Application of artificial neural network for adaptive speed control of PMSM drive with variable parameters

Abstract

PurposeThe aim of the research was to find out a method of adaptive speed control robust against variation of selected parameters of system like moment of inertia, time constant of torque control loop or torque coefficient of the motor.Design/methodology/approachThe main goal of the research was achieved due to application of artificial neural network (ANN), which was trained on line on the base of speed control error. The good results were gained by elaboration of enough fast and precise training algorithm and proper ANN structure.FindingsThe work shows a structure of artificial neural network (ANN), applied as adaptive speed controller, and presents an algorithm of ANN training. Some versions of this algorithm were analysed and verified by simulation and experimental tests.Research limitations/implicationsThe research should be continued to determine a final version of training algorithm and its influence on controller properties.Practical implicationsThe elaborated adaptive controller can be easily used by applying microprocessor system available now on the market. The proposed control solution is robust against parameters variation as well as their imprecise identification. The controller has ability of self‐tuning which can have great practical advantage.Social implicationsSocial implications are difficult to determine.Originality/valueThe paper presents a new solution of adaptive speed controller, which means a new ANN structure and new training algorithm.

Adaptive speed controller for servodrive based on artificial neural network trained online

“…As a results of ANN proper training the difference between both speed waveforms becomes after a few responses small. The transient process involved by a step change of load torque analysed in earlier author's works [6,7] shown interesting properties of the proposed neural controller. One can observe that the load change is accompanying by some changes of selected ANN weight coefficients.…”

“…The scheme of this stand is shown in Fig. 4 [7]. Control algorithms were implemented on signal processor ADSP 21060.…”

“…To reduce a calculation time of training algorithm some simplification was assumed by substituting sign of error gradient in (2) by sign of control error e , what led to equation: (3) is determined individually for each weight and changed in each step of training in agreement with formulas given in [7]:…”

Robust speed controller design for permanent magnet synchronous motor based on gain-scheduled control method via LMI approach

“…Many researches have been conducted to identify the best PMSM speed control method in the last years. Nonlinear control [5][6][7][8], adaptive control [9,10] and robust control [11,12], compared to other methods have been more popular due to their efficient performance. Of course, some other control methods are also worth mentioning such as sliding-mode control (SMC) [6,[13][14][15], neurofuzzy control (NFC) [16][17][18], and generalized predictive and sliding-mode control (GPSMC) [19,20].…”

Repetitive neurocontroller with disturbance feedforward path active in the pass-to-pass direction for a VSI inverter with an output LC filter

“…It has been demonstrated in numerous studies that an FFNN trained in the online mode using error backpropagation (BP) methods such as L-M or resilient BPs (RPROPs) can effectively control non-repetitive processes [22][23][24][25][26][27][28][29][30][31]. Surprisingly, the well-documented usefulness of online trained neurocontrollers for non-repetitive processes in adjustable speed drives and generators has not been followed by a similarly rich literature on neurocontrollers for repetitive processes.…”