Smart Management of Electric Vehicles

Lead Research Organisation: CARDIFF UNIVERSITY
Department Name: Sch of Engineering

Abstract

Distribution networks are typically designed for specific electrical loads using assumptions based on typical load consumption patterns. Battery charging of Electric Vehicles (EVs) will increase the power demand in distribution networks and large scale electric transport will require smart management of the charging infrastructure. Depending on the location and times the vehicles are plugged in, they could cause local constraints on the grid. Analysis tools are required to determine the effects of adding a large number of mobile EVs to the grid, as well as the customers' location, charging time and duration on a daily basis. The main problem in the modelling of the aggregation of EVs is the representation of the uncertainties including: (i) type of residential load, (ii) EV location, (iii) rating of EV charger, (iv) EV charging occurrence and (v) EV charging duration.

An EV aggregator proposed in this research will act as a key mediator between the consumers on one side and the markets and the other power system participants on the other side. The EV aggregator may have to forecast: (i) the electricity consumption of its own customers, for forecasting the aggregator's power balance and (ii) the consumption in the electricity system, for forecasting electricity prices. The impact of EVs is significant for the Distribution Network Operators (DNOs) as there is a need to manage congestion and voltage drops. As the predicted large deployment of EVs could have an important impact on the grid it is expected to adapt the vehicle as much as possible with the existing infrastructure and this can be achieved by the integration of smart grid control techniques. The primary goal of a Smart Grid is the optimal control of the electricity distribution and the charging of EVs can be controlled to reduce peak load.

In order to answer these questions, this project draws on methodologies and results across the boundaries of engineering and informatics. This is an exciting opportunity to bring qualitative and quantitative research methods together to study a complex system covering load forecasting and smart management of EVs. This project aims to (i) investigate control algorithms for smart management of EVs considering the spatial diversity of EVs throughout the network and temporal diversity of EV charging patterns and (ii) demonstrate a practical way of implementing control algorithms to facilitate the future deployment of EVs by laboratory validation.

Planned Impact

This project will investigate control algorithms to facilitate the future deployment of EVs. The research will:
- Produce open source algorithms for the smart management of electric vehicles that could provide the basis for other researchers to verify and build on this work.
- Provide a classification of the historical data set obtained from Plugged in Places projects if these will be available.
- Provide a smart management control algorithms for the charging of electric vehicles.
- Validate the algorithms against real time simulation and laboratory tests.

The main beneficiaries of this research will be the distribution network operators as the day-ahead short-term load forecasting accuracy affects the economic operation and reliability of the system. Furthermore the electric vehicles aggregation control algorithms will assist distribution system operators to avoid system reinforcement.

Additional beneficiaries include government and policy representatives (e.g. WAG, and the European Commission), third sector organisations (e.g. Future Transport Systems, Energy Saving Trust, Carbon Trust), representatives of the energy industry (UPL Utility Infrastructure & Energy Management, Mott MacDonald) and academics across a number of disciplines.

The international collaboration with TECNALIA technology centre in Basque country will provide a basis for future collaboration on research projects.

In addition the applicant will also collaborate with the Cardiff School of Engineering's newly established Knowledge Transfer Centre. This Centre, funded by the Welsh Assembly Government, is designed to facilitate the transfer of technology developed within the School to the commercial sector, and is recruiting experienced Technology Translators to assist in this work. The project will be able to draw on their expertise in knowledge transfer activities in order to achieve maximum impact and help the UK remain at the forefront of engineering technology.

Publications are a vital part of the dissemination strategy. Successful research results will be published in high-impact factor journals.
 
