Precision Position Control of PMSM using Neural Observer and Parameter Compensator

摘要: This paper presents neural load torque compensation method which is composed of a deadbeat observer and gains by parameter estimator. As result, the response PMSM (permanent magnet synchronous motor) obtains better precision position control. To reduce noise effect, post-filter implemented MA (moving average) process. The compensator with an RLSM (recursive least square method) estimator adopted to increase performance main controller. combined high resolve problems. network trained in online phases it feed forward recall error back-propagation training. During normal operation, input-output sampled weighting value multi-times at each sample period accommodate possible variations parameters or torque. proposed control system has robust precise against variation. Stability usefulness are verified computer simulation experiment.