查詢結果分析
相關文獻
- Neural Network Error Accommodation with Fuzzy Logic Elements in Robot Time Optimal Path Tracking
- Vibration Isolation by Using Magnetostrictive Actuator
- 機器故障偵測利用訊號處理與式樣辨識技術
- 智慧型水庫即時操作控制系統
- Applying Fuzzy Logic in the Modified Single-Layer Perceptron Image Segmentation Network
- Intelligent Control Theory in Guidance and Control System Design: An Overview
- 氣體輔助射出成型之智慧型專家系統
- 利用類神經-模糊理論評定契合程度--以管理人員甄選為例
- 利用類神經網路與模糊邏輯預測山崩發生之可能性
- An Adaptive Neural Network Impedance Learning Control Method for Robotic Systems
頁籤選單縮合
題名 | Neural Network Error Accommodation with Fuzzy Logic Elements in Robot Time Optimal Path Tracking=以神經網路與模糊邏輯增進機器人時間最佳化路徑追蹤之容錯度 |
---|---|
作者 | Veryha,Yauheni B.; |
期刊 | Proceedings of the National Science Council : Part A, Physical Science and Engineering |
出版日期 | 20011100 |
卷期 | 25:6 2001.11[民90.11] |
頁次 | 頁367-376 |
分類號 | 448.94 |
語文 | eng |
關鍵詞 | 神經網路; 模糊邏輯; 機器人; 偵錯; 時間最佳化控制; Time optimal control; Adaptive robotic system; Autonomous robotic system; Error detection; Neural network; Fuzzy logic element; |
中文摘要 | 錯誤的偵測、診斷、與承受對自主性機器人系統是相當重要的,系統的錯誤常常 發生於系統關鍵參數或動態產生變化時,也因此會惡化系統的表現,此種現象對於機器人時 間最佳化控制的影響更為顯著,因為在此情形下,系統參數常是在關鍵值,而即使是很小的 誤差也會導致相當大的錯誤。此篇論文乃利用神經網路以及模糊邏輯來處理在計算力矩控制 ( computed torque control )下機器人錯誤診斷的問題, 所提出的學習架構包含一神經 網路作為即時逼近器( on-line approximator )以及在控制部分的模糊邏輯單元來進行機 器人系統的錯誤診斷與容錯。利用神經網路進行的逼近提供了錯誤特徵的模型,可用來偵測 與消除機器人功能上的錯誤。模擬結果印證了所提出的學習機制足以處理在時間最佳化控制 下兩軸機器人的錯誤偵測與容錯。 |
英文摘要 | Error detection, diagnosis and accommodation play key roles in the operation of autonomous robotic systems. System fault, which typically result in changes of critical system parameters or system dynamics, may lead to degradation in performance. This fact is especially important for time optimal robot control when the system parameters reach their critical values and even small changes can lead to accuracy degradation. This paper investigates the problem of error diagnosis in robotic manipulators under computed torque control using neural network and fuzzy logic elements. A learning architecture with neural networks serving as on-line approximators with fuzzy logic elements in the control unit is used for the diagnosis of robotic system errors and error accommodation. Approximation using neural networks provides a model of the error characteristics that can be used for the detection and elimination of errors in robot functioning. Simulation results illustrate the ability of the neural network based error diagnosis method with the fuzzy elements described in this paper to detect and accommodate errors in a two-link robotic manipulator under time optimal control. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。