Title :
Gradient descent based optimization for position and speed estimation of permanent magnet synchronous motor at low speed
Author :
Khan, A.A. ; Mohammed, O.A.
Author_Institution :
Dept. of Electr. & Comput. Eng., Florida Int. Univ., Miami, FL, USA
Abstract :
The optimization of low speed sensorless algorithm parameters is considered to be difficult due to large number of parameters involved. In this paper, a gradient descent based optimization is proposed to optimize the parameters of low speed sensorless algorithm to improve speed and position estimation performance. The finite element based phase variable model of permanent magnet synchronous motor is used for simulation of low speed sensorless algorithm. The proposed optimization method improves both dynamic and steady state response characteristic of estimated position and estimated speed.
Keywords :
finite element analysis; gradient methods; permanent magnet motors; synchronous motors; dynamic state response characteristic; finite element based phase variable model; gradient descent based optimization; low speed sensorless algorithm; permanent magnet synchronous motor; position estimation; speed estimation; steady state response characteristic; Band pass filters; Data mining; Frequency; Mathematical model; Permanent magnet motors; Sensorless control; Signal processing algorithms; Synchronous motors; Torque; Voltage; Optimization; PM motor; Sensorless Control;
Conference_Titel :
Power & Energy Society General Meeting, 2009. PES '09. IEEE
Conference_Location :
Calgary, AB
Print_ISBN :
978-1-4244-4241-6
DOI :
10.1109/PES.2009.5275903
