基于神经网络的非线性动态系统的多内模控制研究

由于非线性动态系统的复杂性,目前还没有统一的控制器设计方法,用传统的线性系统理论来设计,当工作点大范围变化时,很难保证其性能及稳定性.在人工神经网络内模控制系统中建立了神经网络内部模型和神经网络内模控制器,提出了基于多模型的内模控制方法,它对非线性动态过程的控制具有良好的性能.在此基出上,给出了方便易行控制算法,仿真分析结果验证了控制算法的有效性.

Multiple Internal Model Control Study for Nonlinear Dynamic System Based on Neural Network

There is not the unified designing method of the controller at present because of the complication of the nonlinear dynamic system. It is difficult for the control process that is designed by traditional linear theory to ensure good performance and stability over a wide range of operating conditions. Internal model and internal model controller have been built in the control system of artificial neural network, and the control method is presented based on many internal models, which is with superior control function for nonlinear dynamic process. Based on this, the control algorithm presented can be easily realized. Simulation result verifies the effectiveness of the proposed method.