Identification of Hysteresis Dynamics Using Duffing-Like Model

• Main Authors: Chien. Yuan Che, 簡遠哲
• Other Authors: Tu. Jia Ying
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/cjxb69

碩士 === 國立清華大學 === 動力機械工程學系 === 104 === The purpose of this study is to identify the parameters of hysteresis dynamics using the duffing-like model. In the mechanical and material systems, hysteresis curves often behave in the input-output graph, such as magnetorheological(MR) damper and ferromagnetic material components. However, the factors that cause hysteresis in these systems are different. In the literature, Bingham model and Bouc-Wen model are used to identify hysteresis curve, these models have complex mathematical equations including piecewise, nondeterministic, and discontinuous functions that make system analysis difficult. Therefore, this study uses the Duffing-like model to identify hysteresis curve, and provides the new standard operating procedures of hysteresis system identification. The Duffing-like model was developed based on the Duffing equation, showing a scenario of continuous nonlinear ordinary differential equation. Then, this study applies the Lyapunov theorem to show the stability analysis of Duffing-like model and obtain the operating conditions related to the Duffing-like equation parameters. Next, Duffing-like model is built in Simulink for simulation. The simulation results show the relationship between the parameters of Duffing-like model and hysteresis curve. The series of simulation results provide the information about establishing the procedure to fit the hysteresis curve. Finally, this study identifies hysteresis curve with the three systems, magnetoresistive sensor, ferromagnetic material, and magnetorheological damper with the provided procedure and using Duffing-like model. Comparing identification results to the experimental results, and discusses the practical applicability and reliability of hysteresis dynamic identification using Duffing-like model.