Probabilistic modeling of nodal charging demand based on spatial-temporal dynamics of moving electric vehicles

• Main Author: Musa, Ryan Rezal
• Other Authors: Soh Cheong Boon
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Electrical and electronic engineering::Power electronics
• Online Access: https://hdl.handle.net/10356/149700
• Institution: Nanyang Technological University

Global warming has long been a hot topic in discussion around the world very frequently in the past recent years and numerous times people have tried to make the situation improve and get better. The world soon discovered that transportation contributed to a huge portion of greenhouse gases production. These gases lead to the depletion in the ozone layer and therefore global warming. Because of this negative effect transportation had on the environment, many power service providers and automobile companies joined forces to invent a new technology in transportation that would not have harsh effects on the environment called Electric Vehicles (EV). In the recent years, many countries and major influential companies have slowly increased the awareness of the health on the environment, and this led to the public becoming more welcoming to the idea of EVs usage. Because of this, the EV companies are always in the midst of Research and Development (R&D) of the upgrades and improvements of current EV technologies as well as constantly teaching customers to be more open to EV technology and to expose them to realise the positive effects of EVs. This led the companies to come up with more upgrades to further enhance and improve the current versions of EVs and build more charging stations to be on par with the demand for more EVs from the consumers. The sudden increase in usage of EVs and the sudden surge of injection of high amounts of EVs into the traffic transportation grid and as random moving loads into the power system have made the stress of their negative effects more prominent and is gaining more urgent attention from power service providers all around the globe. Because of the spatial-temporal random behaviours of EVs, it is very difficult to identity and to pinpoint the locations of the time and space changing effects. Many studies before this have used the method of checking the whole of the system of EV charging demand on the foundation of the data analysis together with the fixed charging venues and time slots. However, in this project’s case, this report is based on the mechanics of a probabilistic model for nodal charging demand which foundation is on the method of spatial-temporal dynamics of moving EVs. After the integrated system with graph theory is introduced, a spatial-temporal model of moving EVs as loads will be developed on the foundation of random trip chain and the Markov decision process. (MDP). From the probabilities of a single EV as well as multiple EVs’ charging behaviour, the nodal EV charging demands are figured out. The system studies in this project shows us that this model can be used to check the nodal charging demand because of the spatial-temporal dispersion of randomly moving EVs. Around the world in the past few years, EV companies realised that the demands for EVs are growing, and it will keep growing in years to come, the companies predict. This makes it possible to have a chance of power congestion and overload due to the growing EV charging demands that also lead to increase in EV charging stations and their usage. To make matters even worse, that, combined with the impossible task of predicting the random acts of EV drivers, to make an intelligent guess and to pinpoint the place and time that a potential power congestion may occur would be very difficult. So, in this project, a probabilistic model based on spatial-temporal dynamics will be produced together with MATLAB, a programming software to perform the simulation of various situations in this case and then to show visually and to tell us where in the power grid system the distribution in the form of graphs and charts of the load demands at a particular charging location and at a particular time of the day. Applying this to real life situations, this type of probabilistic models would be useful for the many power service providers out there as they can use them to predict when the peak periods are going to happen and then take the corresponding necessary preventive measures to avoid potential power congestions from occurring. This report consists of research information and theories related to the study of this project. The progress of this project will also be recorded inside this report as well as the results attained.