Sensorless speed estimation for long term flywheel energy storage system in standby mode.

Abstract

A novel technique for sensorless speed estimation is presented in this thesis for squirrel cage induction machine (SCIM) drived long-term flywheel energy storage system (FESS) in standby mode. The SCIM model for long-term large-capacity FESS is presented. Based on dynamic model , a hybrid rotor flux observer and speed observer are derived. The hybrid rotor flux observer takes advantages of both the current model and voltage model flux observers by seamlessly incorporating these two models together for a better flux estimation performance even at low speed range. The fundamental speed observer is derived from the dynamic model for speed estimation with a fast response time for a tradeoff of the adaptive capabilities. In order to observe the speed in standby mode, a modified field-oriented control (FOC) scheme is presented. The hybrid flux observer and speed observer are tested in association with the modified FOC. The proposed control technique adopts approaches in an effort to minimize the impact generated by the excitation and speed estimation process to the FESS. Simulation and experiments are conducted to verify the feasibility of the proposed speed estimation at the standby mode. It is also observed that a step change of excitation current has a significant impact to the existing FESS. A ramp control for excitation current is added to avoid the possible oscillation of the estimated speed and the disturbance to the FESS. The speed estimation settling time is optimized based on the experiment and simulation.