Giosprite have developed an Ecosystem of smart network technologies spanning transport, tourism, environmental and social care to help local authorities and businesses gain valuable insight from Internet of Things data quickly and easily.
There is an ever growing desire to increase the penetration of renewable energy sources into our electricity networks to combat climate change. However, the intermittent nature of renewable energy causes a problem when trying to integrate it without compromising the power quality and stability of the grid. The solution may be localised networks (microgrids) with energy storage systems and distributed generation. Microgrids require an energy management system to optimally control energy usage.
This project looks at the use of artificial intelligence and deep reinforcement learning for the energy management system, which allows for the control of energy based on multiple factors, including supply, demand, energy price and weather data, using SEND and Keele Campus as a case study.
The project will use initial data from SEND to train the algorithms based on real-life data. The outcome of the project could be an intelligent energy management system that could be used within SEND to optimally manage energy usage within the campus.
Project progress (June 2019):
- Review of literature on demand response and microgrids and machine learning.
- Prototype agent-based system created using Python.
- Simulations performed based on both simulated energy price and real price data.
- Working to improve using a multi-agent system and incorporating other aspects of demand response.
Business Contact(s): Nick Wilcox
Keele Graduate Researcher: Dan Harrold