基于满意优化的非线性预测控制

对多约束的非线性系统建立T—S模糊模型，采用局部线性化方法在每个采样点对非线性系统进行线性化，从而得到系统的线性化模型。针对传统的优化方法无法直接处理带有一定模糊不确定性的优化问题，该文在广义预测控制滚动优化的机制下把具有模糊边界约束的有限预测时域的优化问题，转化为等价的确定性规划问题，通过模糊规划方法来求解多约束的目标函数，从而用满意优化取代传统的二次型性能指标来求解模糊约束条件的预测控制。仿真结果表明了该方法的有效性。

Nonlinear Predictive Control Based on Satisfactory Optimization

A T - S fuzzy model is constructed for multi - constrained nonlinear system, and local linearization at each sampling point is applied to the system, then the linearlzation model of system is attained. The traditional optimization fails to handle the optimal problem for fuzzy uncertainty. In this paper, the optimization problems of limited time domain with the fuzzy boundary constraints in the GPC are transferred into the equivalent determined problems. The method of fuzzy program in satisfactory control gains result of multi - constrained objective function. The traditional quadratic performance target is substituted by satisfactory optimization control in predictive control for gaining results based on fuzzy constraint conditions. The simulation results demonstrate the effectiveness of this method.