Research On Active Anti-rollover System Based On Integrated Chassis Control For SUV

In recent years, vehicle rollovers happen frequently and bring huge losses to the livesand property of the occupants. Compared with other vehicles, Sport Utility Vehicle (SUV)has higher center of gravity. Therefore, the wide application of SUV leads to the increase ofthe vehicle rollover accident. The lateral stability of vehicle has become an important issuein traffic safety field. In order to improve the vehicle driving safety and reduce the casualtiesand property losses caused by rollover accidents, the engineers have carried on furtherresearches on rollover prevention technology.In this paper, partly sponsored by National Science Foundation of China for YoungScholars (51105169;51205156) and Jilin Provincial Research Foundation for YoungScholars (201101028), the study for rollover prevention technology is carried out. Firstly,according to the analysis and simulation of rollover warning algorithm based on TTR(Timeto Rollover), a multi-parameter modified NN(Neural Network)-TTR rollover warningalgorithm is presented. Then, the overall framework of the active anti-rollover controlsystem based on integrated chassis control is designed using the model predictive controlmethod. The active steering and differential braking system are integrated to imply thetracking of the ideal yaw rate and sideslip angle. Thirdly, the offline simulation platform andhardware in the loop platform are designed and the above-mentioned algorithm is verified.The main contents of the dissertation are as follows:1. Rollover warning algorithmRollover warning is a key of the rollover prevention technology and its accuracy andreal-time performance are the critical factors to influence the control quality. So the threedegrees of freedom vehicle model is established. In order to make the algorithm moreuniversal, the lateral load transfer rate is selected as the rollover index to judge the rollover risk. Then, TTR-based rollover warning algorithm is studied. To make the warning moreaccurate and ideal, bionic intelligent neural network and vehicle state information areintroduced and the modified TTR is used to trigger the active anti-rollover control system.2. Active anti-roll control algorithmResearch on active anti-rollover control algorithm is conducted and the overallframework is presented. In this paper, the model predictive control method which uses thetwo degrees of freedom vehicle model as a prediction model and the yaw rate and sideslipangle as the feedback control targets is used to design the integrated chassis controller. Thenthe additional front wheel angle and the active yaw moment are achieved. Once the NN-TTRexceeds the reference value, the controller will be triggered and active steering anddifferential braking will be applied to the vehicle to avoid rollover accident.3. Off-line simulation of rollover warning and active anti-roll control algorithmThe rollover warning algorithm and active anti-rollover control algorithm are validatedusing Matlab/Simulink and CarSim co-simulation. The simulation results show that theTTR-based rollover warning algorithm can alarm the risk of rollover, but it’s not accurateenough and the NN-TTR algorithm can significantly improve warning accuracy. Also, theactive anti-rollover control system can integrate active steering and differential brakingeffectively, and improve the handling and stability performance.4. Hardware-in-the-loop test of active anti-roll control algorithmThe hardware in the loop test platform based on dSPACE is designed. HIL tests basedon algorithm prototype are performed. The typical conditions chosen are step steering andfishhook steering. The test results indicate that the controller designed in this paper hasexcellent control effects which can effectively improve the vehicle roll stability.