Description of driving patterns through characteristic parameters Academic Article in Scopus uri icon

abstract

  • © ECOS 2019 - Proceedings of the 32nd International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems. All rights reserved.There is an increasing interest to describe properly local driving patterns. A driving pattern is the way drivers drive their vehicles in a region and it is related to the human aspect involved in the vehicles usage. Driving patterns are described trough characteristic parameters (CPs) which are variables resulting from any combination of speed-time such as the average speed, average positive acceleration, among others. Thus, it is customary to describe how people drive in a given city specifying their average speed and average acceleration. However, it is still unknown which set of CPs properly describe the driving patterns. Driving patterns are also described by driving cycles (DCs). They are speed-time series that can also be described by their characteristic parameters (CP*s). Then a DC represents a driving pattern when CPi=CP*i. Existing DCs are used by carmakers to evaluate the performance of their vehicles as part of the regulatory process to sell a new vehicle technology in a market. However, the DCs currently used by carmakers do not represent the local driving patterns. The correct description of driving patterns is a requirement to assess the real energy and environmental performance of the vehicles in that region. Errors in the description of the driving patterns have led to large differences between fuel consumption and the emissions reported by manufacturers and the observed in the normal use of the vehicles. This study focus in the identification of the minimum CPs that properly describe a driving pattern, from the point of view of their use to correctly evaluate the vehicle energy consumption and tailpipe emissions. Twenty-five widely used CPs were considered. We used measurement data of fuel consumption, emissions and speed, of a fleet of 15 vehicles operating in two different urban regions during two months. Using that database, we constructed the representative DCs using Micro-trips method and selecting any combination of 2 or 3 CPs as assessment criteria. We also used the methodology proposed by [1] to evaluate the degree of representativeness of the resulting DCs. More than 1140 iterations where performed considering different combinations of CPs. Results showed that by using standard deviation of the acceleration, the percentage of time in idling, the average speed and the vehicle specific power (VSP) as CPs as assessment criteria, the resulting DC represents the driving pattern in the region of interest with an average-average relative difference (ARD - ) lower than 10%.

publication date

  • January 1, 2019