A Study on the Rollover Warning and Active Roll Control Systems of Articulated Vehicles

• Main Authors: Tao Chou, 周道
• Other Authors: Tzyy-Wen Chu
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/39334587799399552834

博士 === 國防大學理工學院 === 國防科學研究所 === 102 === A rollover index (RI) combined with grey system theory, called a grey rollover index (GRI), is proposed to assess the rollover threat for articulated vehicles with a tractor-semitrailer combination. This index can predict future trends of vehicle dynamics based on current vehicle motion; thus, it is suitable for vehicle rollover detection. Two difficulties are encountered when applying the GRI for rollover detection. The first difficulty is effectively predicting the rollover threat of the vehicles, and the second difficulty is achieving a definite definition of the real rollover timing of a vehicle. The following methods are used to resolve these problems. First, a nonlinear mathematical model is constructed to accurately describe the vehicle dynamics of articulated vehicles. This model is combined with the GRI to predict rollover propensity. Finally, TruckSim® software is used to determine the real rollover timing and facilitate the accurate supply of information to the rollover detection system through the GRI. This index is used to verify the simulation based on the common manoeuvres that cause rollover accidents to reduce the occurrence of false signals and effectively increase the efficiency of the rollover detection system. Furthermore, a rollover control system with integration of active steering and differential braking was proposed to prevent rollover propensity, in which a closed-loop model following control structure was adopted so that the output response of the controlled vehicle may follow that of an ideal virtual vehicle model. In order to achieve this objective, a multi-input multi-output (MIMO) eigenstructure assignment control strategy was proposed. The main concept is to search appropriate eigenvelues and eigenvectors with respect to the time-domain specifiction in combination with mixed-sensitivity H∞ control design considering the requirement in the frequency-domain. The purpose of the control strategy is to attain the gain matrix of the closed-loop system so that the system can satisfy both time- and frequency-domain specifications. Simulation results show that the proposed control design can recover a vehicle under critical rollover condition to a stable condition, which proves the appropriateness of the control design.