基于SMC的模型预测控制性能验证方法研究  

Research on Performance Verification Method of Model Predictive Control Based on SMC

作　　者：袁武[1] 曾山 YUAN Wu;ZENG Shan(Wuhan Technical College of Communications,Wuhan 430065,China;School of Mathematics&Computer Science,Wuhan Polytechnic University,Wuhan 430023,China)

机构地区：[1]武汉交通职业学院,湖北武汉430065 [2]武汉轻工大学数学与计算机学院,湖北武汉430023

出　　处：《智慧电力》2020年第11期74-79,共6页Smart Power

基　　金：国家自然科学基金资助项目(U1833119)。

摘　　要：针对FCSMPC性能验证问题,引入了一种基于统计模型检测(SMC)的验证方法。以三相逆变器及其FCSMPC为例,完成了FCSMPC算法设计,并通过Simulink基准模型开展了仿真计算,并进行了实验。在UPPAAL SMC工具箱中,将系统建模为随机时间自动机网络,并与等效的Simulink基准模型和实验结果相互映证。通过UPPAAL SMC验证系统完成了对FCSMPC算法中控制参数设计和鲁棒性的查询概率检验,检验结果验证了系统动态和模型参数扰动下FCSMPC算法的控制性能。Targeting the problem of the performance verification of finite control set model predictive control(FCSMPC),the paper proposes a verification method based on statistical model checking(SMC).Taking three-phase inverter with FCSMPC as an example,the design of the FCSMPC algorithm is done,the simulation calculation is made with Simulink benchmark model,and the experiment is carried out.In UPPAAL SMC toolbox,the system is modeled as a network of stochastic timed automata,which has a mutual corroboration with the equivalent Simulink benchmark model and experimental results.The UPPAAL SMC verification system is used to complete the query probability verification of the control parameter design and robustness for the FCSMPC algorithm.The results verify the control performance of the FCSMPC algorithm with the system dynamics and model parameter disturbance.

关 键 词：有限集模型预测控制 控制性能 统计模型检测 逆变器 

分 类 号：TM464[电气工程—电器]