Torque control of electric motors without using torque sensor

Accurate torque control of a brushless motor requires the motor´s torque characteristics, which can be captured by a periodic function in motor angle. This paper presents a direct adaptive controller for torque control of brushless motors, which estimates the Fourier coefficients of the periodic function based on the measurements of motor phase voltage and angle. It will be analytically shown that the proposed adaptive controller achieves torque tracking regardless of the trajectories of input signals. Moreover, the adaptive controller does not rely on the modeling of the mechanical load, and that makes control implementation simple and modular. Experimental results obtained from the McGill/MIT motor have demonstrated that motor torque converges to the command torque.