On Load Forecasting,cluster Analysis And Modeling For Tou Models

Since the 70's of 20th century, Demand Side Management as a subject extensively studied by scholars, and the TOU(Time of Use) important model because of its economic value became one of its top priority. The rapid development has made on the TOU, study contains two major parts with the design work and realization. The part of realization mainly contains of two algorithms, one of which is accurate for the prediction of power load data, because all aspects of equipment and conditions are not satisfied with the model data to processed in time; second is cluster analysis algorithm, with it a good time can be divided into the TOU hours for processing. Design means to make the precise time and reasonable price model, to develop a reasonable price in order to obtain the largest economic effects.In this paper, that is, around the above issues, on the following three key questions:(1) The second chapter is to solve power load forecasting data issue. Using the raw data, predict the future electric load data. This article was selected to build the model to approach, the investigation on the BP network based on intelligent algorithm combines for the two main purpose which are eliminate the hidden node BP network and to select the rapid speed of BP network's convergence issues, to improve the precision of the load data and to obtain more accurate data of power load forecasting.(2) The third chapter is to explore the issue of cluster analysis algorithm. Clustering from a number of diverse methods, pick up the simplicity of programming cluster analysis methods, have been given the power load forecast data by multi-objective cluster analysis to determine each time region. Contrast the past improved method the new fuzzy clustering algorithm methods reducing the cycle number, and more conducive to programming, get the precise results of cluster analysis, a clear peak, flat, valley region of the time.(3) The fourth chapter is to explore the issue of time-price model. In this paper, based on reasonable assumptions, on the premise of power load price elasticity of electricity demand and proportion of peak electricity price than the allay electricity price to calculation of the parameters from the TOU model, to obtained a good prediction with emulator. The model can calculate the profitability of power companies, the electricity through various group settings, to select the most profitable that the optimal combination TOU. The model of TOU played an important reference role in to power companies.On the three key issues, time-price model is achieving from design to solve a series of key issues. However, through the MATLAB simulation, the model achieved satisfactory results.