基于最优控制的ANN驾驶员模型与仿真分析

ANN Model for Pilot Based on Optimal Control and Simulation Analysis

摘要: 在分析驾驶员行为特性和行为操纵的基础上，根据预测跟随理论，建立了驾驶员预测控制神经网络（ANN）模型；提出了用最优控制方法确定ANN模型参数的计算方法，采用遗传算法（GA）进行全局优化保证参数的收敛．对飞机俯仰角操纵进行仿真计算．结果表明，所建立的驾驶员模型考虑了系统的非线性因素，实现了多输入多输出功能，具有智能特点．

Abstract: Based on the analysis of pilot behavior character and control mode,the predictive control ANN(artificial neural network)model is established according to the theory of prediction-following.Method to calculate the parameters of ANN model is presented by optimal control.GA(genetic algorithm)is used in global optimization to ensure the convergerce of parameters.An example of controlling attack angle is simulated.Numerical simulation shows that nonlinear characters are considered,and MIMO(multi-input multi-output)function is realized.The model has intelligent characteristics.