A knowledge representation model for the optimisation of electricity generation mixes

Energy models such as MARKAL, MESSAGE and DNE-21 are optimisation tools which aid in the formulation of energy policies. The strength of these models lie in their solid theoretical foundations built on rigorous mathematical equations designed to process numerical (quantitative) data related to economics and the environment.Nevertheless, a complete consideration of energy policy issues also requires the consideration of the political and social aspects of energy. These political and social issues are often associated with non-numerical (qualitative) information. To enable the evaluation of these aspects in a computer model, we hypothesise that a different approach to energy model optimisation design is required.A prototype energy model that is based on a semantic representation using ontologies and is integrated to engineering models implemented in Java has been developed. The model provides both quantitative and qualitative evaluation capabilities through the use of logical inference. The semantic representation of energy policy goals is used (i) to translate a set of energy policy goals into a set of logic queries which is then used to determine the preferred electricity generation mix and (ii) to assist in the formulation of a set of equations which is then solved in order to obtain a proposed electricity generation mix.Scenario case studies have been developed and tested on the prototype energy model to determine its capabilities. Knowledge queries were made on the semantic representation to determine an electricity generation mix which fulfilled a set of energy policy goals (e.g. CO2 emissions reduction, water conservation, energy supply security). An optimised electricity generation mix, based on the queried results, was then generated using a custom developed linear programming optimisation module.

► Prototype energy model which uses semantic representation (ontologies).
► Model accepts both quantitative and qualitative based energy policy goals.
► Uses logic inference to formulate equations for linear optimisation.
► Proposes electricity generation mix based on energy policy goals.