Research On Actuator Control System Of Guided Rocket Jet Vane Based On FNN-PID

The jet vane of guided rocket is one of the thrust vector control devices.The jet vane control system has become the core executive body for its low quality,small size,torque and other characteristics.However,it is difficult to improve the control accuracy for some defects of their own,such as the nonlinear and time-varying features.In this paper,the research for actuator control system of guided rocket jet vane is carried out and a way of the actuator system design method based on the fuzzy neural network PID(FNN-PID)control is proposed.And the system simulation and the experimental research have been completed.The main work contents and related research results are as follows.(1)The structural characteristics and working principle of the brushless DC servo actuator are analyzed in detail.This paper takes the model MX-106 high-precision actuator as the core of the executive component,establishing the various parts of mathematical model the actuator system.Furtherly,the transfer function of actuator control system is deduced.Finally,a simulation model of brushless DC actuator system is built in Matlab simulation platform.(2)The brushless DC actuator system adopts three closed-loop control structure.The actuator system PID controller,fuzzy PID controller and fuzzy neural network PID controller are designed.And the working principle of the above three kinds of control methods are analyzed.Using Matlab simulation toolbox to build three kinds of controller simulation model,the simulation test is carried out in the no-load condition and the loaded condition respectively.The response characteristics and tracking characteristics of the system are analyzed.And the differences between the control effects of different controllers are compared.(3)This paper develops a set of ARM actuator control panel taking LPC2292 as the core.The function and realization of some modules is introduced including PWM generation module,isolation and drive module and communication interface module etc.Software program is designed on the slave computer to realize the application of the three control algorithms and data acquisition system is developed on the upper computer based on LAB VIEW.Through the feasibility analysis of the load,a special torsion spring is designed to simulate the load of jet vane,establishing the rudder deflection angle detection platform.(4)A jet vane test system platform is set up.And the PID controller,fuzzy PID controller and fuzzy neural network PID controller are applied to the test system separately.On the basis of this system,prototype test is conducted under both no-load and loading situation.The simulation and experimental result indicates that compared to these two actuator systems adopting the PID controller and fuzzy PID controller,the actuator system with fuzzy neural network PID controller has short response time,small overshoot,small phase shift,better tracking and response characteristics.The control method has a strong adaptability to the external disturbance and has good robustness,which greatly improves the control performance of the actuator system.