Description This research made technological advances in data analysis and control technique development.
A short-term EV load forecasting model was developed using Support Vector Machines (SVM), an artificial intelligence technique. Charging events for one year were created using national statistical data, to cover the lack of real historical data of EV's charging events. The SVM model was trained and provided a forecast for the day-ahead EV demand. This forecast was compared with the actual demand of that day, as well as with the forecast output of the Monte Carlo technique. The results prove the effectiveness and accuracy of the SVM proposed model, over a more statistical approach.
A decentralized control framework was developed for a mixture of responsive and unresponsive EVs. The control algorithm was enhanced by an EV load forecasting module to increase its effectiveness. The main aim of the control algorithm was to achieve a valley-filling effect on the demand curve.
The effectiveness of the control algorithm was tested in a UK generic distribution network considering a geographical area with 3072 customers. Two case studies were investigated. The first case study considered an EV fleet charging at 11kW charging stations comprising of responsive and unresponsive EVs. It was demonstrated that when the EV load forecast option is activated the EVs are adapting their charging schedule to reduce the impact of the unresponsive EVs on the demand curve. The second case study investigated the effect of the charging station's power rate on the effectiveness of the decentralized control model. It was shown that when the forecasting module is activated there is a demand peak reduction for every combination of responsive/unresponsive EVs considering charging rates of 3kW, 11kW and 22kW.

A multi-agent algorithm (MAS) was developed for the management of EV charging. The MAS consisted of an EVs/DG aggregator agent and "Responsive" or "Unresponsive" EVs agents. The EVs/DG aggregator agent was responsible to design the appropriate virtual pricing policy so that it can maximize its profit. "Responsive" EVs agents were able to respond rationally to the virtual pricing signals, whereas "Unresponsive" EVs agents were defining their charging schedule regardless the virtual cost. Three cases studies were experimentally validated at the Electric Energy Systems Laboratory of the National Technical University of Athens. The first case study investigated the impact of EVs charging on the UK distribution network when the EVs/DG Aggregator was located in the MV/LV transformer. It was demonstrated that the location of the EVs/DG aggregator agent affects the demand and voltage profiles of the LV feeders. The second case study demonstrated the value of the EVs load forecasting in the control strategy. When the EVs/DG aggregator has load forecasting capabilities, the responsive EVs agents are adapting their charging schedule to reduce the impact of the unresponsive EVs agents on the demand curve. The third case study tested the capability of responsive EVs agents to charge preferentially from renewable energy sources. The results demonstrated their capability to reschedule their charging demand according to a real time PV generation profile.
Exploitation Route The control algorithms developed in this project can be used by the charging stations providers to enhance their control platform.
The short-term load forecasting algorithm can be used by the DNOs to predict the electric vehicles charging demand.
It was demonstrated that the smart management of EVs charging based on aggregation enhanced by EV load forecasting could be seen as a win-win strategy for both the distribution network operation (DNO) and the vehicle owners.
Sectors Education

Energy

Environment

Transport

Other

 
Description Influence on local government policies: I was a member of Low Carbon Vehicles Expert Steering Group at Welsh Government and the report that we provided advised Edwina Hart AM CStJ MBE, Minister for Economy, Science and Transport. She said [...] I welcome the report which makes a range of recommendations, a number of which I am pleased to say we are already taking forward with individual businesses as part of my portfolio responsibilities. A major theme of the report is the desirability of having a more integrated approach in terms of government advice and support, which I support. This is reflected in other key recommendations which have significant cross portfolio implications both in terms of policy and funding.I commend these recommendations for future consideration, in recognition that they require looking at in more detail and within a joined up approach, in order to be progressed effectively." http://gov.wales/about/cabinet/cabinetstatements/previous-administration/2016/lowcarbonvehicles/?lang=en
First Year Of Impact 2016
Sector Energy,Environment,Transport
Impact Types Societal

Economic

 
Description Deployment of low carbon vehicles
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Contribution to a national consultation/review
 
Description Agent-based controllers for EVs and micro-generators
Amount £24,953 (GBP)
Funding ID 130801 
Organisation Innovate UK 
Sector Public
Country United Kingdom
Start 05/2012 
End 04/2013
 
