皮带轮系统之非线性动态特性之判别.pdfVIP

桃園創新學報 第三十二期 皮帶輪系統之非線性動態特性之判別 ────────────────────────────────────────── 皮帶輪系統之非線性動態特性之判別 黃元瑞 * 、童景賢 桃園創新技術學院機械系 摘 要 本文以柴比雪夫遞迴式小波類神經模糊網路模式(Chebyshev Recurrent Wavelet Neuro-Fuzzy network, CRWNF) [1]為平皮帶輪系統之判別模型，以探討 平皮帶輪系統之非線性動態特性。柴比雪夫遞迴式小波類神經模糊網路模式係 由柴比雪夫遞迴式小波類神經網路與模糊理論組合而成。其間針對平皮帶輪非 線性動態系統，運用柴比雪夫遞迴式小波類神經網路具學習之能力以達到模型 之最佳化效果。將其效能與適應性網路模糊推理 (ANFIS)之效能加以比較，經 比較模擬結果，驗證柴比雪夫類神經模糊網路對具有不確定性之平面皮帶輪動 態系統有較優之鑑別模型。 關鍵字：遞迴式小波類神經網路、模糊理論 Email : hyr@.tw Tel : 03-4361070-6302 Fax : 03-4384670 137 桃園創新學報 第三十二期 皮帶輪系統之非線性動態特性之判別 ────────────────────────────────────────── Modeling Belt-Drive System with Nonlinear Dynamic Characteristics Yuan-Ruey Huang*, Jingsyan Torng Department of Mechanical Engineering, Taoyuan Innovation Institute of Technology, Chungli , Taiwan, R.O.C. Abstract In this study, the modeling of belt-drive system based on Chebyshev Recurrent Wavelet Neuro-Fuzzy Network (CRWNF) [1] is proposed. CRWNF is composed of Recurrent Wavelet Neural networks and Fuzzy theory. Using Recurrent Wavelet Neural networks with dynamical mapping mechanism as inferring mechanisms can identify nonlinear dynamical system effectively. To verify the performance of the proposed CRWNF for modeling belt-drive system, that of the adaptive neural fuzzy inference system is also presented. According to the simulation results CRWNF can identify nonlinear dynamical system more effectively. Keywords: recurrent wavelet neural network 、fuzzy theory Email : hyr@.