Multi-Response Parameters Optimization Based on PCA and Neural Network

doi:10.16182/j.issn1004731x.joss.201801022

Abstract

Abstract: A multi-response parameters optimization method based on principal component analysis (PCA) and neural network is proposed. It is used to optimize temperature and time parameters in complex thermal polymerization process. By using the method of weighted PCA, two response indexes, capacity value and loss tangent value, are converted into a single quality performance index. The main effect value is used to identify the search range. The radical basis function (RBF) neural network model is established to search and identify the optimal process parameters. Results show that response indexes are improved and the optimization effect is obvious. Therefore, this study method is a convenient and effective method to solve the complicated nonlinear response process parameters optimization, and has practical application value.

Key words: PCA, neural network, metallized film capacitor, parameter optimization

CLC Number:

TP114.2

Yu Jianli, Huang Hongqi, Miao Manxiang. Multi-Response Parameters Optimization Based on PCA and Neural Network[J]. Journal of System Simulation, 2018, 30(1): 176-184.

References

[1] 李智威, 李化, 杨佩原, 等. 不同场强下金属化聚丙烯膜电容器泄漏特性[J]. 电工技术学报, 2013, 28(9): 274-280. Li Z W, Li H, Yang P Y, et al. Leakage Characteristic of Metallized Polypropylene Film Capacitors under Different Electric Fields[J]. Transactions of China Eelectrotechnical Society, 2013, 28(9): 274-280.
[2] 彭宝华, 周经伦, 冯静, 等. 金属化膜脉冲电容器剩余寿命预测方法研究[J]. 电子学报, 2011, 39(11): 2674-2679. Peng B H, Zhou J L, Feng J, et al. Residual Lifetime Prediction of Metallized Film Pulse Capacitors[J]. Acta Electronica Sinica, 2011, 39(11): 2674-2679.
[3] 樊红杰. 金属化薄膜电容器电容量衰减的解决方案[J]. 电子元件与材料, 2011, 30(9): 40-42. Fan H J. Design and optimization of capacitance decay program for metallized film capacitors[J]. Electronic Components and Materials, 2011, 30(9): 40-42.
[4] Candioti L V, De Zan M M, Camara S, et al. Goicoechea. Experimental Design and Multiple Response Optimization. Using the Desirability Function in Analytical Methods Development[J]. Talanta (S0039-9140), 2014, 124: 123-138.
[5] Salmasnia A, Bashiri M.A new desirability function-based method for correlated multiple response optimization[J]. The International Journal of Advanced Manufacturing Technology(S0268-3768), 2015, 76: 1047-1062.
[6] 张流洋, 马义中, 汪建均, 等. 考虑协变量的相关多响应稳健优化设计[J]. 系统工程, 2015, 33(2): 126-133. Zhang L Y, Ma Y Z, Wang J J, et al. Correlated multi-response robust design and optimization with considering covariates[J]. Systems Engineering, 2015, 32(2): 126-133.
[7] 汪建均, 马义中, 翟云焕. 相关多质量特性的优化设计[J]. 管理工程学报, 2011, 25(2): 66-73. Wang J J, Ma Y Z, Zhai Y H. Optimization Design of Correlated Multiple Quality Characteristics[J]. Journal of Industrial Engineering and Engineering Management, 2011, 25(2): 66-73.
[8] 田亮, 刘芳, 刘鑫屏, 等. 高速率变负荷模式下火电机组协调控制系统参数优化[J]. 系统仿真学报, 2015, 27(7): 1532-1540. Tian L, Liu F, Liu X P, et al. Parameter Optimization on Coordinated Control System of Thermal Power Units in High Rate Variable Load Operation Mode[J]. Journal of System Simulation, 2015, 27(7): 1532-1540.
[9] 刘长良, 马增辉. Smith预估模型参数仿真分析及多目标优化[J]. 系统仿真学报, 2014, 26(8): 1706-1712. Liu C L, Ma Z H. Simulation Analysis and Multi-objective Optimization of Smith-Predictor Parameters Tuning[J]. Journal of System Simulation, 2014, 26(8): 1706-1712.
[10] Shit J S, Tzeng Y F, Yang J B.Principal Component Analysis for Multiple Quality Characteristics Optimization of Mental Inert Gas Welding Aluminum Foam Plate[J]. Materials and Design(S0261-3069), 2011, 32(3): 1253-1261.
[11] Gauri S K, Pal S.The Principal Component Analysis (PCA)-Based Approaches for Multi-response Optimization: Some Areas of Concerns[J]. The International Journal of Advanced Manufacturing Technology(S0268-3768), 2014, 70: 1875-1887.
[12] 张迎冬, 何桢. 改进的主成分分析法在多响应优化中的应用[J]. 组合机床与自动化加工技术, 2012(11): 97-100. Zhang Y D, He Z. Application of improved principal component analysis method to the multi-response optimization[J]. Modular Machine Tool ＆ Automatic Manufacturing Technique, 2012(11): 97-100.
[13] Zhang Y Y, Pan B C.Modeling Batch and Column Phosphate Removal by Hydrated Ferric Oxide-based Nanocomposite Using Response Surface Methodology and Artificial Neural Network[J]. Chemical Engineering Journal(S1385-8947), 2014, 249: 111-120.
[14] Peng A H, Xiao X M, Yue R.Process Parameter Optimization for Fused Deposition Modeling Using Response Surface Methodology Combine with Fuzzy Inference System[J]. The International Journal of Advanced Manufacturing Technology(S0268-3768), 2014, 73: 87-100.
[15] 欧阳林寒, 马义中, 汪建均, 等. 基于模型不确定性的响应曲面建模[J]. 系统工程与电子技术, 2015, 37(8): 1818-1824. Ou Yang L H, Ma Y Z, Wang J J, et al. Response Surface Modeling Base on Model Uncertainty[J]. Systems Engineering and Electronics, 2015, 37(8): 1818-1824.
[16] 樊兆峰, 马小平, 邵晓根. 非线性系统RBF神经网络多步预测控制[J]. 控制与决策, 2014, 29(7): 1274-1278. Fan Z F, Ma X P, Shao X G. RBF neural network multi-step predictive control for nonlinear systems[J]. Control and Decision, 2014, 29(7): 1274-1278.
[17] 张永志, 董俊慧. 基于模糊C均值聚类的模糊RBF神经网络预测焊接接头力学性能建模[J]. 机械工程学报, 2014, 50(12): 58-64. Zhang Y Z, Dong J H. Modeling Fuzzy RBF Neural Network to Predict of Mechanical Properties of Welding Joints Based on Fuzzy C-means Cluster[J]. Journal of Mechanical Engineering, 2014, 50(12): 58-64.