Scholarlink

<p>This paper proposes a novel control approach integrating fuzzy logic with Long Short-Term Memory (LSTM) networks to enhance the performance of multi-segment robot systems comprising rigidly connected two-wheeled self-balancing units. Tailored for load-carrying applications, this method tackles synchronization, stability, and adaptability challenges under varying loads and terrains. By integrating Long Short-Term Memory (LSTM) for motor speed prediction with fuzzy inference, it leverages fuzzy logic’s strength in managing uncertainty and nonlinearity, delivering refined control outputs that outperform traditional deterministic approaches. Experimental results on a physical prototype demonstrate improved speed tracking, reduced skidding, and enhanced stability compared to non-predictive methods, validated across indoor floors and uneven asphalt. Beyond this application, the predictive fuzzy framework offers a generic solution for robotic systems requiring robust, adaptive control in dynamic environments, merging fuzzy logic’s uncertainty handling with LSTM’s predictive power. This hybrid approach advances motor control precision and efficiency, with potential implications for scalable robotic automation.</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

Fuzzy Control System for Multi-segment Load-Carrying Robot with Rigid Connections Integrated with LSTM

  • 期刊论文

  • 发表日期：

    2025年9月2日

查看全文
  • Chi Yu Chen,
  • Kuo Yi Chen

摘要

This paper proposes a novel control approach integrating fuzzy logic with Long Short-Term Memory (LSTM) networks to enhance the performance of multi-segment robot systems comprising rigidly connected two-wheeled self-balancing units. Tailored for load-carrying applications, this method tackles synchronization, stability, and adaptability challenges under varying loads and terrains. By integrating Long Short-Term Memory (LSTM) for motor speed prediction with fuzzy inference, it leverages fuzzy logic’s strength in managing uncertainty and nonlinearity, delivering refined control outputs that outperform traditional deterministic approaches. Experimental results on a physical prototype demonstrate improved speed tracking, reduced skidding, and enhanced stability compared to non-predictive methods, validated across indoor floors and uneven asphalt. Beyond this application, the predictive fuzzy framework offers a generic solution for robotic systems requiring robust, adaptive control in dynamic environments, merging fuzzy logic’s uncertainty handling with LSTM’s predictive power. This hybrid approach advances motor control precision and efficiency, with potential implications for scalable robotic automation.