Description Ebb and Flow Energy Systmes
Amount £38,639 (GBP)
Funding ID 131426 
Organisation Innovate UK 
Sector Public
Country United Kingdom
Start 07/2013 
End 10/2013
 
Description Grid Economics, Planning and Business Models for Smart Electric Mobility
Amount £1,005,771 (GBP)
Funding ID EP/L001039/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Public
Country United Kingdom
Start 12/2013 
End 12/2016
 
Description Industrial collaboration with Cenex 
Organisation Cenex
Country United Kingdom 
Sector Private 
PI Contribution The analysis of Electric Vehicles charging events using data mining techniques.
Start Year 2014
 
Description Schneider Electric 
Organisation Schneider Electric Ltd UK
Country United Kingdom 
Sector Private 
PI Contribution I have a Non-Disclosure Agreement with Schneider Electric for using a large sample of electric vehicles charging events for training the Short Term Load Forecasting Algorithm developed in Task 1.
Collaborator Contribution Schneider Electric Ltd UK provided a large sample of electric vehicles charging events that were used for validating the short-term load forecasting algorithm.
Impact The development of short term load forecasting algorithm for electric vehicles charging
Start Year 2012
 
Description Invited Expert to the European CEN-CLC eMobility working group on Smart Charging 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? Yes
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact I am working at an invited expert to the CEN-CLC eMobility working group on Smart Charging. The group produced a report with the purpose of documenting different aspects of Electro Mobility Smart Charging. A group of about 35 experts was formed and divided into 6 tasks-groups (definition of smart charging, link to smart grid, generic role model, reference architecture, charging types/scenarios and final report) to support an efficient working method for the defined scope of work.



The scope of WG Smart Charge report is:
• Define and document a generic role models for different actors and their roles in the domains of E-mobility and power system (Smart Grid)
• Define and document a reference architectures for smart charging of electric vehicles, which will be aligned with the Smart Grid (M/490) reference architecture (correlation with other smart grid functionalities is required in order to maximize system-wide impact and benefits)
• Collect and adapt a set of typical,
Year(s) Of Engagement Activity 2012,2013,2014
 
Description Member of the IEEE P2030.1 working group for the development of standards for electric source transportation. 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? Yes
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact I am participating for the development of the IEEE P2030.1 Draft Guide for Electric-Sourced Transportation Infrastructure as a member of Task Force 2 Grid Impact and Task Force 3 Road Map. .

I am a member of TF2: Impacts on Energy Supply, Transmission, Distribution and Customer Sectors and I contributed in writing this section of the guide.
Year(s) Of Engagement Activity 2011,2012,2013
URL http://standards.ieee.org/develop/project/2030.1.html
 
Description Smart Management of Electric Vehicles 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? Yes
Geographic Reach International
Primary Audience Participants in your research and patient groups
Results and Impact I was invited as a speaker at the "Autonomic Road Transport Systems (ARTS) Summer School, Paris-Marne La Valée.
My talk was extremely well received and the audience asked a lot of questions. This talk triggered captivating conversations afterwards.



After my talk it was established a collaboration with Orebro University, Sweden for the development of a model for integrating the electricity and transport system.
Year(s) Of Engagement Activity 2013
URL http://helios.hud.ac.uk/cost/summerschool.php
 
Description Smart Management of Electric Vehicles Workshop 1 
Form Of Engagement Activity A formal working group, expert panel or dialogue
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Participants in your research and patient groups
Results and Impact The first workshop presented the results from Task 1 Electric Vehicles load forecast.

The industrial participants were from E.ON, UPL, Schneider Electric, KAM Futures and Energy Saving Trust.

The international participants were from NTUA Greece.

The research results outcomes facilitated further collaboration with industrial partners, Schneider Electric, KAM Futures and Energy Saving Trust.
Year(s) Of Engagement Activity 2013
URL http://news.engineering.cf.ac.uk/index.php/events/348-smart-management-of-electric-vehicles-dissemin...