Skip to content

Application of Machine Learning Techniques in Increased Sustainability for Residential Energy Usage S&T14

  • School: School of Science and Technology
  • Study mode(s): Full-time / Part-time
  • Starting: 2022
  • Funding: UK student / EU student (non-UK) / International student (non-EU) / Fully-funded

Overview

NTU's Fully-funded PhD Studentship Scheme 2022

Project ID: S&T14

As climate changes we observe modifications in the energy consumption which not only impact the energy grid but also our comfort. The shifting energy demand patterns create new challenges and opportunities for utilities trying to balance supply with demand within the smart grid. These are accentuated by the increasingly use of renewables which provide an additional factor of uncertainty in the balancing equation. Traditionally, it has been addressed through demand response in which energy curtailment is achieved at the customer end of the distribution network through price incentives, automated control, or voluntary involvement. The success of these techniques was often overestimated by theoretical models mostly due to the limitations in performing real-time monitoring and control of customer smart appliances. As 5G gets introduced these limitations will no longer have an impact but other challenges related to data analytics and the use of on-premises renewables will arise. As customer demand and comfort is affected by the use of both inhouse renewables and utility-provided energy, energy management will have to factor in the increasingly uncertainty in the forecast of customer-side energy for personal use as well as the overall demand across the distribution grid.

The proposed studentship will investigate novel machine learning ensemble models for incorporating personal renewable energy generation into the variable customer demand forecast to ensure accurate predictions based on a series of factors such as customer comfort and local weather (wind, solar). The scalable models will be integrated at the edge of the network as part of the smart home to enable near real-time forecasts based on IoT devices for monitoring solar and wind on premise. These predictions will be integrated in the home energy consumption and sent to (simulated) utilities for demand response actions. In addition, the recommendation system will provide tenants with advice on consumption peaks by including available renewables to ensure the drain from the distribution grid is minimized. The successful candidate will have experience in applied machine learning and IoT programming. Knowledge of smart grids is a plus but not mandatory. He/she will develop a working IoT prototype and ML algorithm initially tested within the NTU Clifton Smart Campus and later, if viable, on the Nottingham LoRaWAN IoT network. Research will involve a combination of simulated and real-life data ingested in the working protype.As climate changes we observe modifications in the energy consumption which not only impact the energy grid but also our comfort. The shifting energy demand patterns create new challenges and opportunities for utilities trying to balance supply with demand within the smart grid. These are accentuated by the increasingly use of renewables which provide an additional factor of uncertainty in the balancing equation. Traditionally, it has been addressed through demand response in which energy curtailment is achieved at the customer end of the distribution network through price incentives, automated control, or voluntary involvement. The success of these techniques was often overestimated by theoretical models mostly due to the limitations in performing real-time monitoring and control of customer smart appliances. As 5G gets introduced these limitations will no longer have an impact but other challenges related to data analytics and the use of on-premises renewables will arise. As customer demand and comfort is affected by the use of both inhouse renewables and utility-provided energy, energy management will have to factor in the increasingly uncertainty in the forecast of customer-side energy for personal use as well as the overall demand across the distribution grid.

The proposed studentship will investigate novel machine learning ensemble models for incorporating personal renewable energy generation into the variable customer demand forecast to ensure accurate predictions based on a series of factors such as customer comfort and local weather (wind, solar). The scalable models will be integrated at the edge of the network as part of the smart home to enable near real-time forecasts based on IoT devices for monitoring solar and wind on premise. These predictions will be integrated in the home energy consumption and sent to (simulated) utilities for demand response actions. In addition, the recommendation system will provide tenants with advice on consumption peaks by including available renewables to ensure the drain from the distribution grid is minimized. The successful candidate will have experience in applied machine learning and IoT programming. Knowledge of smart grids is a plus but not mandatory. He/she will develop a working IoT prototype and ML algorithm initially tested within the NTU Clifton Smart Campus and later, if viable, on the Nottingham LoRaWAN IoT network. Research will involve a combination of simulated and real-life data ingested in the working protype.

School strategic research priority

The supervision team is enthusiastic about developing and promoting Sustainable Engineering as a growth research area in the
Imaging, Materials and Engineering Centre, aligned to Zero Carbon.

Entry qualifications

For the eligibility criteria, visit our studentship application page.

How to apply

For guidance and to make an application, please visit our studentship application page. The application deadline is Friday 14 January 2022.

Fees and funding

This is part of NTU's 2022 fully-funded PhD Studentship Scheme.

Guidance and support

Download our full applicant guidance notes for more information.

Still need help?

+44 (0)115 941 8418