{"title":"使用混合兰克-高斯 PSO 算法和 ANN 预测并优化 WAAM 组件的表面波浪度","authors":"","doi":"10.1016/j.istruc.2024.107247","DOIUrl":null,"url":null,"abstract":"<div><p>Surface waviness is crucial for the quality of components produced via the wire and arc additive manufacturing (WAAM) process. This study presents a novel method for predicting surface waviness, employing an advanced model to optimize process parameter configurations using a combination of Rank-Gaussian particle swarm optimization (RGPSO) and an Artificial Neural Network (ANN). The novelty of this process is that the RGPSO not only optimizes the hyperparameters of the ANN model to enhance prediction performance, but also addresses surface waviness optimization. The RGPSO algorithm's optimization performance is evaluated using 23 benchmark functions, demonstrating competitiveness against nine comparative meta-heuristic algorithms. Experimental data on surface waviness from the literature are utilized to train, test and validate three different prediction models, including a standalone ANN model, a PSO optimized ANN (PSO-ANN) model, and an RGPSO optimized ANN (RGPSO-ANN) model. The results indicate that the developed RGPSO-ANN model achieves the highest accuracy in terms of the metrics <span><math><mi>RMSE</mi></math></span> (0.019), <span><math><mi>R</mi></math></span> (0.996), <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mrow><mfenced><mrow><mn>0.990</mn></mrow></mfenced></mrow><mo>,</mo><mspace></mspace><mi>MAE</mi></mrow></math></span> (0.013), <span><math><mrow><mi>RMSLE</mi><mrow><mfenced><mrow><mn>0.013</mn></mrow></mfenced></mrow><mo>,</mo><mspace></mspace></mrow></math></span>and <span><math><mi>MAPE</mi></math></span> (3.46 %). It performs better than the PSO-ANN model (<span><math><mi>RMSE</mi></math></span> 0.026, <span><math><mi>R</mi></math></span> 0.975, <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mn>0.982</mn><mo>,</mo></mrow></math></span> <span><math><mi>MAE</mi></math></span> 0.019, <span><math><mrow><mi>RMSLE</mi><mspace></mspace></mrow></math></span>0.019<span><math><mrow><mo>,</mo><mi>a</mi></mrow></math></span>nd <span><math><mi>MAPE</mi></math></span> 5.73 %), and better than the ANN (<span><math><mi>RMSE</mi></math></span> 0.046, <span><math><mrow><mi>R</mi><mspace></mspace></mrow></math></span>0.991, <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mspace></mspace><mn>0.944</mn><mo>,</mo></mrow></math></span> <span><math><mi>MAE</mi></math></span> 0.034, <span><math><mrow><mi>RMSLE</mi><mn>0.032</mn></mrow></math></span> and <span><math><mi>MAPE</mi></math></span> 9.31 %). The RGPSO, PSO and other optimization algorithms are then applied to minimize the surface waviness of a WAAM component. RGPSO achieved the optimal value (0.1631 mm), which corresponds to a 12.6 % reduction compared to the best value obtained using PSO.</p></div>","PeriodicalId":48642,"journal":{"name":"Structures","volume":null,"pages":null},"PeriodicalIF":3.9000,"publicationDate":"2024-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Prediction and optimization of surface waviness of WAAM components using a hybrid Rank-Gaussian PSO algorithm and ANN\",\"authors\":\"\",\"doi\":\"10.1016/j.istruc.2024.107247\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>Surface waviness is crucial for the quality of components produced via the wire and arc additive manufacturing (WAAM) process. This study presents a novel method for predicting surface waviness, employing an advanced model to optimize process parameter configurations using a combination of Rank-Gaussian particle swarm optimization (RGPSO) and an Artificial Neural Network (ANN). The novelty of this process is that the RGPSO not only optimizes the hyperparameters of the ANN model to enhance prediction performance, but also addresses surface waviness optimization. The RGPSO algorithm's optimization performance is evaluated using 23 benchmark functions, demonstrating competitiveness against nine comparative meta-heuristic algorithms. Experimental data on surface waviness from the literature are utilized to train, test and validate three different prediction models, including a standalone ANN model, a PSO optimized ANN (PSO-ANN) model, and an RGPSO optimized ANN (RGPSO-ANN) model. The results indicate that the developed RGPSO-ANN model achieves the highest accuracy in terms of the metrics <span><math><mi>RMSE</mi></math></span> (0.019), <span><math><mi>R</mi></math></span> (0.996), <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mrow><mfenced><mrow><mn>0.990</mn></mrow></mfenced></mrow><mo>,</mo><mspace></mspace><mi>MAE</mi></mrow></math></span> (0.013), <span><math><mrow><mi>RMSLE</mi><mrow><mfenced><mrow><mn>0.013</mn></mrow></mfenced></mrow><mo>,</mo><mspace></mspace></mrow></math></span>and <span><math><mi>MAPE</mi></math></span> (3.46 %). It performs better than the PSO-ANN model (<span><math><mi>RMSE</mi></math></span> 0.026, <span><math><mi>R</mi></math></span> 0.975, <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mn>0.982</mn><mo>,</mo></mrow></math></span> <span><math><mi>MAE</mi></math></span> 0.019, <span><math><mrow><mi>RMSLE</mi><mspace></mspace></mrow></math></span>0.019<span><math><mrow><mo>,</mo><mi>a</mi></mrow></math></span>nd <span><math><mi>MAPE</mi></math></span> 5.73 %), and better than the ANN (<span><math><mi>RMSE</mi></math></span> 0.046, <span><math><mrow><mi>R</mi><mspace></mspace></mrow></math></span>0.991, <span><math><mrow><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup><mspace></mspace><mn>0.944</mn><mo>,</mo></mrow></math></span> <span><math><mi>MAE</mi></math></span> 0.034, <span><math><mrow><mi>RMSLE</mi><mn>0.032</mn></mrow></math></span> and <span><math><mi>MAPE</mi></math></span> 9.31 %). The RGPSO, PSO and other optimization algorithms are then applied to minimize the surface waviness of a WAAM component. RGPSO achieved the optimal value (0.1631 mm), which corresponds to a 12.6 % reduction compared to the best value obtained using PSO.</p></div>\",\"PeriodicalId\":48642,\"journal\":{\"name\":\"Structures\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":3.9000,\"publicationDate\":\"2024-09-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Structures\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2352012424013997\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Structures","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2352012424013997","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Prediction and optimization of surface waviness of WAAM components using a hybrid Rank-Gaussian PSO algorithm and ANN
Surface waviness is crucial for the quality of components produced via the wire and arc additive manufacturing (WAAM) process. This study presents a novel method for predicting surface waviness, employing an advanced model to optimize process parameter configurations using a combination of Rank-Gaussian particle swarm optimization (RGPSO) and an Artificial Neural Network (ANN). The novelty of this process is that the RGPSO not only optimizes the hyperparameters of the ANN model to enhance prediction performance, but also addresses surface waviness optimization. The RGPSO algorithm's optimization performance is evaluated using 23 benchmark functions, demonstrating competitiveness against nine comparative meta-heuristic algorithms. Experimental data on surface waviness from the literature are utilized to train, test and validate three different prediction models, including a standalone ANN model, a PSO optimized ANN (PSO-ANN) model, and an RGPSO optimized ANN (RGPSO-ANN) model. The results indicate that the developed RGPSO-ANN model achieves the highest accuracy in terms of the metrics (0.019), (0.996), (0.013), and (3.46 %). It performs better than the PSO-ANN model ( 0.026, 0.975, 0.019, 0.019nd 5.73 %), and better than the ANN ( 0.046, 0.991, 0.034, and 9.31 %). The RGPSO, PSO and other optimization algorithms are then applied to minimize the surface waviness of a WAAM component. RGPSO achieved the optimal value (0.1631 mm), which corresponds to a 12.6 % reduction compared to the best value obtained using PSO.
期刊介绍:
Structures aims to publish internationally-leading research across the full breadth of structural engineering. Papers for Structures are particularly welcome in which high-quality research will benefit from wide readership of academics and practitioners such that not only high citation rates but also tangible industrial-related pathways to impact are achieved.