基于并行支持向量机的多变量非线性模型预测控制

提出一种基于并行支持向量机的多变量系统非线性模型预测控制算法．首先，通过考虑输入、输出间的耦合，建立基于并行支持向量机的多步预测模型；然后，将该模型用于非线性预测控制,提出新的适用于并行预测模型的反馈校正策略，得到最优控制律．连续搅拌槽式反应器（CSTR）的控制仿真结果表明,该算法的性能优于基于并行神经网络的非线性模型预测控制和基于集成模型的非线性模型预测控制．

Multivariable nonlinear model predictive control based on parallel support vector machines

A nonlinear model predictive control algorithm based on parallel support vector machines for multivariable systems is presented. Multi-step predictive models based on parallel support vector machines are established by considering the coupling effects among outputs and inputs. Then, the predictive models are applied to nonlinear predictive control, a novel feedback correction strategy suitable for the models is presented, and optimal predictive control law is obtained. Simulation results of CSTR control show that the presented algorithm has better performance than nonlinear model predictive control based on parallel neural networks and integrating model.