基于自适应神经模糊系统的高超声速飞行器再入预测制导

针对高超声速飞行器再入运动过程模型的非线性特性，提出了一种基于自适应神经模糊系统(ANFIS)的再入预测校正制导方法。在以能量为自变量的三自由度再入方程的基础上分别设计了纵向制导律和侧向制导律。以能量和剩余航程偏差为输入参数，侧倾角调节量为输出参数，设计了ANFIS控制器，并将其应用于纵向制导。侧向制导基于横程与能量的近似线性关系，设计了由分段漏斗形横程走廊控制的侧倾角反转逻辑。仿真结果表明，所设计的制导律具有制导指令解算速度快，制导和落点精度高且对再入初始偏差及过程扰动不敏感的优点。

ANFIS-based Predictive Reentry Guidance for Hypersonic Vehicles

An adaptive neuro-fuzzy inference system (ANFIS) based predictor-corrector guidance is proposed for the nonlinearity of the reentry kinematic equations of hypersonic vehicles. The energy-based three degrees of freedom reentry kinematic equations are established, and the longitudinal and lateral guidance laws are designed. For the longitudinal guidance, an ANFIS controller is designed by taking energy and range error as system inputs and bank angle variation as system output. The lateral guidance is based on the approximate linearity between crossrange and energy. A lateral reversal logic based on the cross range corridor is designed. Simulation results show that the proposed guidance law has the advantages of less calculating time and high guidance precision, and is not sensitive to the initial errors and disturbances.