Search

The bird's-eye view (BEV) method, which is a vision-centric representation-based perception task, is essential and promising for future Autonomous Vehicle perception. It has advantages of fusion-friendly, intuitive, end-to-end optimization and is cheaper than LiDAR. The performance of existing BEV methods, however, would be deteriorated under the situation of a tire blow-out. This is because they quite rely on accurate camera calibration which may be disabled by noisy camera parameters during blow-out. Therefore, it is extremely unsafe to use existing BEV methods in the tire blow-out situation. In this paper, we propose a geometry-guided auto-resizable kernel transformer (GARKT) method, which is designed especially for vehicles with tire blow-out. Specifically, we establish a camera deviation model for vehicles with tire blow-out. Then we use the geometric priors to attain the prior position in perspective view with auto-resizable kernels. The resizable perception areas are encoded and flattened to generate BEV representation. GARKT predicts the nuScenes detection score (NDS) with a value of 0.439 on a newly created blow-out dataset based on nuScenes. NDS can still obtain 0.431 when the tire is completely flat, which is much more robust compared to other transformer-based BEV methods. Moreover, the GARKT method has almost real-time computing speed, with about 20.5 fps on one GPU. [ABSTRACT FROM AUTHOR]

The development of the automotive technology makes the vehicle mobility possible. The current technologies commonly used include active suspension, intelligent differential lock, central tire inflation system, and torque control-based engine technology. These new technologies cannot be applied unless the road character is predicted by the vehicle. Soil support strength "cone index" (CI, pound/inch²) can be classified into 3 categories: 30-75, 75-125, and > 125. Based on the CI classification, the road also can be classified. On the road with CI Category 1, the vehicle usually cannot go through; on the road with CI Category 2, the vehicle can pass through quickly by the soft road pattern; and the road with CI Category 3 can barely affect the mobility of the vehicle. In this study, to identify the soft road, the rolling coefficients threshold based on the experiences was established between Category 1 and Category 2. It assumed that the mobility of the vehicle would be better improved if a vehicle identified the road feature and then adjusted the vehicle-self running at the soft pattern automatically. To increase the mobility of the military off-road vehicle, one identification method of the soft roads based on real-time monitoring of the rolling resistance was put forward. Since the tire barely deformed, it could be regarded as a rigid-body when a vehicle was running on a soft road. The width of the tire could affect the coefficient of the rolling resistance, but the character parameters of soft roads were the major factors affecting coefficient of the rolling resistance. When the vehicle was driven on soft road, the movement of the vehicle could be prevented by soft soil because not all soil counterforce were effective, for example, some of the resistance forces did not produce propulsive force but were consumed to make soil deformation. The character parameters of soft roads in Institute of Changchun Wetland were measured and the CI was calculated. The rolling resistance of a wheel on real compact terrain was calculated based on Bakker's terramechanics pressure-sinkage equation. The results showed that the rolling resistance was less than 0.03 when the vehicle ran on concrete road, less than 0.1 on hard road, and greater than 0.2 on sand. Therefore, the rolling resistance of 0.2 was considered as the critical value for soft road identification. In a real condition when a vehicle ran on a road, by real-time monitoring the running condition of a vehicle and its engineer on real roads, the coefficient of the rolling resistance could be calculated based on the longitudinal force equation using the engine transient output torsion. If the obtained coefficient of the rolling resistance was greater than 0.2, the vehicle would run at a soft road pattern and pass through the soft road more quickly. In a test study, the accuracy of a vehicle successfully identifying road condition reached 90%, confirming that the rolling resistance-based soft road identification method proposed here was reliable. [ABSTRACT FROM AUTHOR]

A finite element model for the gear and shaft system is developed, on which a finite element analysis is conducted to calculate the deformation of gear shaft and the contact stress of gear teeth, and analyze the effects of gear shaft deformation on the contact state of teeth surfaces. Through comparing the results with that of calculation by classical method, it is shown that using the finite element model built can not only calculate shaft deformation and teeth contact stress, but also comprehensively analyze the effect of shaft deformation on the load distributions in teeth contact area and between teeth as well as the contact stresses of teeth surfaces, providing bases for more thoroughly and accurately analyze the teeth surface contact state and load distribution of transmission gears and predict the fatigue life of gears. [ABSTRACT FROM AUTHOR]