A novel induction motor IRFOC with on-line parameter estimation taking into account thermal and deep-bar effects

Sensorless drive control has been widely studied in recent years due to numerous advantages regarding potential failures of position sensors, especially in applications such as automotive or aerospace. Among vector control drives, indirect rotor-flux-oriented control (IRFOC) is one of the most popular and tested options, whose main weaknesses are the dependence on motor parameters and the low-speed behaviour. Parameter values change due to several reasons, such as thermal, saturation or deep-bar effect. Saturation can be avoided by controlling the flux, and the other effects are taken into account. A thermal model that focuses on obtaining stator and rotor representative temperatures is selected. Concerning the deep-bar effect, an approximate solution to a previously developed one is presented which shows a good compromise between accuracy and simplicity. In order to obtain a better behaviour in the low speed region, a speed estimator is built so that not only the dq electrical equations are considered, but also the mechanical one. It allows considering the static friction that is not properly monitored by the current measurements when their peak value is low. The proposed design includes motor overcurrent protection and its validity and usefulness is experimentally tested on a TMS320C31 DSP from the Simulink/Matlab environment.