一种基于多面体描述系统的非线性预测控制

A Nonlinear Model Predictive Control Based on Polytoptic Description System

利用泰勒级数对系统线性化;采用对每个偏导数取最大最小的方法构造多面体,描述系统包裹原非线性系统;对由多面体描述系统各个顶点构成的多个线性预测模型设计相同的反馈控制器,该反馈控制器可以通过有线性矩阵不等式约束的半正定线性规划得到,满足系统实时优化的需要.系统的稳定性由多面体描述系统的稳定性保证.

Presented in the paper is a nonlinear model predicative control method. On the basis of which the nonlinear system is linearized through the tailor series, the polytoptic description systems including initial nonlinear system are constructed by using the maximum and minimum of each partial derivative and the same feedback control law is designed on each vertex of the polytoptic description system. The control law can be obtained through a semi-definite programming with several linear matrix inequalities, which can satisfy the demand of online optimization, and the stability of nonlinear system is based on that of the polytoptic description system.