工业大系统多模型-分散辨识方法

Industrial Large-scale Systems of Multi-model-decentralized Identification Method

为了更加快速准确控制,使系统更加稳定地工作在最佳工作状态,在已经成熟的分散辨识方法的基础上,通过介绍非线性系统固定多模型控制算法,把两者结合起来,将复杂的非线性问题转化为我们熟悉的线性化问题,近而利用分散辨识方法,得到系统工作的最优工况。且此方法在非线性系统中非常实用,并且辨识精度高。仿真研究表明了此方法的有效性和实用性。

In order to make the system more stable work in the best condition more quickly and accurately,the complex non-linear problem was transformed into the familiar linearization problem by combining the mature decentralized identification method with the nonlinear system fixed multi-model control algorithm.The optimal working conditions of system were obtained by using decentralized identification methods.This method was very useful in nonlinear systems,and had high identification accuracy.Simulation shows the effectiveness and practicability.