基于多维泰勒网的空空导弹惯性参考自适应控制

Inertial Reference Adaptive Control of AAM Based on Multi-dimensional Taylor Network

针对空空导弹飞行控制系统的模型不精确、快时变、非线性等特点,提出了基于多维泰勒网的惯性参考自适应控制方案。首先,建立特征点处的空空导弹数学模型;然后基于三通道解耦模型,设计多维泰勒网控制器,将其与PID控制器和H∞控制器进行控制效果比较,并对这三种控制器的参数均采用粒子群算法来优化。最后,对每种控制器的参数均采用惯性参考自适应调整策略,通过空空导弹六自由度模型的大机动仿真实验来比较三种控制方法的控制效果。仿真实验结果表明,在这三种控制方案中基于多维泰勒网的惯性参考自适应控制的控制效果最好。

An inertial reference adaptive control scheme based on multi-dimensional Taylor network is proposed to solve the problem of imprecision,fast time-varying and nonlinearity of air-to-air missile flight control system model.Firstly,the mathematical model of air-to-air missile at feature point is established.Then,based on the three-channel decoupling model,a multi-dimensional Taylor network controller is designed,and its control effect is compared with those of PID controller and H∞one.The parameters of these three controllers are optimized by particle swarm optimization.Finally,the inertial reference adaptive adjustment strategy is adopted for each controller’s parameters,and the control effects of the three control methods are compared through the large maneuver simulation experiment of the air-to-air missile model with six degrees of freedom.Simulation results show that the control effect of inertial reference adaptive control based on multi-dimensional Taylor network is the best in the three control schemes.