论文部分内容阅读
针对常规无模型自适应控制(MFAC)算法只利用了受控系统I/O数据的零阶信息,许多系统内部信息没有充分利用的缺陷,提出了一种基于紧格式线性化的无模型动态矩阵控制方法。该方法利用紧格式线性化方法把非线性系统转化为一组由伪梯度向量描述的线性系统,将动态矩阵控制算法的预测模型和该伪梯度向量相结合,推导出无模型动态矩阵控制的动态模型。在此基础上,利用滚动优化和反馈校正等策略,推导出了无模型动态矩阵控制律的解析式。由于该方法中引入了受控系统I/O数据所带的系统内部信息,极大地提高了预测模型的精确性,增强了系统的鲁棒性,从而有效改善了控制器的控制性能。仿真结果表明该算法具有一定的结构自适应性、良好的伺服性和较强的鲁棒性。
In view of the fact that the MFAC algorithm only uses the zero-order information of the I / O data of the controlled system and many of the system’s internal information is not fully utilized, a compact model-based dynamic matrix Control Method. The method uses a tight linearization method to transform a nonlinear system into a set of linear systems described by pseudo-gradient vectors. By combining the predictive model of the dynamic matrix control algorithm with the pseudo-gradient vector, the dynamics of model-free dynamic matrix control model. On this basis, using the methods of rolling optimization and feedback correction, the analytic formula of modelless dynamic matrix control law is deduced. Because this method introduces the system internal information brought by the I / O data of the controlled system, the accuracy of the predictive model is greatly improved and the robustness of the system is enhanced, so as to effectively improve the control performance of the controller. Simulation results show that the proposed algorithm has some structural adaptability, good servo and strong robustness.