基于T-S模糊模型的约束预测控制研究

【作者】 董雪雅；

【导师】 麦继平；

【作者基本信息】 天津工业大学 ， 控制理论与控制工程， 2008， 硕士

【摘要】 基于线性模型的预测控制研究已经相当成熟并得到了广泛的工业应用,然而在实际控制系统中,被控对象往往具有非线性、时变性和不确定性,针对高度非线性系统则很难取得令人满意的控制效果。而T-S模糊模型可以以任意精度逼近非线性系统,并且由于其结构的特殊性,可以通过局部动态线性化,把非线性系统表示成为线性形式,使得对其设计GPC控制器成为可能。本文首先简要介绍了广义预测控制的基本原理、基本算法步骤和模糊系统模型的基本原理,然后针对非线性系统提出了几种基于T-S模糊模型的有约束广义预测控制算法,并通过matlab仿真试验验证了这些算法的有效性。全文主要创新点有:1.针对单变量非线性系统,在利用T-S模糊模型充分逼近的基础上,基于一步近似计算的思想,通过对输入变量的适当处理,提出了一种带约束输入的快速广义预测控制算法,该方法充分考虑了控制输入及其增量受约束的情况,并且计算量不大。2.针对多变量非线性系统,基于T-S模糊模型建模,提出了一种带约束输入的快速广义预测控制算法,该方法充分考虑了控制输入及其增量受约束的情况,并避免了非线性搜索方法求解受约束的优化问题和求Diophantine方程,并且计算量不大。在本文的最后,总结全文,并提出在该方向上需进一步做的工作。更多还原

【Abstract】 Research on linear predictive control has become mature and linear MPC has gained wide applications in industrial processes. However, mostprocesses in industry are nonlinear, time-variant and bear uncertainty,for a highly nonlinear system, it may not give rise to satisfactory dynamic performance. The T-S fuzzy model can approaching any nonlinear systems, and its structure is simple. So the controlled plant can be expressed as a linear model, then the GPC controller can designed for it. In this dissertation, the development about predictive control and the T-S fuzzy system’s strongpoints and deficiencies are introduced firstly. The basic identification steps of Takagi-Sugeno fuzzy model are presented in detail. And then some new algorithms of constrained fuzzy generalized predictive control are presented for nonlinear systems. The simulation results show their superior performance for nonlinear systems. In conclusion the main contents are as follows:1) one kind of quickly constrained generalized predictive control algorithm is presented based on the T-S fuzzy model which is used to approach the SISO nonlinear systems, its computer load is not too large;2) one kind of quickly constrained generalized predictive control algorithm is presented based on the T-S fuzzy model which is used to approach the MIMO nonlinear systems,it avoids the nonlinear search and need not to solve Diophantine functions by using a soft gene of input,so its computer load is not too large;In the last section of this dissertation, a conclusion is presented, and some jobs needed to be done in the future are drawn.更多还原