论文部分内容阅读
线性控制算法具有计算简单、实时性好等优点,但是工业过程通常具有复杂的非线性特性,而且非线性程度越强,实际过程和模型之间的误差越来越大,常规的线性控制器性能会随着非线性程度的增强而显著降低。对于一般的非线性控制策略而言,它们具有建模成本较高、难以选取合适的权重因子等难点,本文采用分段线性化的方法,将非线性系统过程分为多个分段的线性化模型,将动态矩阵控制策略推广到各个分段线性模型,设计了分段线性化多模型DMC算法,通过组合以实现其预测控制的目的。为了验证该算法的有效性,本文在Matlab/Simulink软件平台上进行了1套强非线性pH值中和过程实验装置的较全面的控制仿真,包括阶跃及方波给定值跟踪、抗强干扰,以及存在较大模型失配时的给定值跟踪性能仿真。仿真结果表明,针对强非线性的过程,分段线性化的DMC比不分段单模型的DMC过程具有更好的控制性能。
The linear control algorithm has the advantages of simple calculation and good real-time performance. However, the industrial process usually has complex nonlinear characteristics. And the stronger the nonlinearity is, the bigger the error between the actual process and the model is. The conventional linear controller performance Will decrease significantly with the degree of non-linearity. For the general nonlinear control strategy, they have the difficulty of modeling the high cost and difficult to select the appropriate weight factors. In this paper, the method of piecewise linearization is used to divide the nonlinear system process into multiple piecewise linearization Model, the dynamic matrix control strategy is extended to each piecewise linear model. A piecewise linearized multi-model DMC algorithm is designed to achieve the purpose of predictive control. In order to verify the effectiveness of the algorithm, a complete set of control simulation of a set of strong non-linear pH value neutralization experimental device is carried out on Matlab / Simulink software platform, including step and square wave setpoint tracking, Interference, and given value tracking performance simulation in the presence of large model mismatch. Simulation results show that, for the strong nonlinear process, piecewise linear DMC has better control performance than non-segmented single-model DMC process.