Self-Commissioning Algorithm for Inverter Non-Linearity Compensation in Sensorless Induction Motor Drives

Sensorless field oriented control (SFOC) of induction motor (IM) drives estimate the mechanical position of the rotor and the field orientation by means of the motor voltage and current measures. In many cases, the motor voltages measures are substituted by the inverter commands, and the flux estimation is based on the stator model of the machine, i.e. the integration of back EMF voltages. With closed loop current control, the inverter nonlinearities (dead-time and on-state voltage drops) introduce a distortion in voltage estimation whose effect is particularly evident at low speed: lost of accuracy in flux estimation and reduced performance of the field oriented control. Most of the techniques for the compensation of the inverter nonlinear effects in the literature are based on off-line identification of the inverter model, with heavy offline post-processing. The goal of the paper is to present a simple and accurate method for on-line identification of the inverter parameters at the drive start-up. The method requires very few extra processing blocks, integrated into the main control code and it is based on the evaluation of the feedback signal of any closed loop flux observer. It is thus a general method, valid for all those drives where the flux observer is based on back-EMF integration, e.g. synchronous IPM drives. The self-commissioning algorithm is presented and tested here for a sensorless induction motor drive, controlled by a fixed-point industrial DSP. The feasibility and effectiveness of the proposed solution is demonstrated by experimental results.