Motors Using a Deep Learning-Based Torque Control with Torque Ripple Reduction under Nonlinear Magnetic Conditions

Notice

This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.

Year : 2026 | Volume : 04 | 02 | Page :
By

Ankita Madhukar Gaikwad,

Bhushan Kadam,

  1. Student, Department of Electrical Engineering, Sanjivani College of Engineering, Kopargaon, Maharashtra, India
  2. Assistant Professor, Department of Electrical Engineering, Sanjivani College of Engineering, Kopargaon, Maharashtra, India

Abstract

This research discusses a deep learning strategy for torque management to minimize the effect of torque ripple in a nonlinear electric motor. Nonlinear electric motor losses may include: magnetic saturation, harmonic flux losses and inverter losses. In many cases when the system parameters deviate and/or instability issues occur, the traditional method with a model-based approach or PI control may encounter challenges. In this case, the authors proposed a hybrid approach based on deep neural networks that have convolutional and recurrent layers to learn in real time and correct for torque drift as it occurs. To precisely forecast system behavior under various operating situations, the suggested architecture combines sophisticated feature extraction and temporal learning capabilities. Without the need for regular human adjustment, the deep learning model adjusts to nonlinear dynamics, parameter uncertainties, and external disturbances by continually evaluating motor operating data. This capacity to adapt improves the control system’s durability and dependability while preserving steady-state and transient performance. High-performance electric drive applications can benefit from the hybrid architecture’s quick convergence and effective processing of complicated information. Verification of the implementation of the proposed technique in MATLAB/Simulink simulated scenarios shows that the control scheme could maintain lower torque ripple, reduced vibrations of the motor, and higher efficiency than a standard model-based control approach took. This approach to intelligent control represents a new feature due to the control of an adaptive self-learning electric drive targeted for industrial and automotive usage.

Keywords: Deep learning, torque ripple, nonlinear magnetic condition, motor control, neural network, adaptive control.

How to cite this article: Ankita Madhukar Gaikwad, Bhushan Kadam. Motors Using a Deep Learning-Based Torque Control with Torque Ripple Reduction under Nonlinear Magnetic Conditions. International Journal of Advanced Control and System Engineering. 2026; 04(02):-.
How to cite this URL: Ankita Madhukar Gaikwad, Bhushan Kadam. Motors Using a Deep Learning-Based Torque Control with Torque Ripple Reduction under Nonlinear Magnetic Conditions. International Journal of Advanced Control and System Engineering. 2026; 04(02):-. Available from: https://journals.stmjournals.com/ijacse/article=2026/view=253341

References

  1. De Kock HW, Rix AJ, Kamper MJ. Optimal torque control of interior permanent magnet synchronous machines in the full speed range. In2008 18th International Conference on Electrical Machines 2008 Sep 6 (pp. 1-6). IEEE.
  2. Qu J, Zhang P, Jatskevich J, Zhang C. Torque-ripple reduction of permanent magnet synchronous machine drives based on novel speed harmonic control at low-speed operation. IEEE Transactions on Industrial Electronics. 2022 Dec 20;70(8):7683-94.
  3. Kieu NV, Bao NH, Bach TQ. A Model Advanced Predictive Current Control Strategy for PMSM Drives with Torque Ripple and Current Harmonic Reduction. Journal of Robotics and Control (JRC). 2025 Oct 21;6(5):2444-56.
  4. Perumal S. Model Predictive Torque Control of Permanent Magnet Synchronous Motor Drives Using Deep Reinforcement Learning for High-Dynamic Industrial Applications. National Journal of Electric Drives and Control Systems. 2026 Mar 2:1-9.
  5. Hidaka Y, Sato T, Sugahara K. Data-Driven Topology Optimization With Warm Starting CMA-ES for Constraint Optimization in Rotating Machines. IEEE Access. 2026 Feb 18.
  6. M. El-Sayed, A. Tahan, and K. Khorasani, “Deep reinforcement learning for optimal torque control of permanent magnet motors,” IEEE Transactions on Neural Networks and Learning Systems, vol. 33, no. 9, pp. 4558–4571, Sept. 2022.
  7. Liu T, Husain I, Elbuluk M. Torque ripple minimization with on-line parameter estimation using neural networks in permanent magnet synchronous motors. InConference Record of 1998 IEEE Industry Applications Conference. Thirty-Third IAS Annual Meeting (Cat. No. 98CH36242) 1998 Oct 12 (Vol. 1, pp. 35-40). IEEE.
  8. Chen Y, Xiao Y, Ji C, Su B. An online torque prediction method of PMSM using PINNs. IEEE Access. 2025 May 9;13:196796-808.
  9. Sangar B, Singh M, Sreejeth M. An improved ANFIS model predictive current control approach for minimizing torque and current ripples in PMSM-driven electric vehicle. Electrical Engineering. 2024 Oct;106(5):5897-907.
  10. Velarde-Gomez S, Giraldo E. Real-time identification and nonlinear control of a permanent-magnet synchronous motor based on a physics-informed neural network and exact feedback linearization. Information. 2024 Sep 19;15(9):577.
Ahead of Print Subscription Review Article
Volume 04
02
Received 06/08/2026
Accepted 13/08/2026
Published 25/08/2026
Publication Time 19 Days

Login

My IP

PlumX Metrics