An Algorithm for Terrain Recognition of Heavy Duty Diesel Vehicles Using Chinese Regulatory Standard Data
Liang Yang, Huajun Wang, Wenbin Wang, Tiegang Gao, Hang Gao · 2022 Asia Conference on Algorithms, Computing and Machine Learning (CACML) · 2022
An algorithm for terrain recognition of heavy-duty diesel vehicles using Chinese regulatory standard data is proposed in this paper. First of all, clean the data; Then, select and calculate the required eigenvalues according to the different characteristics of diesel vehicles on different terrain; Finally, an improved depth learning network based on multi-headed self-attention is designed to estimate the proportion of driving mileage on different terrain. In addition, a terrain database of heavy diesel vehicles driving roads is constructed. The experimental results show that the proposed algorithm has low computational complexity, universal applicability, and good recognition performance on the constructed terrain database. At present, there is no similar algorithm for terrain recognition of heavy-duty diesel vehicles using Chinese regulatory standards data. The algorithm in this paper is the first in the industry.