Abstract
The accuracy estimation of induction motors’ efficiency is beneficial and crucial in the industry for energy savings. The requirement for in situ machine efficiency estimation techniques is increasing in importance because it is the precondition to making the energy-saving scheme. Currently, the torque and speed identification method is widely applied in online efficiency estimation for motor systems. However, the higher precision parameters, such as stator resistance Rs and equivalent resistance of iron losses Rfe, which are the key to the efficiency estimation process with the air gap torque method, are of cardinal importance in the estimation process. Moreover, the computation burden is also a severe problem for the real-time data process. To solve these problems, as for the torque and speed-identification-based efficiency estimation method, this paper presents a lower time burden method based on Quantum Particle Swarm Optimization-Trust Region Algorithm (QPSO-TRA). The contribution of the proposed method is to transform the disadvantages of former algorithms to develop a reliable hybrid algorithm to identify the crucial parameters, namely, Rs and Rfe. Sensorless speed identification based on the rotor slot harmonic frequency (RSHF) method is adopted for speed determination. This hybrid algorithm reduces the computation burden by about 1/3 compared to the classical genetic algorithm (GA). The proposed method was validated by testing a 5.5 kW motor in the laboratory and a 10 MW induction motor in the field.