Real-Time Backstepping Control Based MRAS Speed Estimation of an Induction Machine

Abstract

This work presents a real-time evaluation of a sensorless MRAS-backstepping control applied to an induction process (IM). To run the system in safe mode, a backstepping method is first implemented. It guarantees decoupled torque and flux management of the IM without the need for an additional regulator by removing the proportional/integral controller (PI) and decreasing the number of gains that are highly reliant on the temperature-dependent IM parameters. The backstepping control is designed to attain the required speed and flux, and it is ideally suited for regulating an uncertain system that may be affected by external disturbances. Its formulation is based on the application of Lyapunov functions to assure the overall stability of the system. The second section presents a model reference adaptive system (MRAS) to estimate rotor speed and eliminate the sensor, the system's weakest link. Utilizing dSPACE 1104 cards, the suggested control procedure is evaluated. The achieved experimental results indicate that the entire system runs smoothly, with the proposed MRAS-backstepping providing high performance, the system successfully tracking the speed profile under tolerable load, and the system maintaining its dynamic behavior at different speed levels.