Scholarlink

<p>The structural systems inherit challenges such as highly uncertain dynamics, triggering of unmodeled dynamics in high-frequency excitation, and various natural disturbances. Traditional active mass drivers (AMD) that depend on predetermined structure’s characteristics lose their reliability in these conditions. Considering these challenges, in this paper, a hybrid adaptive non-singleton Type-3 (T3) fuzzy logic controller (HANT3-FLC) is presented. The suggested controller has three parts: First, an online-learned adaptive non-singleton T3 fuzzy logic system (FLS)-based fractional-order proportional–integral–derivative (PID) controller (ANT3-FPID) is designed. The designed learning scheme uniquely accelerates adaptation by simultaneously tuning fuzzy rules and dynamically adjusting membership functions (MFs). Furthermore, the developed non-singleton fuzzification aims to effectively minimize the detrimental effects of sensor noise, which is a critical concern in practical structural systems. The second part is designed to deal with unmodeled dynamics. The displacement error of ANT3-FPID is modeled using a first-order dynamic FLS, enabling the design of a complementary adaptive FLC (AT3-FLC). Finally, in the last part, an adaptive parallel supplementary compensator is designed based on stability analysis through the Lyapunov theory. The compensator is developed through online estimation of the maximum approximation error, guaranteeing robust performance. The efficacy and robustness of the proposed controller are validated through comprehensive simulations and real-world implementation on a structural control system.</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

Deep online learning type-3 fuzzy structural control strategy for active vibration suppression: real-world validation

  • 期刊论文

  • 发表日期：

    2025年11月14日

查看全文
  • Chunwei Zhang,
  • Shuhui Pei,
  • Ardashir Mohammadzadeh

摘要

The structural systems inherit challenges such as highly uncertain dynamics, triggering of unmodeled dynamics in high-frequency excitation, and various natural disturbances. Traditional active mass drivers (AMD) that depend on predetermined structure’s characteristics lose their reliability in these conditions. Considering these challenges, in this paper, a hybrid adaptive non-singleton Type-3 (T3) fuzzy logic controller (HANT3-FLC) is presented. The suggested controller has three parts: First, an online-learned adaptive non-singleton T3 fuzzy logic system (FLS)-based fractional-order proportional–integral–derivative (PID) controller (ANT3-FPID) is designed. The designed learning scheme uniquely accelerates adaptation by simultaneously tuning fuzzy rules and dynamically adjusting membership functions (MFs). Furthermore, the developed non-singleton fuzzification aims to effectively minimize the detrimental effects of sensor noise, which is a critical concern in practical structural systems. The second part is designed to deal with unmodeled dynamics. The displacement error of ANT3-FPID is modeled using a first-order dynamic FLS, enabling the design of a complementary adaptive FLC (AT3-FLC). Finally, in the last part, an adaptive parallel supplementary compensator is designed based on stability analysis through the Lyapunov theory. The compensator is developed through online estimation of the maximum approximation error, guaranteeing robust performance. The efficacy and robustness of the proposed controller are validated through comprehensive simulations and real-world implementation on a structural control system.