Scholarlink

<p>This paper presents a novel Adaptive Neuro-Fuzzy Sliding Mode Observer (ANFIS-SMO) designed for sensorless Field-Oriented Control (FOC) of Permanent Magnet Synchronous Motor (PMSM) drives. The proposed approach integrates an Adaptive Neuro-Fuzzy Inference System (ANFIS) with a Sliding Mode Observer (SMO) to enhance rotor position and speed estimation accuracy, particularly under parameter uncertainties and dynamic load conditions. Unlike conventional SMO, where chattering introduces current distortions that manifest as torque ripple, the ANFIS-SMO dynamically adjusts observer gains through fuzzy inference, thereby suppressing chattering and ensuring smoother current and torque responses. As a result, the method simultaneously achieves torque ripple mitigation and high-precision rotor angle estimation. Comparative analysis with traditional sensorless control methods, including Phase-Locked Loop (PLL) and conventional SMO, demonstrates that ANFIS-SMO significantly reduces torque ripple (up to 40%), improves speed tracking accuracy (by 25%), and maintains rotor position estimation errors below 2%. Furthermore, it provides faster settling times and reduced overshoot compared to existing techniques, highlighting its potential for high-performance PMSM applications in electric vehicles and industrial drives.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Torque Ripple Mitigation and Rotor Position Error Reduction in PMSM Sensorless Drives via ANFIS-SMO

  • 期刊论文

  • 发表日期：

    2026年1月20日

查看全文
  • Brijendra Sangar,
  • Shilpa Ranjan,
  • Madhusudan Singh,
  • Mini Sreejeth

摘要

This paper presents a novel Adaptive Neuro-Fuzzy Sliding Mode Observer (ANFIS-SMO) designed for sensorless Field-Oriented Control (FOC) of Permanent Magnet Synchronous Motor (PMSM) drives. The proposed approach integrates an Adaptive Neuro-Fuzzy Inference System (ANFIS) with a Sliding Mode Observer (SMO) to enhance rotor position and speed estimation accuracy, particularly under parameter uncertainties and dynamic load conditions. Unlike conventional SMO, where chattering introduces current distortions that manifest as torque ripple, the ANFIS-SMO dynamically adjusts observer gains through fuzzy inference, thereby suppressing chattering and ensuring smoother current and torque responses. As a result, the method simultaneously achieves torque ripple mitigation and high-precision rotor angle estimation. Comparative analysis with traditional sensorless control methods, including Phase-Locked Loop (PLL) and conventional SMO, demonstrates that ANFIS-SMO significantly reduces torque ripple (up to 40%), improves speed tracking accuracy (by 25%), and maintains rotor position estimation errors below 2%. Furthermore, it provides faster settling times and reduced overshoot compared to existing techniques, highlighting its potential for high-performance PMSM applications in electric vehicles and industrial drives.