SCIENCE@home

A Comparative Study of Machine Learning Algorithms for Controlling Torque of Permanent Magnet Synchronous Motors through a Closed Loop System

Ahmad Bilal,Asad Waheed,M. H. Shah

Published 2019 in Intellect

ABSTRACT

Industries and manufacturers are being continuously challenged by the increasing demand of need for the development of fault and control predictive models that can help industries to have a better control over machines and their applications. The traditional classical models, are often based on physical and mathematical modeling of system and need high expertise and in-depth understanding of system in order to develop control solutions. However, with swift advancements in methods like data mining and machine learning, various techniques and algorithms have been developed which has revolutionized the fault diagnostics and control techniques for industries. A trained machine learning model, with an effective algorithm, can help to automatically adjust the current and voltage for motor power drivers, in order to compensate the change in torque due to temperature. The proposed paper evaluates the efficiency of different machine learning algorithms, in order to develop a torque control method to overcome the effect of change in temperature parameters of various part in the synchronous motor by using methods like linear regression, decision trees, support vector machine, and ensemble of tress. The best results for training of a machine learning based model was obtained by medium Gaussian and coarse tree algorithm in term of minimized error and optimized prediction speed respectively. Synchronous motors have emerged as one of the most popular choices in industries and house hold due to their efficiency and high power ratings. However, for electric motors, a rise in temperature can change the output torque which is not a desirable feature and results in degradation of output efficiency. The change in torque also causes the current values to fluctuate which leads to power wastage and can even harm the motor.

PUBLICATION RECORD

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-66 of 66 references · Page 1 of 1

CITED BY

Showing 1-6 of 6 citing papers · Page 1 of 1