灰色误差校正多变量动态矩阵控制

针对多变量系统耦合严重和特性时变的特点,提出了一种灰色误差校正多变量动态矩阵控制方法。该方法将原系统分解为多个多输入单输出子系统,根据动态矩阵控制设计了多个子系统的控制器,对多个控制器方程联立求解得到预测控制增量序列。以灰色多步误差预测校正替代原有算法的当前一步误差校正,用来克服对象时变所造成的模型失配的影响,并给出了灰色多步误差预测方法的稳定性分析。对300MW机组负荷系统的仿真表明,该方法减小了负荷和主汽压间的耦合;当对象特性变化时,能够实现超前调节,缩短调节时间。

Gray error correction multivariable dynamic matrix control

Gray error correction multivariable DMC(dynamic matrix control) control is presented in order to overcome the multivariable system's shortcomings of serious coupling and time-varying parameters.In this method,multivariable system was firstly decomposed into several multiple-input and single-output subsystems and several DMC controllers were designed accordingly.By solving the DMC controllers' equations,predictive control series were obtained.Gray multiple step error correction method took the place of the conventional single step error correction method to minimize the impact of model mismatch when system's characteristic changed.The stability of the gray error correction method was also presented.The simulation of 300MW unit load system proves that the new method can not only lighten the system's coupling problem but also realize leading regulate and shorten the regulate time.