| Synthetic aperture radar is a radar that obtains high-resolution images similar to traditional optical photography under extremely low-visibility meteorological conditions.However,the necessary condition for obtaining a clear image is that the radar antenna maintains a uniform linear motion in the reconnaissance area.However,during the working process,the radar antenna continuously deviates from the original working position under the disturbance of airflow,which causes the image generated by the radar antenna to be blurred.The airborne radar antenna stabilization platform is a device that isolates antenna jitter and keeps the radar antenna working stably and having a clear image.On the premise that the mechanical processing technology and assembly combination of the stable platform meet the technical requirements,the stability accuracy of the stable platform is further improved through the optimized control algorithm of the motor.This paper takes the CZ-08 DF airborne radar antenna stabilization platform as the research object,and proposes an improved PID neural network control strategy for the defects of traditional PID control.The frame structure and working mechanism of the CZ-08 DF airborne radar antenna stabilization platform are analyzed,and various interference factors on the stabilization platform are studied.Through the platform system coordinate system and the coordinate transformation matrix,the angular velocity disturbance isolation conditions are determined,and the angular velocity amount that needs to be compensated on the stable platform is obtained through analysis.The DC servo motor,PWM power amplifier circuit and inertial element in the stable platform system are mathematically modeled respectively,the mathematical model of the system is analyzed,and the double-loop control of digital speed loop and digital position loop is designed.The PID control and neuron network theory are analyzed,and the structure,forward algorithm,back-propagation algorithm and initial value of network connection weight of PID neuron network are studied.In view of the shortcomings of the neural network error back-propagation algorithm,an improved PID neuron network control is proposed.On the basis of the original algorithm,the momentum term,the adaptive learning rate and the improvement method of the excitation function are added.The model of the CZ-08 DF airborne radar antenna stabilization platform control system is simplified,and the specific transfer function is obtained from the motor parameters.Based on MATLAB software,the PID controller and the improved PID neural network controller are simulated and compared in terms of step response and disturbance suppression capability.The improved PID neural network control algorithm has better performance in the disturbance suppression of radar antennas and stabilizes the platform.The stabilization accuracy is higher,and the radar antenna imaging is clearer. |
PDF Full Text Request