simulation results, the RNN-based proxy is used for 2D and 3D production optimization problems. Optimizations are performed using a particle swarm optimization (PSO) algorithm with a filter-based nonlinear-constraint treatment. The trained proxy is extremely fast, although optimizations that apply the RNN-based proxy at all iterations are found to be suboptimal relative to full simulation-based (standard) optimization. Through use of a few additional simulation-based PSO iterations after proxy-based optimization, we achieve NPVs comparable with those from simulation-based optimization but with speedups of 10 or more (relative to performing five simulation-based optimization runs). It is important to note that because the RNN-based proxy provides full well-rate time sequences, optimization constraint types or limits, as well as economic parameters, can be varied without retraining. NOTE: This paper is published as part of the 2021 Reservoir Simulation Conference Special Issue.

Dense conjugate initialization for deterministic PSO in applications: ORTHOinit+

This paper describes a class of novel initializations in Deterministic Particle Swarm Optimization (DPSO) for approximately solving costly unconstrained global optimization problems. The initializations are based on choosing specific dense initial positions and velocities for particles. These choices tend to induce in some sense orthogonality of particles’ trajectories, in the early iterations, in order to better explore the search space. Our proposal is inspired by both a theoretical analysis on a reformulation of PSO iteration, and by possible limits of the proposals reported in Campana et al. (2010); Campana et al. (2013). We explicitly show that, in comparison with other initializations from the literature, our initializations tend to scatter PSO particles, at least in the first iterations. The latter goal is obtained by imposing that the initial choice of particles’ position/velocity satisfies specific conjugacy conditions, with respect to a matrix depending on the parameters of PSO. In particular, by an appropriate condition on particles’ velocities, our initializations also resemble and partially extend a general paradigm in the literature of exact methods for derivative-free optimization. Moreover, we propose dense initializations for DPSO, so that the final approximate global solution obtained is possibly not too sparse, which might cause troubles in some applications. Numerical results, on both Portfolio Selection and Computational Fluid Dynamics problems, validate our theory and prove the effectiveness of our proposal, which applies also in case different neighborhood topologies are adopted in DPSO.

Optimization of decoupling point position using metaheuristic evolutionary algorithms for smart mass customization manufacturing.

In this paper, we present two metaheuristic evolutionary algorithms-based approaches to position the customer order decoupling point (CODP) in smart mass customization (SMC). SMC tries to autonomously mass customize and produce products per customer needs in Industry 4.0. SMC shown here is from the perspective of arriving at a CODP during manufacturing process flow designs meant for fast moving and complex product variants. Learning generally needs several repetitive cycles to break the complexity barrier. We make use of fruit fly and particle swarm optimization (PSO) evolutionary algorithms with the help of MATLAB programming to constantly search better fitting consecutive process modules in manufacturing chain. CODP is optimized by increasing modularity and reducing complexity through evolutionary concept. Learning-based PSO iterations are performed. The methods shown here are recommended for process flow design in a learning-oriented supply chain organization which can involve in-house and outsourced manufacturing steps. Finally, a complexity reduction model is presented which can aid in deploying this concept in design of supply chain and manufacturing flows.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00521-020-05657-1.

Practical demand response potential evaluation of air-conditioning loads for aggregated customers

The evaluation of demand response (DR) potential is of vital significance for evaluating the possible power reduction, forming constraints for optimal economic dispatching, and targeting appropriate customers effectively. To improve the applicability and reliability of DR of air conditioning loads (ACLs), this paper presents a novel approach of the DR potential for aggregated customers, considering both the performance of physical response and incentive effect. For physics-based part, in order to better reveal the multi-uncertainties and various conditions, a load decomposition methodology is proposed, which non-intrusively disaggregates the whole energy usage into baseload and ACLs. Subsequently, a segment analysis methodology, including a modified regression method and a PSO iteration method, are developed for the static and dynamic parameter estimation, respectively. For incentive-based part, the electricity elasticity of both the electricity price and incentive are considered to calculate the load difference after changes of electricity price and incentive, which better consider the actual response willingness and capacity of customers. Furthermore, case studies based on low voltage distribution area in Jiangsu province, China validate a better performance of accuracy and robustness using the proposed methods, which are further applied to DR implementation and the results strongly prove that much difference (a maximum of more than 400 kW) appears resulting from different operational conditions and DR strategies. By comparison, the practical DR potential is much affected by the incentive, rather than the physics, which mainly stimulates the customers’ intention. And the equilibrium can be achieved by choosing the optimal incentive price to stimulate greater potential and save the cost simultaneously.

Chaotic image encryption algorithm based on hybrid multi-objective particle swarm optimization and DNA sequence

This paper studies an image encryption algorithm based on multi-objective particle swarm optimization (MOPSO), DNA encoding sequence and one-dimensional Logistic map. First, the key of this paper consists of the sub-key sequence selected by particle swarm optimization (PSO), a hash value of the plaintext image, and a shuffle mark bit. Create random DNA mask images using Logistic map and DNA encoding. Then use it and the block-shuffled plaintext DNA encoding sequence to operate to form an encryption system. In PSO, the position value of a particle represents a position of the plaintext image, the iterative PSO algorithm is based on the information entropy and correlation coefficient. Finally, obtains the best ciphertext, and returns the value of the best particle at this time. Simulation experiment and security analysis show that the correlation coefficient and entropy of ciphertext are excellent, and it can resist all kinds of typical attacks and has better encryption effect.

Particle Swarm Optimization (PSO) for improving the accuracy of ChemCam LIBS sub-model quantitative method

Laser-induced breakdown spectroscopy (LIBS) is a powerful tool for qualitative analysis of chemical composition on planetary surface. Specifically, the quantitative compositional analysis method is a significant challenge for LIBS instrument onboard the Mars Science Laboratory (MSL) rover Curiosity ChemCam. Partial Least Squares (PLS) sub-model strategy is one of the outstanding multivariate analysis methods for calibration modeling, which is firstly developed by ChemCam science team. However, a troubling key issue is there are many parameters that need to be optimized, which increases the uncertainty of predicting outcomes and is time-consuming. In this study, an automatic parameters selection method based on Particle Swarm Optimization (PSO) tool is introduced. In the process of PSO iteration, RMSE minimization is taken as fitness, and finally the optimal sub-model parameters set is searched. In this way, the authors also get the best PLS latent variables of each sub-model by traversal method. Finally, the PSO PLS sub-model (PSO-PLS-SM) gets significant improvement in accuracy for the expanded Chemcam standards (408). And the RMSE of Si, Al, Ca, Na elements has been reduced by more than 20% relative to the conventional predictions.