基于多层局部回归神经网络的多变量非线性系统预测控制

以罐式搅拌反应器为例，针对复杂多变量系统的强耦合性、非线性、时变性等问题，研究了多变量非线性系统的预测控制及改善控制性能的方法，采用多层局部回归神经网络离线建立预测模型，以偏差补偿和模型修正相结合的方式对预测模型进行误差补偿，以要线校正用于预测控制，通过对性能指标中的偏差项负指数加权，进一步改善预测控制性能，住址结果表明了控制算法的有效性。

Predictive Control of Multivariable Nonlinear System Based on Multilayer Local Recurrent Neural Networks

Taking the stirred tank reactor for example, the predictive control of MIMO nonlinear system based on \{multilayer\} local recurrent neural networks is presented. Aiming at the difficulties in modeling the complex MIMO nonlinear system, the multilayer local recurrent neural network is used to build the predictive model of the process off line. In feedback correction, considering the requirements of the accuracy and practicability, error compensation and model correction are adopted to correct the predictive model online for the predictive control. We draw the conclusion that negative exponential weighting of future tracking errors can improve the control performance of the control systems. The results of simulation show the effectiveness of the control algorithm.