Position-Sensorless Rotor Flux-Oriented Control Of Induction Machines

Applying position-sensorless rotor flux-oriented control(RFOC)to induction machine(IM)drives is able to reduce hardware cost,improve reliability of drives and realize highperformance closed-loop speed regulation.Thus,position-sensorless RFOC has received various of concerns from researchers and engineers for a long period.Simultaneous rotor flux observation and speed estimation based on the IM fundamental model are general and widespread,but they require carefully designed flux observer and controllers,and it is quite difficult to guarantee their stability at low speeds.Aiming at these issues,position-sensorless IM RFOC based on adaptive reduced-order observer is investigated.First,the IM fundamental model is derived,which is used to analyze the principle of RFOC.The implementation of RFOC is also illustrated.Next,the association between speed estimation and rotor flux observation is explored,and a speed estimation method based on model adaptive reference control(MRAC)is derived.Since the voltage model(VM)can not calculate rotor flux accurately at low speeds,error of rotor flux observation is introduced as a feedback for the VM,which leads to a direct flux-oriented(DFO)adaptive reduced-order observer.Small signal analysis is performed to tune the observer gain reasonably,it shows that a flux observer independent of speed is helpful for analysis and complex gain is essential for local stability of all operation points except where the synchronous angular velocity is zero.The accuracy of observer input,controller feedback and electric and magnetic parameters of the IM also affects performance of closed-loop IM drives.Nonlinearities of the inverter produces errors between stator voltage and reference voltage,so it is modeled as a simple form and a compensation method based on the feedback of approximate voltage model is proposed.The deterioration of speed in closed-loop drives brought by inevitable sampling error of stator currents is illustrated.Moreover,parameter sensitivity of flux observer is analyzed through frequency response,and it proves that a flux observer that is dominated by current model(CM)at low speeds,whereas by VM at high speeds is preferable for reduce parameter sensitivity.The experiment rig is built using a PE-Expert3 digital controller.Indirect flux-oriented(IFO)control with CM and position sensor,compensation of inverter nonlinearities and position-sensorless DFO control with reduced-order observer are tested by experiments.Experimental results indicate that the theoretical analysis is valid and the proposed position-sensorless RFOC system is capable of keeping stable and good dynamic and steady state performance in a large operation range.