Research Of FRARMA Forecasting Algorithm In Medium And LongTerm Electric Power Load Forecasting

Generally, Long-term electricity load power forecasting is the months and years power forecasting, which is the basement of electricity departments, and supporting data for making sale and marketing plans. The development of long-term electricity load and power forecasting is influenced by some infectors, such as economy, society and climate. Fuzzy clustering can description this influence, but there are many defects. Maximum the fuzzy clustering's advantage and compensate its defect, there will be having meaning both in theory and practice.There are some drawbacks of classical fuzzy clustering algorithm as follow:First, the computing of independent variable weights is unreasonable. Second, the set of horizontal section members is slurred. Third, the computational methods of correlation factor are single. As to compensate for these aforementioned drawbacks, a new algorithm named improved fuzzy clustering algorithm is improving in this essay. The new algorithm uses association analysis to compute the independent variable weights, sets up a method warehouse and uses it to calculation the correlation factors, and selects distinct members of the equivalent matrix as the set of horizontal section. ARMA model is based on the time sequence, the prediction models can describe the dynamic process, the linear modeling sequence must is a long time. the sequence of power has an obvious trend of development, is not a time sequence, though by the high priestess sequence of difference to the tendency to get a stable time sequence, but also eliminate the original sequence of long-term characteristics, causes a lot of information. There are some drawbacks of ARMA model in long-term electrical load forecasting of power system, in view of that, this paper propose the models of RARMA. Firstly, the model bring in regression model warehouse to fit the history data of power. Secondly, use the best fitting model's stable random as the modeling data of ARMA model. Finally, get the RARMA model by combing the two models. The demonstration indicates that the new algorithm increased the accuracy of forecasting result. Further study indicates that, the factors in fuzzy clustering algorithm are stable time sequences. With the ARMA model we can resolve the defects of fuzzy clustering algorithm, and then, improve the model of FRARMA. Based on the study work, all the algorithms are programmed and applied to the system of Liaoning electricity load power forecasting, and the results indicates that all the algorithms are valuable. The model of FRARMA both has theory and practice meanings.