基于自回归小波神经网络的空中目标威胁评估

Aerial Target Threat Assessment Based on Self-recurrent Wavelet Neural Network

针对空战中目标威胁评估系统非线性、评估难度大等特点,提出了自回归小波神经网络(Self-recurrent Wavelet Neural Network,SRWNN)的空中目标威胁评估方法。通过分析SRWNN结合递归神经网络(Recurrent Neural Net RNN)的吸引子动力学和WNN快速收敛的特点,建立了SRWNN模型,提出了SRWNN的参数优化学习算法,以实现增强自学习能力的目的,然后分析了威胁评估的影响因素,给出了基于SRWNN的空中目标威胁评估算法的程序设计。仿真实验结果表明,与WNN相比,该算法提高了系统的稳定性,加快了收敛速度,增强了预测精度。

Aiming at the characteristics of the non-linearity of the target threat assessment system in air combat and the difficulty of evaluation, an aerial target threat assessment method based on self-recurrent wavelet neural network was proposed. The SRWNN model is combined with the attractor dynamics of recursive neural network and the characteristics of fast convergence of WNN. The SRWNN model was established, and the parameter optimization learning algorithm of SRWNN was proposed to achieve the purpose of enhancing the self-learning ability. By analyzing the influencing factors of threat assessment, the program design of air target threat assessment algorithm based on SRWNN is given. Simulation results show that compared with WNN, this algorithm improves the stability of the system, speeds up the convergence speed, and enhances the prediction accuracy.