FUKUYAMA Yoshikazu
Department Undergraduate School , School of Interdisciplinary Mathematical Sciences Position Professor |
|
Date | 2018/09/27 |
Presentation Theme | Daily Peak Load Forecasting by Artificial Neural Network using Correntropy based Differential Evolutionary Particle Swarm Optimization for Reduction of Engineering on Data Including Untypical Values |
Conference | Workshop on Electric Power Technology and Power System Technology of IEE of Japan |
Promoters | Technical Committees of Electric Power Technology and Power System of Power & Energy Society of IEE of Japan |
Conference Type | Workshop/Symposium |
Presentation Type | Speech (General) |
Contribution Type | Collaborative |
Venue | Nagoya Institute of Technology |
Publisher and common publisher | Daiji Sakurai, Yoshikazu Fukuyama, Tatsuya Iizaka, Tetsuro Matsui, |
Details | This paper proposes daily peak load forecasting by Artificial Neural Network (ANN) using correntropy based differential evolutionary particle swarm optimization (DEEPSO) for reduction of engineering on data including untypical values. When untypical data exist in the training data, estimation accuracy of daily peak load forecasting can be affected by the untypical data. Therefore, engineers have to remove the untypical values in order improve estimation accuracy. It is a heavy burden to engineers. Correntropy has a possibility to solve this problem. The effectiveness of proposed method is verified by comparison with conventional least mean squre (LMS) based ANNs trained using back-propagation (BP), particle swarm optimization (PSO), and DEEPSO. |