Sensorless control for limp-home mode of EV applications

Abstract

Over the past decade research into electric vehicles’ (EVs) safety, reliability and availability has become a hot topic and has attracted a lot of attention in the literature. Inevitably these key areas require further study and improvement. One of the challenges EVs face is speed/position sensor failure due to vibration and harsh environments. Wires connecting the sensor to the motor controller have a high likelihood of breakage. Loss of signals from the speed/position sensor will bring the EV to halt mode. Speed sensor failure at a busy roundabout or on a high speed motorway can have serious consequences and put the lives of drivers and passengers in great danger. This thesis aims to tackle the aforementioned issues by proposing several novel sensorless schemes based on Model Reference Adaptive Systems (MRAS) suitable for limp-home mode of EV applications. The estimated speed from these schemes is used for the rotor flux position estimation. The estimated rotor flux position is employed for sensorless torque-controlled drive (TCD) based on indirect rotor field oriented control (IRFOC). The capabilities of the proposed schemes have been evaluated and compared to the conventional back-Electromotive Force MRAS (back-EMF MRAS) scheme using simulation environment and a test bench setup. The new schemes have also been tested on electric golf buggies. The results presented for the proposed schemes show that utilising these schemes provide a reliable and smooth sensorless operation during vehicle test-drive starting from standstill and over a wide range of speeds, including the field weakening region. Employing these new schemes for sensorless TCD in limp-home mode of EV applications increases safety, reliability and availability of EVs.