Research on the BP neural Network of Bus Unsafe Driving Behavior

Abstract

The security issue of urban bus is increasingly prominent in our country. Promoting scientific and reasonable unsafe driving events record system is an effective way to improve the level of public transport safety management. At the same time, the reasonable record system contribute to make the correct bus driving behavior, and it is also can reduce the loss of public traffic resources. At first, this paper puts forward using tri-axial acceleration sensor to collect the data of the car. Then classifying of bus driving behavior, setting up a series of driving behavior model, including the brake model, throttle model, a sharp turn model, according to the driving behavior model of data processing, and further by BP neural network to establish a BP neural network to effectively supervise bus driver's driving behavior.

DOI : http://dx.doi.org/10.11591/telkomnika.v12i3.4657