A safe driving speed prediction method for the Hong Kong-Zhuhai-Macao bridge based on genetic algorithm and BP neural network

The Hong Kong-Zhuhai-Macao Bridge, as the world's most up-to-date cross-sea bridge, spans the Lingdingyang sea, and traffic is easily affected by typhoons, heavy rain and other weather conditions. Besides, there are various road conditions such as tunnels and turning overpasses on the bridge. With the increase of traffic flow, it is important to control the speed of the bridge through traffic for the safe driving of vehicles. The purpose of this paper is to design a vehicle driving speed model to calculate the safe driving strategy on the Hong Kong-Zhuhai-Macao Bridge based on the bridge design. Two models are established in this paper: Model Ⅰ: Curve driving speed model based on cars of different masses; Model Ⅱ: Safe driving speed model based on cars under the influence of different typhoon wind directions. For model Ⅰ, this paper gets the theoretical driving speed of the car through the physical force analysis of the car and the linear speed prediction model of the flat and longitudinal combination. For model Ⅱ, this paper predicts the maximum safe speed of the car through genetic algorithm and BP neural network with the input vehicle type, road curvature parameters and car entry speed, and gets the safe speed of the car under different wind levels by combining with wind model. Finally, an AHP model is established by combining different condition factors with the car-following model, and the capacity of the bridge is considered comprehensively to find out the duration of motor vehicles passing the bridge.