A python short-term load forecasting model for Northern Ireland’s Electricity Demand

  • A python short-term load forecasting model for Northern Ireland’s Electricity Demand
Internships Summer 2017/18

Proposed Project Title:

  • A python short-term load forecasting model for Northern Ireland’s Electricity Demand

Principal Supervisor(s):
  • Professor Seán McLoone


Project Description:

The EPIC cluster is working with SONI (System Operator for Norther Ireland) on a research project to improve the performance of their short-term load forecasting models. Here short-term refers to forecasting electricity demand a day ahead, which is a key period for planning the purchase of electricity through the Single Electricity Market (SEM).  Large forecasting errors can add significantly to the cost of generation either due to the need to purchase expensive generation at the last minute, or paying for generation that is then not required. It also puts the satiability of the power system at risk. Hence, accurate load forecasting is essential for efficient and stable power system operation.

The performance of existing load forecasting methods used by SONI has been deteriorating in recent years due to the impact of the rapid growth in small scale (un-metered) distributed generation on the network, and in particular small scale (roof-top) photovoltaics. These have the effect of suppressing the perceived load on the power system and are difficult to predict as their contribution is not directly observable or controllable by the system operator.

The overall aims of this summer internship project are: (1) to develop a python implementation of a promising new load forecasting methodology recently developed by the EPIC cluster and; (2) to evaluate its performance on load forecasting data from Northern Ireland as well as a number of international benchmark load forecasting datasets.  Interested students will also have the opportunity to explore machine learning based enhancements to the proposed methodology.


Objectives:
  1. Become familiar with the principles of load-forecasting, and the operation of a number of the regression based load forecasting models developed within the EPIC cluster (developed in Matlab)
  2. Develop a stand-alone python implementation of a promising new sliding window updating methodology so that it can be evaluated in the field
  3. Conduct a number of studies to evaluate the performance of the method using load forecasting data for Northern Ireland.
  4. Evaluate the methodology on number of international benchmark load forecasting datasets to establish if the approach is applicable more generally or is specific to the Northern Ireland context.
  5. Using Python libraries, explore if machine learning approaches can be used to improve the performance of the new methodology 

Academic Requirements:

The scheme is open to all EEECS Undergraduates (apart from students on the BIT degree pathway and students who are due to graduate this summer)

You must be a competent programmer and be willing to self-learn python. An interest in machine learning, and an aptitude for data analysis and algorithm development are also desirable.


General Information:

Each internship will last between 6-8 weeks and will pay a weekly stipend of £200.

Accommodation and travel costs are not provided under this scheme.

Start date: To be negotiated

Duration: 8 weeks

Location: EPIC PHD laboratories, Ashby Building.

Further information available at: http://www.qub.ac.uk/schools/eeecs/Research/


Contact details:

Supervisor Name: Seán McLoone
Address:

Queens University of Belfast
School of EEECS,
Ashby Building,
Stranmillis Road,
Belfast
BT9 5AH

Email: s.mcloone@qub.ac.uk

For further information on Research Area click on link below:

http://qub.ac.uk/research-centres/EPIC/