关键词: 应急车辆；YOLOX 算法；DeepSORT 算法；车辆统计监测系统；ETA 算法

Abstract: With the development of economy and society, road congestion caused by urban traffic saturation is increasing, which hinders the passage of emergency vehicles. In order to improve the efficiency of emergency vehicle traffic, taking Lhasa as a case, the existing signal control system is assisted and optimized on the basis of YOLOX algorithm and DeepSORT algorithm, and the vehicle statistical monitoring system architecture is constructed. Then the estimated time to arrival (ETA) algorithm is used to calculate and estimate the arrival time, so as to realize the intelligent networking control of emergency vehicles and traffic lights. Finally, a mathematical model of multi-objective traffic signal timing is established. By reducing the road saturation on the driving path of emergency vehicles, other vehicles in front of the vehicle have space to make way for emergency vehicles, so that emergency vehicles can drive at high speed. The results show that during the peak period of traffic flow, the average delay of vehicles can be reduced by 33.9 % and the traffic capacity of vehicles can be increased by 42.9 % after using the system. The system provides new ideas for the optimization of urban traffic flow and the improvement of governance level, and provides impetus for the development of urban traffic intelligence.

Key words: emergency vehicles; YOLOX algorithm; DeepSORT algorithm; vehicle statistical monitoring system; ETA algorithm