The Urban Simulations Group focuses on modelling and optimisation for architectures and infrastructure of all aspects of electro-mobility. Despite the increasing popularity of electric vehicles, there does not exist a model case of city-wide use of/transition to electric vehicles; it is, however, imperative that we understand how design choices at all levels affect the entire electric transportation infrastructure. Development of efficient, effective and accurate simulation methods and tools gives experts - be they scientists, engineers or decision makers - with access to a holistic view of the electric transportation system. This would not only enhance their ability to explore the different aspects of their problems and solutions but also open up several new avenues of exploration .
Thanks to state-of-the-art computer technology, simulation methodology, and general software tool development, we have the opportunity – for the first time in history – to use simulation as a powerful base technology,for predicting, anticipating, evaluating and assessing the consequences of decisions. Through simulation, we can help to steer the development and deployment of electric vehicles in a direction that maximizes societal benefit and that can substantiate discussions in society about technology assessment.
The central theme of the work done in the Urban Simulations Group and presented on this website is hence the idea of multi-scale simulation. This refers to the ability to integrate models and simulation tools from different levels of scale (internal battery and/or fuel cell processes, models of battery modules, models on the vehicle system level, models of the power grid, etc.). This also encapsulates the idea that the total simulation involves different scales and granularities, various methodologies, diverse computer platforms.
SEMSim – Scalable Electromobility Simulation
Central to the work in the Urban Simulations Group is the development of a novel multi-scale simulation platform aimed for conducting large-scale exploratory simulation experiments: the Scalable Electromobility Simulation (SEMSim) platform. SEMSim is based on a discrete-event agent-based traffic simulation which will integrate various vehicle component models and driver behavior models that are relevant to electromobility (see figure below). In addition, the traffic simulation can be coupled with other simulations, such as a power system simulation, for example.
There are three major challenges which are addressed by the work done in our group: (1) computational performance, (2) simulation coupling, (3) data analysis and synthesis.
Computational performance is crucial in the context of exploratory simulation which requires the execution of many thousand simulation scenarios. This requires the simulation to be able to run significantly faster than real-time in order to be able to explore a given scenario space within a reasonable amount of time. For this purpose, SEMSim is based on a parallel discrete-event simulation engine which is designed for the execution in high-performance computing environments.
Simulation coupling is concerned with the ability to have different simulation instances coupled such that they can be executed together. While some models, such as vehicle component models, can be directly integrated into the agent-based traffic simulation; others like the power and environment simulation are decoupled and connected through a High Level Architecture (HLA).
Data is crucial for SEMSim: (1) data is needed during the modeling process in order to create and validate the various models in SEMSim; and (2) data with specific characteristics is needed to reflect certain assumptions as part of exploratory what-if scenario analysis. Although several public and private entities already collect various data, access to it may be restricted. Therefore, some of the work that is done involves investigating alternative methods for data collection through crowd sourcing and participatory sensing. In addition, we investigate methods and algorithms for analyzing and synthesizing artificial data sets.
The modelling and simulation tools developed by RP 5 are aimed for use in planning of infrastructure and vehicle design. With the help of the exploratory simulation facilities, we will be able to detect and analyze inter-system dependencies from a holistic viewpoint. For example, the SEMSim platform allows us to investigate how design decisions regarding battery capacity will influence the transportation system as well as the charging infrastructure and the power system. The ability to analyze these effects can help decision makers to make an informed decision regarding the future of electro-mobility.