Research On The Control Of Anti-jamming Position Following System Of A Rocket Weapon

With the development of national defense technology,rocket weapon system for the automation level and the tracking accuracy requirements are increasingly high.The position servo system is a necessary part of the whole system.In order to obtain good control performance,we must put forward higher requirements on the position servo system.When the rocket launches,the centroid position,stiffness,damping and moment of inertia of the system are greatly changed.The system parameters are uncertain,and it is affected by the continuous gas flow in the launch condition,which cause the vibration of the directional device and the subsequent launch in this environment to reduce the hit accuracy.Therefore,how to overcome the disturbance and uncertainty of the system,and to improve tracking accuracy and anti-jamming capability of the rocket position servo system,are the current problem needed to study.In this paper,based on a certain type of rocket position servo system,the mechanical structure of multiple rocket launcher and the composition and working are introduced,and mathematical model of AC servo motor and the mathematical simulation model of a certain type of rocket AC servo system are established.Besides,the load disturbance of the position servo system is analyzed,which lays the foundation for the identification and control strategy.The off-line training and online adjustment identification method is adopted.RBF neural network is used to identify system firstly.Considering the problem of certain parameters of RBF neural network,an improved particle swarm algorithm is used to optimize the centers,widths and weight.The off-line trained parameters is regard as the initial value of the online identifier,which avoids the vibration phenomenon and accelerates up the convergence speed.In order to restrain the load disturbance of the position servo system,the disturbance of the position servo system is reduced.A single neuron Auto Disturbance Rejection Controller(ADRC)based on RBF neural network is designed.By using the improved fal function of extended observer in the ADRC,the gas flow impulse,system friction and load change of the system are attributed to the extended state.Taking into account the many parameters of the auto disturbance rejection controller and the difficulty to determine the parameter,in this paper,a single neuron controller(SNC)is used to replace the nonlinear state error feedback controller(NLSEF),whose weight uses the identification information of the RBF neural network online identifier to adjust automatically online.Combined with the national 973 project of the experiment,the proposed control strategy is verified in the experiment.The experimental results show that the control strategy can effectively restrain the load disturbance and have strong anti-interference ability.