Use this URL to cite or link to this record in EThOS:
Title: An informed long-term forecasting method for electrical distribution network operators
Author: Akperi, Brian Temisan
ISNI:       0000 0004 6425 1888
Awarding Body: Durham University
Current Institution: Durham University
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Northern Powergrid (NPG) is an electrical distribution network operator in the UK servicing Yorkshire and the Northeast of England. Currently they produce long-term eight year forecasts for each substation on the network with an emphasis on an annual maximum demand (MD) figure. The current method used by NPG is thought to oversimplify the problem and does not give enough insight into changes in substation demand. In order to inform their current forecast, the novel CL-ANFIS method uses a combination of machine learning techniques for both forecasting and general insight to the drivers of demand. Also introduced here are novel techniques for determination of MD at NPG and methods for handling load transfer periods. In order to address a problem of this size, a twofold approach is taken. One is to address the drivers of demand such as weather, economic or demographic data sets through the use of statistics and machine learning techniques. The other is to address the long-term forecasting problem with a transparent technique that can aid in explaining the drivers of demand on any given substation. Techniques used include cluster analysis on demographic data sets in addition to ANFIS as a forecasting method. The results of the novel CL-ANFIS method are compared against the current NPG forecast and show how more insight into substation demand profiles can drive the decision-making process. This is done through a combination of using a tailored customer database for NPG and leveraging the information provided by the membership functions of ANFIS.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available