Mehdi Shafiei

Position sensorless and adaptive speed design for controlling brushless DC motor drives

This paper proposes a method for direct torque control of Brushless DC (BLDC) motors. Evaluating the trapezium of back-EMF is needed, and is done via a sliding mode observer employing just one measurement of stator current. The effect of the proposed estimation algorithm is reducing the impact of switching noise and consequently eliminating the required filter. Furthermore, to overcome the uncertainties related to BLDC motors, Recursive Least Square (RLS) is regarded as a real-time estimator of inertia and viscous damping coefficients of the BLDC motor. By substituting the estimated load torque in mechanical dynamic equations, the rotor speed can be calculated. Also, to increase the robustness and decrease the rise time of the system, Modified Model Reference Adaptive System (MMRAS) is applied in order to design a new speed controller. Simulation results confirm the validity of this recommended method.

Regenerative braking of electric vehicle using a modified direct torque control and adaptive control theory

This paper represents a novel regenerative braking approach for electric vehicles. The proposed method solves the short-range problem which is related to the battery discharge. The direct torque control switching algorithm is modified to recover electrical energy from electric vehicle, driven by brushless direct-current motor, without using the additional power converter or the other electrical energy storage devices. During regenerative braking process, a switching pattern is applied to the inverter which is different from the normal operation due to the special arrangement of voltage vectors. The new switching pattern is considered to convert mechanical energy into electrical energy. State of charge of the battery is used as a performance indicator of the proposed method. Simultaneously, a model reference adaptive system is designed to tune the systems parameters. Several simulations are conducted to validate the performance and effectiveness of the proposed methods. The results show the high capability of designed methods.