Nonlinear state estimator design and implementation for speed control of a BLDC motor drive using square-root extended filter

Sensorless speed control of a brushless DC (BLDC) motor, basically a combined control-observer design, plays a crucial role in several industries, including electrical and aerospace applications. The key idea is to use only electrical (current) sensors and the remaining mechanical states (speed and rotor position) to be estimated using square-root extended (Figure presented.) filter (SRE (Figure presented.) F) for a BLDC motor. These estimated states are then fed-back to the controller. In most of the designs, it is assumed that the BLDC motor plant parameters are constant; however, in reality, their parameters vary due to various intrinsic and extrinsic conditions. Compared to the conventional extended Kalman filter or its square-root version, the proposed SRE (Figure presented.) F for the BLDC motor has an inherent tendency to deal with parameter uncertainties and hence improves the robustness of the overall system. The robustness of the proposed scheme is tested at various reference speeds for a closed-loop BLDC drive; further parametric uncertainties in stator resistance, detuning of inductance, and change in load are considered. The proposed control-observer strategy using SRE (Figure presented.) F is compared with its counterpart square-root extended Kalman filter (SREKF), and the experimental results validate the efficacy of SRE (Figure presented.) F for speed control of BLDC motor drive during the nominal conditions and in the presence of parametric uncertainties.