| Power system short term load forecasting is an important part of load forecasting, load forecasting for high-precision power saving Nyingchi operating costs, improve power quality, improve security and stability of the system has important practical significance. In this paper, artificial neural network BP neural network and RBF neural networks, established for the local characteristics of the artificial neural network model. This study is divided into two levels: First level is the BP network model with radial basis function network model training mechanism and how the introduction of factors affect the prediction accuracy of research; second level is the access part of the Lingzhi in the case of measured data grid , select the BP neural network and RBF neural network, working 24 hours on normal to fit the short-term load forecast. According to the order paper, the main work of this paper is as follows:(1) Summarizes the load forecasting research background, significance and current development status, analysis of electric power load forecasting the importance of development in Tibet, the Tibetan influence factors and the load forecast to follow the concept of cover.(2) The artificial of neural network and the basic principle of the development process, focusing on analysis of the BP network model and the RBF network model learning mechanism, influencing factors and the existence of limitations.(3) The regional climate of Lingzhi characteristics and artificial neural network based on BP network and RBF network selection forecasting, the network structure, transfer function and other parameters of a study choice. After a lot of trial and simulation results, the final prediction results prove that RBF network is much better than BP network, and use the results of the analysis, the results show that the method of scientific and practical. Prediction model presented in this paper on the grid Linzhi improved load forecasting has some theoretical and practical significance.
|
PDF Full Text Request