期刊名称：Walailak Journal of Science and Technology (WJST)
印刷版ISSN：2228-835X
出版年度：2016
卷号：14
期号：6
页码：463-484
语种：English
出版社：Institute of Research and Development, Walailak University.
摘要：Optimisation of power generation and plant capacity are of primary significance for the improvement of power supplies, cost decisions and economics. This paper develops a robust predictive model for electric power generation and capacity utilisation and integrates the output from predictive models into a multi-objective model. The optimal solution was determined after comparing the performance of a Real Coding Genetic Algorithm (RCGA), Particle Swarm Optimisation (PSO) and Big-Bang Big-Crunch Algorithm (BB-BC). Testing of the proposed model was carried out using data from a Nigerian electric power generation plant with a capacity of about 5 million Megawatt Hours (MWH) to test the presented methodology. Our findings indicate that the Auto-Regressive Integrated Moving average (ARIMA) model adequately predicts the power generation variables and compares favorably with literature results. The RCGA performed better than the BB-BC and PSO algorithms in terms of the quality of solutions for the proposed model. The outcome of this study suggests that computational complexity can be reduced in the evaluation of variables, yet producing a practical, simple and robust model.
Loading...
