论文部分内容阅读
本文给出了一种以最小二乘方法进行参数估计加最小方差为控制策略的多变量自校正调节器MVSTR。调节算法是在一台十六位单板机TP-86A 上实现的,文中给出了闭环模拟实验结果。为了改善MVSTR 的动静态特性,我们在典型的MVSTR 中引入了一个非线性的P-I 反馈调节器。该改进是基于给不同运行条件下的闭环系统配以不同的零极点的思想。模拟实验结果,表明这种改进是成功的。本文还就计算机有限字长所致的计算误差使得参数估计发散,进而导致控制发散的问题进行了讨论,并给出了一种易于在数字计算机上实现的抑制参数发散的方法。
In this paper, we present a multivariable self-tuning regulator MVSTR that uses the least square method for parameter estimation plus the minimum variance as the control strategy. The adjustment algorithm is implemented on a single sixteen single board computer TP-86A. The closed-loop simulation results are given in this paper. In order to improve the dynamic and static characteristics of MVSTR, we introduce a nonlinear P-I feedback controller in a typical MVSTR. This improvement is based on the idea of equipping the closed-loop system under different operating conditions with different zero poles. Simulation results show that this improvement is successful. In this paper, we also discuss the problem of control divergence caused by the calculation error caused by limited computer word length, and then give a method to suppress the parameter divergence which is easy to implement on digital computer.