基于混沌神经网络预测模型的最优控制决策及应用

为了实现非线性、大时滞系统的自适应控制，首先根据具有混沌特性的非线性、大时滞系统的时序列重构相空间，计算出相空间的饱和嵌入维数和最大Lyapunov指数，并以此为指导，建立混沌神经网络预测模型，该模型即便在网络输入不完整或发生变异的情况下，仍能对系统作高精度的短期预测；在此基础上，将预测模型的输出通过反馈校正．再将校正误差和控制增量引入性能函数最优，最后得到最优控制决策，实现了对非线性、大时滞系统高精度的自适应控制。最后将预测控制决策应用到非线性、大时滞的锅炉过热汽温控制中，仿真结果表明了该控制的有效性、快速性和鲁棒性。图6参6

Optimal Control Based on Chaos Neural Network Forecast Model and Its Application

In order to realize adaptive control of nonlinear big-lagged system, a chaotic attractors space is reconstructed, systemic embed dimension and maximal Lyapunov exponent are calculated by the chaotic time series of the nonlinear big-lagged system in this paper. By above all, a chaos neural network model is constructed, which can make high precisin short-term forecast for the nonlinear big-lagged system even by imperfect and variation inputs. On the basis of this, a feedback rectification term is achieved by the forecast model outputs, a optimal controller is designed by the feedback rectification term and control input error being introduced into a performance functin, and robust adaptive forecase control to the nonlinear big-lagged system is realized. Last, the controller is applied to nonlinear big-lagged overheat temperature system of drum boiler, the validity, the high-speed and the robustness are proved by simulate results. Figs 6 and refs 6.