Use this URL to cite or link to this record in EThOS:
Title: Modelling long-term primary energy mix in the electricity supply industry through genetic algorithm based optimisation
Author: Silverton, Charles Lawrence
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2000
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The United Kingdom's Electricity Supply Industry (ESI) was commercially restructured when it was privatised in 1990. Its long-term future now depends upon the actions of competing companies rather than the political decisions of a nationalised industry. Existing models of the industry have not included these market effects as the added complexity has proved difficult to solve. A new type of model is needed to understand the operation and enable forward planning in the ESI. There are many approaches to forecasting ranging from individuals' opinions to mathematical iterations and, more recently, artificially intelligent techniques. Each of these methods have a place in different modelling environments as each has different characteristics. The thesis of this study suggests that forecasting the fuel mix in the ESI is a large non-linear problem that may only be solved by a Genetic Algorithm (GA) based model. GAs use a combination of selection, breeding and mutation to evolve an optimum solution from a population of possible solutions. This thesis report shows how a global utility function reduced the large set of non-linear equations, that described the ESI, into a single optimisation problem solved by a GA. The GA made repeatable optimisations allowing reliable forecasts of different possible future scenarios. The model was further improved by the inclusion of new genetic operators that reduced volatility and gave the GA a memory of previous generations. The model was validated by matching an ex-post forecast with actual past data. It was then used to analyse the ESI's sensitivity to changing environments. This was achieved by building a picture of the future environment from the combined results of multiple scenario forecasts. Although there were politically sensitive outcomes to some scenarios, electricity generation met demand in every case.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available