Robust Sliding Mode Control Guidance Law for Interceptor Based on Wavelet Neural Networks

In this paper a wavelet neural networks (WNN) sliding mode control guidance law with strong robustness is proposed for interceptor considering the dynamics of autopilot. Firstly, using the robust ability of sliding mode control to uncertainties in the control process, a typical three-dimensional sliding mode control guidance law (SMCGL) based on line-of-sight (LOS) is presented. And then wavelet neural networks are employed to predict the uncertain system dynamics on line to relax the requirement of uncertainty bound in the design of a traditional sliding mode controller. The adaptive learning algorithm of the WNN is derived from the sense of Lyapunov stability analysis. Numerical simulation results show that the dynamic behaviors of the proposed guidance law are superior to typical SMCGL in term of miss distance and obviously reduce the chattering phenomenon.