论文部分内容阅读
为解决工业过程控制领域中非线性系统的模型辨识与预测控制问题,提出一种基于BP神经网络模型的预测控制策略,采用一种分段最小二乘支持向量机辨识Hammerstein-Wiener模型系数的方法建立非线性预测控制器.利用BP神经网络训练预测控制输入序列和拟牛顿算法求解非线性预测控制律,从而实现了一种基于支持向量机Hammerstein-Wiener辨识模型的非线性预测控制算法.仿真实验验证了该算法的有效性和可行性.
In order to solve the problem of model identification and predictive control of nonlinear systems in the field of industrial process control, a predictive control strategy based on BP neural network model is proposed. A method of piecewise least square support vector machine to identify Hammerstein-Wiener model coefficients A nonlinear predictive controller based on support vector machine (Hammerstein-Wiener) identification model is developed by using BP neural network to train predictive control input sequence and quasi-Newton algorithm to solve nonlinear predictive control law.The simulation experiment The validity and feasibility of the algorithm are verified.