Attitude stabilization system of laser weapons based on adaptive neural network algorithm

For the time-vary and non-linearity of attitude stabilization system of laser weapons, performances of traditional adaptive control algorithms are not satisfactory. According to the mathematical model of attitude control stabilization loop and the characteristics of operating conditions of laser weapons, an approach of neural network PID model reference adaptive control based on RBF (Radial Basis Function) on-line identification is presented. The result of simulation shows that not only the dynamic and static characteristics of the system can satisfy the index requirements, but also the control effect is improved as compared with the traditional PID control method.