The Study, Based On Load Forecasting Of Electric Energy Intelligent Monitoring And Billing System

This thesis is built on a project named Power Auto Monitoring System of DaQing Refinery. This project is on the first round development now .First of all, this thesis shows the three layers C/S distributed construction of the whole system, and analyzes each part of the system carefully. This construction mode forms three layers as below: client interface layer, the operation logic layer, and the database server layer. This can make the operation logic layer more independent and easier to upgrade.Then, this thesis emphasizes on the study of load forecasting system designed for the project. This load forecasting system belongs to the operation logic layer in the whole system. The method used in the load forecasting is the artificial neural network. By doing fuzzy operation to the sample data, using the LBG algorithm to compartmentalize the sample space, compute the center of each space , using all samples in each sample space to train the neural network according to that sample space, then use those space centers to train a BP network to modify the system. If the system pass the exam, the load forecasting system can run .In the end, this thesis shows a experiment, and give the data and the result of the experiment ,so as to prove the feasibility of the load forecasting system.