Redundant Control In AC Servo Systems

In the modern war,with the military technology and the war mode change,the requirement of weapon system reliability is higher than ever before.The performance of the system controller part is the main factor that affecting the reliability of the weapon.In order to improve the performance of artillery,this paper has studied the artillery control system based on the AC servo system of artillery.To meet the high stability and reliability requirements of the AC servo system,a single control system is difficult to meet the requirements.This paper adopt dual-CPU software and hardware redundancy control program.In this way,it can meet the artillery control of AC servo system in the case of strong interference,such as the deployment of reliability.It has some practical value.The main research work is included in the following aspects in this paper.Firstly,making efforts in the analysis of the working principle and composition about redundant control,and several redundancy realization methods are compared and analyzed.On this basis,the redundancy of dual-CPU implementation is studied and analyzed.The redundant implementation of "hot backup" is determined and build the system structure.Then,the key technologies of redundancy implementation including clock synchronization,information synchronization,fault detection and switching are studied.Using STM32F107VCT6 as the controller chip,two sets of hardware circuits are designed to build dual-CPU hardware redundancy.The communication between the master system and the backup system is carried out through the CAN.The analytic fault detection method is used to compare the output of the master system and the backup system in real time.When the threshold error is exceeded,the system will automatically switch to the backup system to ensure the control system stable and reliable.Then,due to the nonlinear characteristics of the AC servo system,design the BP neural network PID control algorithm and the RBF neural network PID control algorithm.The three control parameters Kp,Ki and Kd of PID are adjusted by neural network to improve the control effect.With the MATLAB simulation analysis,the RBF neural network PID control algorithm has better control effect.So the master and backup systems use this control algorithm to ensure consistent output signals.Finally,to design the redundant control software,including the main program module,data acquisition module,CPU synchronization module,fault detection module,and redundant switching module.Through the semi-physical platform of the AC servo system experiment,simulate the fault situation and detect the real-time output values of the two CPU modules.When the main CPU module fails,the output value changes abruptly.The backup CPU module can monitor in real time and switch the output to the backup value,which ensures the reliability of the system.