Driver evaluation in heavy duty vehicles based on acceleration and braking behaviors

This is the latest version of this item.

Official URL: https://dx.doi.org/10.1109/IECON43393.2020.9255274

In this paper, we present a real-time driver evalua-tion system for heavy-duty vehicles by focusing on the classifica-tion of risky acceleration and braking behaviors. We utilize animproved version of our previous Long Short Memory (LSTM)based acceleration behavior model [10] to evaluate varyingacceleration behaviors of a truck driver in small time periods.This model continuously classifies a driver as one of six driverclasses with specified longitudinal-lateral aggression levels, usingdriving signals as time-series inputs. The driver gets accelerationscore updates based on assigned classes and the geometry ofdriven road sections. To evaluate the braking behaviors of atruck driver, we propose a braking behavior model, which usesa novel approach to analyze deceleration patterns formed duringbrake operations. The braking score of a driver is updated foreach brake event based on the pattern, magnitude, and frequencyevaluations. The proposed driver evaluation system has achievedsignificant results in both the classification and evaluation ofacceleration and braking behaviors.