Mobility data-driven planning of electric vehicle charging infrastructure for decarbonising energy and transport systems

Project supervisors

Dr Hao Wang, Faculty of Information Technology, Monash Energy Institute (Main Supervisor)
Dr Mike Ma, Faculty of Engineering

PhD project abstract

The adoption of electric vehicles (EVs) is widely accepted as a fundamental approach to mitigating climate change for a sustainable future with cleaner energy and transport systems. Furthermore, the technological advancements in batteries and chargers enable the massive deployment of EVs. However, the lack of planning for charging infrastructure and consumers’ concerns known as “range anxiety” are amongst the top barriers to the wide adoption of EVs. This project aims to develop a paradigm for data-driven modelling of EV behavours, optimisation for EV charging station planning, and AI-based pricing design for EV charging coordination. More specifically, this project has the following three research tasks.

  1. We will characterise realistic EV mobility behaviours and charging preferences using real-world spatio-temporal traffic data and open-access consumer surveys.
  2. We will develop an large-scale optimisation problem for determining the optimal placement and sizing of charging infrastructure.
  3. We will design an AI-empowered pricing scheme considering heterogenous EV behaviours and preferences under a dynamical energy-transport environment with uncertainties.

The outcomes of this project will provide new models and algorithms for energy-transport integration, advancing the knowledge of mitigation strategies for sustainable urban development.

Areas of research

Climate change; smart city; electric vehicle; data analytics; transport electrification

Project description

Electric vehicles (EVs) play a significant role in the transition to a sustainable future with cleaner energy and transport systems. On the one hand, the lack of public charging infrastructure and consumers’ concerns known as “range anxiety” are amongst the top barriers to enable such a transition and must be removed to provide confidence in adopting EVs. On the other hand, due to the complex spatiotemporal behaviours of EVs, the unmanaged electric charging demand from EV fleets will significantly impact the existing energy and transport systems. Therefore, reliable charging infrastructure and charging strategies are the prerequisites to the successful adoption of EVs.

This PhD project aims to develop a paradigm for the optimal planning and management of EV charging infrastructure using data analysis, optimisation, and AI. The mobility patterns and charging demand will be analysed to model EV behaviours. An optimisation problem will be formulated to determine the optimal placement and sizing of charging infrastructure considering a variety of constraints in the electric power grids and transport networks. An AI-based EV charging strategy will be designed to guide the EV routing via energy-aware incentive signals by taking EV mobility behaviour and preferences into account.

This project is highly interdisciplinary in nature across mobility behaviour, EV driver preference, power system planning, and transport network modelling. Supported by an interdisciplinary supervisory team, this project will answer the following research questions.

  1. How to characterise realistic EV mobility behaviours and charging preferences using real-world spatio-temporal traffic data and open-access consumer surveys?
  2. How to plan a network of charging stations in order to optimise the convivence and costs of charging without causing significant peak load in the energy grids or congestion in the transport networks?
  3. How to manage EV charging by providing effective pricing signals considering heterogenous EV behaviours and preferences under a dynamical energy-transport environment with uncertainties?

To conduct this project, we need to study the data-driven modelling and solution across sustainable energy and transport systems. The objective is to optimise the placement and sizing of EV charging stations based on the spatial and temporal data and EV driver satisfaction. The methodologies include data analysis, optimisation with a focus on large-scale mixed-integer programming, and AI-based EV coordination to mitigate the negative impact on both energy and urban transport networks.

As part of the PhD project, the following design factors need to be assessed.

  1. Comparison between centrally controlled autonomous EV fleet charging and individual preference based EV charging;
  2. The opportunity of combining electricity generated from renewable sources in the charging infrastructure planning problem;
  3. The impact of pricing schemes, such as flat rates, time-of-use rates, and dynamic rates, on the effectiveness of EV charging management;
  4. The trade-off between the possible impact on energy grids and transport networks, such as peak load and traffic congestion.

Sustainable development is a growing research focus in different faculties across the university, including Information Technology and Engineering. The University and FIT have been significantly investing in major projects and themes such as Net Zero Initiative and IT for Sustainability. This project is perfectly aligned with these research priorities and will be led by academics with a proven track record to contribute to a significant research area of energy and transport systems as part of sustainable urban development.

The outcomes of this project will provide new models and algorithms for energy-transport integration, in particular EV charging infrastructure planning and management. Specifically, the expected deliverables are as follows.

  1. Data analytical method and modelling tool for EV spatio-temporal mobility behaviours and charging preferences; Statistical distribution of origin-destination pairs over zonal models;
  2. An optimisation framework for determining the optimal placement and sizing of charging infrastructure based on mobility scenarios using the benders decomposition method;
  3. A pricing scheme for EV coordination to the available charging infrastructure considering the impact on both energy grids and transport networks using multi-agent reinforcement learning.

Efficient deployment of the network of charging stations is a critical problem, which will play a significant role in prompting the adoption of EVs, managing the charging infrastructure in the near future. The deliverables of this project have a great potential to help the stockholders in the energy, transport sectors, including distribution network operators, transport agencies, policymakers and the government, to facilitate environment friendly EV charging planning and management in Australia.

PhD student role description

This project provides a perfect opportunity for the candidate to work in one of the most exciting problems of our time, i.e., mitigating climate change via transport electrification. In this project, the candidate will work with leading researchers in IT and engineering, access world-class research resources, learn state-of-the-art techniques, such as spatio-temporal data analytics, stochastic optimisation, multi-agent reinforcement learning, in an interdisciplinary area across energy and transport systems.

Role/contribution in the project

  • Collaborate with supervisors and other team members
  • Analyse the mobility data and design optimisation and machine learning algorithms
  • Validate the design on energy-transport simulation platforms
  • Present and promote research findings in international conferences and workshops
  • Publish scholarly works in top venues
  • Work with industry partners in energy and transport sectors if any.

Through this project, the candidate will become an expert in AI and data science for sustainable energy-transport interfacing system

  • with strong ability of problem-solving using AI and data science tools,
  • strong skills in analytical thinking and programming,
  • a deep understanding of critical infrastructure and sustainable development, building a unique blend of expertise and experiences for a future career in academia and/or AI, IT, Energy, and Transport industry.

Required skills and experience

  • Degree in Computer Science, IT, Civil/Transportation Engineering or equivalent
  • Strong programming skills in Python, Julia, etc.
  • Strong mathematical and analytical skills
  • Knowledge of electricity network analysis and simulation with network design knowledge highly desirable
  • Excellent written and verbal communication skills.

Potential start date

1 December 2021

Submit your Expression of Interest