基于RBF网络的UCAV战术机动轨迹快速生成方法

Rapid generation method of UCAV tactical maneuver trajectory based on RBF network

针对无人作战飞机(unmanned combat aerial vehicle,UCAV)战术机动动作数学表征困难,机动生成的计算实时性要求高的主要问题,分析了战术机动轨迹建模的基本原理;提出了战术机动轨迹建模的基本思路;设计了基于UCAV运动动力学模型的机动轨迹最优控制方案;建立了基于遗传算法的飞行操控量求解策略。针对操控量求解的实时性问题,基于径向基核函数神经网络,提出了以适应度函数为预测和评判标准的机动飞行操控量快速求解方法,从而建立了初始状态、性能指标与机动飞行操控量的非线性映射模型,实现了机动轨迹的快速生成和机动曲线的精度控制,并通过仿真验证了该方法的有效性。

Aiming at the main issues of the difficult mathematical representation of the unmanned combat aerial vehicle(UCAV)tactical maneuvering,and the high maneuver-generated computational real-time requirements,the paper analyzes the basic principle of tactical maneuvering trajectory modeling,proposes the basic idea of tactical maneuver modeling,designs the optimal trajectory control scheme based on UCAV kinematic model,and the strategy of solving flight control based on genetic algorithm is established.Focusing on the real-time problem of control quantity solution,based on the radical basis function neural networks,the paper proposes a fast method of solving flight control quantity based on fitness function as prediction and evaluation criterion,and a nonlinear mapping model of initial state,performance index and flight control quantity is established,realizes the rapid generation of maneuvering trajectory and the precision control of maneuver curve,and the effectiveness of this method is verified by simulation.