Short-term power load forecasting based on BAT-BP neural network model
- https://doi.org/10.2991/mmme-16.2016.38How to use a DOI?
- Bat algorithm; BP neural network; Electric power load forecasting
Accurate short-term forecasting of power load has important significance in Safe and stable operation and im-provement of economic benefits of electric power system. The electric power system load forecasting usually adopts BP neural network (BPNN) method, but this method has slow convergence speed, is easy to fall in-to local minimum point and has poor robustness. In order to improve the accuracy of electric power load fore-casting, the BA-BP load forecasting model of bat algorithm optimizing BPNN is proposed. Firstly, for each individual bat containing the BPNN parameters, the individual is encoded in a real number encoding, and the average relative error is set as the fitness function, and then get the Best bat individual by simulating the pro-cess of bats flying, So as to get the optimal parameter of BPNN. According to the optimal parameters to es-tablish prediction model, finally, the performance test is carried out by simulation experiment, the contrast curve of training speed and relative error is obtained, the results proved that the BAT-BP prediction model has a significant advantage over the simple BP neural network, and it can improve the accuracy of the load forecasting results.
- © 2016, the Authors. Published by Atlantis Press.
- Open Access
- This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Cite this article
TY - CONF AU - Jian Di AU - Tao Yao PY - 2016/10 DA - 2016/10 TI - Short-term power load forecasting based on BAT-BP neural network model BT - Proceedings of the 2016 4th International Conference on Mechanical Materials and Manufacturing Engineering PB - Atlantis Press SP - 172 EP - 177 SN - 2352-5401 UR - https://doi.org/10.2991/mmme-16.2016.38 DO - https://doi.org/10.2991/mmme-16.2016.38 ID - Di2016/10 ER -