基于改进差分进化算法的约束动态矩阵控制
|
摘要
针对约束动态矩阵控制的优化问题求解,给出一种改进的差分进化算法。首先,为了平衡算法的收敛速度和全局搜索能力以及提升其稳定性的目的,采用了耦合-最优排序的变异策略以及最优个体保留方法。其次,设计了特定的进化操作、优化与控制一体的目标函数来处理约束条件。最后以带约束条件的三容液位系统为研究对象,仿真验证了本文提出的控制算法较传统的约束动态矩阵控制算法有更强的抗干扰能力、无静差等优点以及改进的差分进化算法比原来的动态差分进化算法有更好的稳定性及更快的收敛速度。
To solve the optimization problem of constrained dynamic matrix control,an improved differential evolution algorithm was proposed. Firstly,in order to balance convergence speed and global searching ability and enhance stability of the algorithm,the coupled-optimal ordering mutation operation and optimal individual retention strategy were utilized. Secondly,a specific evolution operation and objective function that combines optimization and control were designed to deal with the constraints. Finally the three-tank liquid level system with constraints was taken as research object. The simulation verifies the control algorithm that has advantages of no static error and stronger anti-interference ability than traditional constrained dynamic matrix control algorithm,and validating improved differential evolution algorithm has better stability and faster convergence rate than original dynamic differential evolution algorithm.
To solve the optimization problem of constrained dynamic matrix control,an improved differential evolution algorithm was proposed. Firstly,in order to balance convergence speed and global searching ability and enhance stability of the algorithm,the coupled-optimal ordering mutation operation and optimal individual retention strategy were utilized. Secondly,a specific evolution operation and objective function that combines optimization and control were designed to deal with the constraints. Finally the three-tank liquid level system with constraints was taken as research object. The simulation verifies the control algorithm that has advantages of no static error and stronger anti-interference ability than traditional constrained dynamic matrix control algorithm,and validating improved differential evolution algorithm has better stability and faster convergence rate than original dynamic differential evolution algorithm.
引文
[1]席裕庚,李德伟,林姝.模型预测控制---现状与挑战[J].自动化学报,2013,39(3):222-236.
[2]席裕庚.预测控制[M].北京:国防工业出版社,2013,(12):1-16.
[3]李俊利,杨绍武,袁平,等.基于动态矩阵控制的燃煤锅炉蒸汽压力控制[J].控制工程,2016,23(11):1685-1689.
[4]焦巍,刘光斌.非线性模型预测控制的智能算法综述[J].系统仿真学报,2008,20(24):6581-6586.
[5]汪慎文,丁立新,张文生,等.差分进化算法研究进展[J].武汉大学学报(理学版),2014,60(4):283-292.
[6]张春美.差分进化算法理论与应用[M].北京:北京理工大学出版社,2014:1-7.
[7]王毓栋,王京春.基于DE算法的非线性预测控制及其应用[J].控制工程,2008,15(1):25-28.
[8]朱腾,王京春,熊智华.基于改进DE-NMPC的酸碱中和反应p H值控制[J].自动化学报,2010,36(1):159-163.
[9]Danzomo B A,Salami M J E,Khan M R.Identification and predictive control of spray tower system using artificial neural network and differential evolution algorithm[C].Control Conference.IEEE,2015:1-7.
[10]Negri G H,Preuss V H B,Cavalca M S M,et al.Differential evolution optimization applied in multivariate nonlinear model-based predictive control[C].Latin America Congress on Computational Intelligence.IEEE,2015:1-6.
[11]Negri G H,Cavalca M S M,Parpinelli R S.Model-Based Predictive Control Using Differential Evolution Applied to a Pressure System[J].IEEE Latin America Transactions,2016,14(1):89-95.
[12]Yu X,Huang D,Wang X,et al.DE-based neural network nonlinear model predictive control and its application for the p H neutralization reactor control[J].Chinese Control and Decision Conference,2008:1597-1602.
[13]刘振泽,许洋,王峰明.改进差分进化算法在非线性模型预测控制中的应用[J].北京工业大学学报,2015(5):680-685.
[14]陈虹.模型预测控制[M].北京:科学出版社,2013:70-149.
[15]吴亮红,王耀南.动态差分进化算法及其应用[M].北京:科学出版社,2014:9-20.
[16]Qin,A.K,Huang,et al.Differential evolution algorithm with strategy adaptation for global numerical optimization[J].IEEE Transactions on Evolutionary Computation,2009,13(2):398-417.
[2]席裕庚.预测控制[M].北京:国防工业出版社,2013,(12):1-16.
[3]李俊利,杨绍武,袁平,等.基于动态矩阵控制的燃煤锅炉蒸汽压力控制[J].控制工程,2016,23(11):1685-1689.
[4]焦巍,刘光斌.非线性模型预测控制的智能算法综述[J].系统仿真学报,2008,20(24):6581-6586.
[5]汪慎文,丁立新,张文生,等.差分进化算法研究进展[J].武汉大学学报(理学版),2014,60(4):283-292.
[6]张春美.差分进化算法理论与应用[M].北京:北京理工大学出版社,2014:1-7.
[7]王毓栋,王京春.基于DE算法的非线性预测控制及其应用[J].控制工程,2008,15(1):25-28.
[8]朱腾,王京春,熊智华.基于改进DE-NMPC的酸碱中和反应p H值控制[J].自动化学报,2010,36(1):159-163.
[9]Danzomo B A,Salami M J E,Khan M R.Identification and predictive control of spray tower system using artificial neural network and differential evolution algorithm[C].Control Conference.IEEE,2015:1-7.
[10]Negri G H,Preuss V H B,Cavalca M S M,et al.Differential evolution optimization applied in multivariate nonlinear model-based predictive control[C].Latin America Congress on Computational Intelligence.IEEE,2015:1-6.
[11]Negri G H,Cavalca M S M,Parpinelli R S.Model-Based Predictive Control Using Differential Evolution Applied to a Pressure System[J].IEEE Latin America Transactions,2016,14(1):89-95.
[12]Yu X,Huang D,Wang X,et al.DE-based neural network nonlinear model predictive control and its application for the p H neutralization reactor control[J].Chinese Control and Decision Conference,2008:1597-1602.
[13]刘振泽,许洋,王峰明.改进差分进化算法在非线性模型预测控制中的应用[J].北京工业大学学报,2015(5):680-685.
[14]陈虹.模型预测控制[M].北京:科学出版社,2013:70-149.
[15]吴亮红,王耀南.动态差分进化算法及其应用[M].北京:科学出版社,2014:9-20.
[16]Qin,A.K,Huang,et al.Differential evolution algorithm with strategy adaptation for global numerical optimization[J].IEEE Transactions on Evolutionary Computation,2009,13(2):398-417.