基于NSGA权值修正的MPC方法及其在光氢系统应用

MPC METHOD BASED ON NSGA WEIGHT CORRECTION AND ITS APPLICATION IN PHOTOHYDROGEN SYSTEM

针对模型预测控制(MPC)优化控制方法中权值修正问题,以光氢系统为研究背景,搭建一种非支配排序遗传算法(NSGA)修正MPC中权值的复合模型。以NSGA为外层函数,以储能电池出力能力评价系数、储能电池充/放电总能量和碱性电解槽(AEL)功率波动率为目标函数,将权值作为NSGA-MPC复合模型的输入变量进行选择、交叉、变异;以MPC优化控制方法为内层函数,针对不同权值,将计算得出的不同输出变量作为NSGA中目标函数的输入变量,经过遍历、寻优,最终得到最优权值。将最优权值结合现有研究成果中MPC优化控制所用权值围绕三方面进行对比分析,包括:目标函数值、MPC优化控制追踪效果、电解槽和储能电池功率波动率。结果表明,所得最优权值一定程度上降低了电解槽和储能电池的功率波动,优化了MPC控制追踪效果。

Addressing the weight adjustment issue in MPC(model predictive control)optimization,a composite model is constructed for NSGA(non-dominated sorting genetic algorithms)based weight modification,taking the photo-hydrogen system as the research background.NSGA is used as the outer function,with the evaluation coefficients of the energy storage battery’s output capability,the total energy of the energy storage battery’s charge/discharge,and the AEL power fluctuation rate as objective functions.The weights are treated as input variables for the NSGA-MPC composite model,where they undergo selection,crossover,and mutation.The MPC optimization control method serves as the inner function,and for different weights,the computed output variables are used as input variables for the NSGA’s objective functions.After traversing and optimizing,the optimal weights are obtained.The optimal weights are compared with the existing research results in MPC optimization control regarding three aspects:objective function values,MPC tracking performance,and power fluctuation of the electrolyzer and energy storage battery.The results indicate that the optimal weights obtained in this study have partially reduced the power fluctuation of the electrolyzer and energy storage battery,thereby optimizing the MPC control tracking performance.