Model based PID tuning of Antenna Control System for tracking of Spacecraft

Abstract

Proportional, Integral and Derivative (PID) Controller are widely used in tracking mechanism of spacecraft. PID controller popularity is attributed to its simplicity but optimum tuning of PID parameters is a complicated process. Computational techniques and control theory can be highly useful for tuning the PID constants.Antenna Control Servo System (ACSS) is used for continuous tracking of spacecraft during launch phase and on orbit phase. System performance can be easily evaluated and improved using simulations if mathematical model of system is known. Classical approach of physical modeling of system using lumped parameters is tedious and time consuming, moreover many parameters of system are not known for developing the model.The paper introduces a novel approach of developing mathematical model of ACSS from time domain test data using computational techniques. System Identification tool of MATLAB is used for estimating the model which is faster and flexible. Random data is generated in position control loop both in azimuth and elevation axis separately and taken for modeling.The antenna control system model is developed from test data of system logged with 0.2s sampling time. The transfer function is obtained between the servo amplifier output current and antenna angular velocity. Model is estimated as first order with zero order hold. Convergence criteria of the model are verified to ascertain model validity. Separate models are developed for azimuth motor1, azimuth motor2, elevation motor1 and elevation motor2 from respective motor data.4 simulation schematics are developed in SIMULINK, for each motor separately. Each one comprises of estimated model (current to velocity transfer function), servo amplifier model, PID controller logic with velocity limiter, acceleration limiter and anti-windup logic. Simulations are carried out in SIMULINK for studying closed loop behavior of each motor separately with known PID coefficients. The simulated results are then compared with actual antenna system response using same PID coefficients and performance is evaluated. Step response, ramp response and sine wave response of each motor is analyzed in simulation model. The paper concludes with the significance and future work scope of this research work.