Scholarlink

<p>This study presents a PRISMA-guided systematic review and quantitative synthesis of Sliding-Mode Observer (SMO)–based fault-tolerant control strategies published between 2020 and 2025. A structured database search identified 312 records, of which 63 peer-reviewed studies met predefined inclusion criteria following duplicate removal, multi-stage screening, and quality assessment. Only studies reporting quantitative validation of state estimation, fault diagnosis, or fault reconstruction performance were included. To enable cross-study comparability, performance metrics were normalized using Relative Root Mean Square Error (RRMSE), Settling Time Ratio (STR), Chattering Energy Index (CEI), and Robustness Margin (RM). Results indicate that advanced SMO variants—such as adaptive, super-twisting, higher-order, and hybrid AI-enhanced designs—achieve 30–70% lower estimation error and significantly faster fault recovery compared to classical observers. Multi-observer and redundancy-based architectures further improve resilience in sensor-degraded environments across motor drives, UAVs, robotics, and microgrids. Despite strong simulation performance, limited hardware-in-the-loop and experimental validation reveals a persistent simulation-to-reality gap. This review provides a transparent, benchmark-oriented roadmap for developing experimentally validated and resource-efficient SMO-based fault-tolerant control systems for safety–critical applications.</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

Review of sliding-mode observer and fault-tolerant control in sensor-driven platforms

  • 期刊论文

  • 发表日期：

    2026年5月30日

查看全文
  • Varuna Kumara,
  • Manjunatha Badiger

摘要

This study presents a PRISMA-guided systematic review and quantitative synthesis of Sliding-Mode Observer (SMO)–based fault-tolerant control strategies published between 2020 and 2025. A structured database search identified 312 records, of which 63 peer-reviewed studies met predefined inclusion criteria following duplicate removal, multi-stage screening, and quality assessment. Only studies reporting quantitative validation of state estimation, fault diagnosis, or fault reconstruction performance were included. To enable cross-study comparability, performance metrics were normalized using Relative Root Mean Square Error (RRMSE), Settling Time Ratio (STR), Chattering Energy Index (CEI), and Robustness Margin (RM). Results indicate that advanced SMO variants—such as adaptive, super-twisting, higher-order, and hybrid AI-enhanced designs—achieve 30–70% lower estimation error and significantly faster fault recovery compared to classical observers. Multi-observer and redundancy-based architectures further improve resilience in sensor-degraded environments across motor drives, UAVs, robotics, and microgrids. Despite strong simulation performance, limited hardware-in-the-loop and experimental validation reveals a persistent simulation-to-reality gap. This review provides a transparent, benchmark-oriented roadmap for developing experimentally validated and resource-efficient SMO-based fault-tolerant control systems for safety–critical applications.