On Numerical Descriptions of Road Transport Missions

Licentiate thesis, 2017

This thesis addresses some issues of current interest in energy consumption prediction through simulation. First, we review the situation for rating, regulation and legislation of CO$_2$-emissions for cars and heavy vehicles. We explain some of the problems with the current description (or lack thereof) of the road and surroundings for such tests, called driving cycles. With that in mind, two main research questions are formulated. The first we call the `representation problem': what to include in a numerical description of a transport mission, and how to represent it mathematically? In answer, a proposal for a format is derived; the operating cycle-format. It is a physical description of the transport mission that consists of four parts: road, weather, traffic and mission, with the important property that it is independent of both driver and vehicle. Furthermore, it is explained how to build a simulation model capable of using the new mission description. Next, this is applied in a case-study of a real-world cargo transport, and the simulation results are used in a product development situation to improve energy consumption. In this specific case, a fuel consumption improvement of 16\% is achieved. The second question we call the `classification problem': how should a mission executed in a specific region be labelled (i.e. described on a high level) depending on its characteristics? In answer, two classification methods are discussed: the Global Transport Application (GTA) and stochastic models. The basic structure of GTA is explained, and it is applied to the same log file that was used in the case-study. The principles of classification through stochastic models is described by explicit construction of such a model for topography. An example of how the methods can be applied in sales-to-order is made, by investigating how to best choose buffer size for a hybrid truck. Finally, a process for both efficient product development and sales-to-order is outlined, that combines the format proposal and the two classification methods. If the output of the process is used in an optimisation process, the result is a vehicle configuration tailored for the transport mission in question.