Effects of errors in position- and navigation data on predictive vehicle operational strategy optimization

Abstract

Predictive operational strategies enable fuel savings by optimizing the powertrain operation of a vehicle taking lookahead data like velocity limits and slopes of the planned route into account. Gear changes, free-wheeling, speed and operational modes are optimized along that route to achieve minimal fuel consumption. However, these optimizations assume complete information on the route properties ahead. To evaluate the impact of look-ahead data quality on the predictive operational strategies of passenger vehicles, the fuel consumption of different optimized speed trajectories are simulated. The fuel consumption of a speed trajectory with perfect knowledge is then compared to the energy needs of different speed trajectories generated on the same route by optimization methods based on erroneous look-ahead data. In the tested scenarios, optimized velocity trajectories show in most of the cases same or better fuel consumption than constant, non-optimized speed profiles, even if they are based on erroneous map data. Errors in look-ahead data can be shown to have bigger effects on heavy vehicles.