Driving Assistance Decision Method Based on Dynamic Driving Map

Aiming at the problem of vehicle driving map modeling, this paper puts forward a method of constructing dynamic driving map(DDM). Firstly, the intention of road vehicles is identified based on hidden Markov model, which is regarded as one layer of driving map. Then, this paper hierarchically processes and fuses the current environment from four aspects: vehicle movement information, lane information, vehicle and obstacle information and vehicle prediction information, and obtains the drivability probability of each unit. Then, based on DDM, the behavior decision-making cost equation is put forward, and the decision-making model of vehicle driving behavior with minimum cost is designed by using the idea of dynamic programming. Finally, through the simulation verification, the results show that the dynamic driving map designed in this paper can accurately and effectively express the complex road environment compactly, and give reasonable driving advice according to the specific environment, which has good robustness.