多变量网络化DMC算法的鲁棒稳定性研究

摘要

在工业控制系统优化算法的研究中,引入网络化控制系统,针对多输入多输出动态矩阵控制(DMC)算法的鲁棒稳定性问题,首先根据有限脉冲响应(FIR)模型导出了网络控制系统通信信道上存在随机网络诱导时延的网络化DMC算法；其次,在考虑预测模型与系统实际模型之间存在模型失配情况下,利用线性矩阵不等式方法( LMI)分析该网络化DMC算法的鲁棒稳定性问题,进而建立了网络化DMC算法的鲁棒稳定性的充分条件.数值仿真验证表明解决工业控制系统性能优化的有效性.%The problem of robust stability of multi-input and multi-output dynamic matrix control ( DMC) algorithm was investigated for networked control systems. Firstly, based on finite impulse response (FIR) model, a networked DMC algorithm was obtained for the system where there exists random network-induced delay on the communication channel. Secondly, when the model mismatch between prediction model and system model was considered, and the robust stability property of the networked DMC algorithm ws analyzed with the method of linear matrix inequality (LMI). Then, a sufficient condition of robust stability was established for the networked DMC algorithm. Finally, a numerical simulation was used to demonstrate the effectiveness of the results obtained here.