Servo System Control Using SPC and ASPC

This paper designs a Subspace Predictive Controller (SPC) and an Adaptive SPC (ASPC) to control the angular position of a servomechanism. The results are obtained experimentally using Lab Volt servo kit. The performance of proposed controllers is compared to standard Model Predictive Control (MPC). As SPC and ASPC are based on nonparametric Subspace Identification (SID), they avoid the need to design an explicit state observer and hence they ease the process of tuning the controller. The design ASPC aims to compensate the mismatch between the real system and the nonparametric model by updating the controller with new closed loop input-output measurement.