Torque sensorless control for an electric power assisted bicycle with instantaneous pedaling torque estimation

A typical electric power-assisted bicycle is equipped with an expensive torque sensor to measure a rider's pedaling torque. In this paper, we propose a method to estimate the pedaling torque from Hall effect sensors mounted on the motor, reducing the cost of electric power-assisted bicycles. It is necessary to separate pedaling torque and load torque from disturbance torque, which is estimated by a disturbance observer. We employ the Fourier series expansion to extract the periodic components of disturbance torque associated with the crank angle, which corresponds to the average pedaling torque. A conditional recursive least squares method is then used to estimate the parameters of the load torque model. Finally, the instantaneous pedaling torque is obtained basedon the disturbance observerand the estimated load torque. We confirmed the validity of the proposed method by simulations and experiments. Our method can also be applied to a normal bicycle by replacing its front wheel with a motorized wheel. © 2017 The Institute of Electrical Engineers of Japan.