• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2021, Vol. 57 ›› Issue (22): 277-283.doi: 10.3901/JME.2021.22.277

Previous Articles     Next Articles

Model Experimental Study on Flow Field and Force of Wet Clutch

LI Shenlong1,2, WU Wei2,3, HU Jibin2,3, YUAN Shihua2,3, WEI Chunhui3, HU Chenhui3   

  1. 1. China North Vehicle Research Institute, Beijing 100072;
    2. National Key Laboratory of Vehicular Transmission, Beijing Institute of Technology, Beijing 100081;
    3. School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081
  • Received:2020-11-26 Revised:2021-10-22 Online:2021-11-20 Published:2022-02-28

Abstract: The driver's intention information of the front vehicle is very important to judge the danger of the rear vehicle active collision warning and avoidance model. To solve the problem of high false alarm rate and delayed braking, an active collision warning and avoidance model considering the driver's intention of the front vehicle is proposed. First, BP neural network and hidden Markov model are selected as the main model of the driving behavior layer and the driving intention layer, then, the intention recognition model is built by using the intention observation data of the front vehicle's brake pedal, accelerator pedal and speed collected by the driving simulator as input, so as to realize the recognition of the front vehicle driver's intention of accelerating driving, uniform driving, normal braking and emergency braking. Secondly, based on the intention recognition result of the driver in front and road attachment information transmitted to the rear vehicle by the Internet of Vehicles, the vehicle active collision warning and avoidance model considering the intentions of the driver in front is established to dynamically judge the collision risk and adjust the warning and braking logic. Finally, in order to verify the accuracy of the proposed driver intention recognition model and the effectiveness of the active collision warning and avoidance model, a co-simulation platform based on Simulink, Carsim and PreScan is constructed and the multi-condition experiment test is carried out. The results show that the average recognition accuracy of the proposed BP-HMM model is 94.17%, which is significantly better than that of BP or HMM models. The average positive alarm rate of the active collision warning and avoidance model is 93.43%. Compared with the TTC, Mazda and the model consider the driver's intention of following vehicle, the average false alarm rate is reduced by 16.12%, 23.43% and 26.67%, respectively. In the automatic emergency braking test condition, collision can be successfully avoided. The shortest relative distance between the two vehicles is mostly kept within the range of 2-8 m, with an average of 3.698 m, which has higher safety and stability.

Key words: wet clutch, axial force, drag torque, flow pattern

CLC Number: