Driver driving behavior automatic analysis method based on long-term and short-term memory network

Abstract

The invention discloses a driver driving behavior automatic analysis method based on a long-term and short-term memory network. The method comprises the steps that step one, feature data is extractedand preprocessed; step two, a sliding window is designed; and step three, a driving behavior automatic analysis network is established. The method particularly comprises the following steps that firstly, driving data containing driving behaviors of a driver is extracted from an automobile CAN bus; the driving behaviors of the driver are scored by using the riding experience of a passenger, and theriding experience is taken as a label for evaluating the driving data of the driver; a driver driving behavior automatic analysis network is established based on the long-term and short-term memory network, deep training of the driving data and label data of a large number of drivers is carried out, driving behavior habits of the drivers are learned, and a driving behavior automatic analysis library is established; and finally, when any driver drives an automobile, the established driver driving behavior automatic analysis network can achieve automatic analysis of the driver driving behaviorsaccording to the driving data and judge whether the driver has an excellent driving behavior habit or not.

Description

technical field [0001] The invention belongs to the technical field of electronic information, in particular to an automatic analysis method of driver's driving behavior based on a long-short-term memory network. Background technique [0002] The integration of machine learning and the automotive industry has injected new impetus into the intelligent networked automotive industry. However, there are still many difficult problems to be solved, such as automatic analysis of driver's driving behavior. Since passenger comfort has always been the focus of the automotive industry, automated analysis of drivers' driving behavior has great application value. For example, for private cars, automated analysis of driving behavior can improve drivers' driving habits and reduce traffic accidents; for companies such as buses and online car-hailing companies, drivers with excellent driving behavior can be recruited to bring passengers more comfortable rides Experience, thereby improving ...

AI Technical Summary

Problems solved by technology

[0005] 1) The car CAN bus data contains a lot of car status information, how to extract the driver's driving behavior from the car CAN bus data is a major challenge

[0006] 2) The quality of the driver's driving behavior is affected by multiple factors such as vehicle acceleration, speed, and steering wheel angular velocity. How to identify the driver's driving characteristics from a large number of on-board sensors is a challenging task

Images