Short term load forecasting of anomalous load using hybrid soft computing methods

International Conference on Innovation in Engineering and Vocational Education

自然科学;教育

Rasyid, S.A.^1 ; Abdullah, A.G.^1 ; Mulyadi, Y.^1

Faculty of Technology and Vocational Education, Universitas Pendidikan Indonesia, Jalan Setiabudhi 229, Bandung

40154, Indonesia^1

关键词: Electrical energy;    Electricity companies;    Electricity demands;    Generation cost;    Hybrid soft computing methods;    Load forecasting;    National electricity markets;    Short term load forecasting;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/128/1/012007/pdf
DOI  :  10.1088/1757-899X/128/1/012007

来源: IOP

【 摘 要 】

Load forecast accuracy will have an impact on the generation cost is more economical. The use of electrical energy by consumers on holiday, show the tendency of the load patterns are not identical, it is different from the pattern of the load on a normal day. It is then defined as a anomalous load. In this paper, the method of hybrid ANN-Particle Swarm proposed to improve the accuracy of anomalous load forecasting that often occur on holidays. The proposed methodology has been used to forecast the half-hourly electricity demand for power systems in the Indonesia National Electricity Market in West Java region. Experiments were conducted by testing various of learning rate and learning data input. Performance of this methodology will be validated with real data from the national of electricity company. The result of observations show that the proposed formula is very effective to short-term load forecasting in the case of anomalous load. Hybrid ANN-Swarm Particle relatively simple and easy as a analysis tool by engineers.