Efficient determination of driving resistance through system identification based on driving route information and weather data

Abstract

Predictive driving functions based on load profile prediction can make an important contribution to the urgently needed further reduction in energy consumption of road vehicles. In this context, this paper presents a method that precisely and efficiently predicts the driving resistances of an upcoming trip by means of online parameter estimation and system identification based on the driving route and weather data. A parameter estimation has been implemented that determines at the beginning of each trip different parameters such as the total vehicle mass, the dynamic rolling radius as well as the absolute and relative inclination of the vehicle relative to the road. It then performs an online calibration of the speed and acceleration calculation. In addition, a method was developed that allows the prediction of driving resistances based on the estimated parameters as well as environmental data such as road roughness and wind conditions. The developed functions were validated against real-world measurements under different environmental conditions and were able to reduce the prediction error in parameter estimation to a median of less than 2 %, while the prediction error of the driving resistances was reduced from sometimes more than 40 % to a few percentage points compared to conventional approaches.