Research On Uncertainty Modeling Of Power Grid Planning Based On Probabilistic Box Theory

Network planning is an important part of the entire power system, but a good grid plan would affect operation of the entire electricity market. Currently, the main problem of the grid planning is the diversity and uncertainty indicators, In this paper, on the basis of domestic and international, grid indicators for uncertainty information modeling study based on probability box theory:(1)Grid indicators data subject to distribution, using the law data construction box of the algorithm; (2) Grid indicators data without the law, using the original data construction box of the algorithm; (3) Grid indicators large amount of data, using non-dimensional construction box of the algorithm.First, the paper describes the network planning in the indicators classification and characteristics of uncertainty information, introduces some existing treatment methods, compare their advantages and disadvantages, and explains the advantages of these methods with respect to probability theory box; Then, the paper constructed the network planning indicators architecture, show the diversity and complexity of the power indicators, Illustrates the power indicator accurately or not will have a significant affect the accuracy of network planning programs, It presents three modeling algorithm combines probability box theory building probabilistic box, and load, line loss rate, the voltage passing rate example of the display; Finally, to adapt to changes in load variation using SVM pattern recognition methods through the program, to verify the feasibility of the grid indicators probability box modeling, Then select the actual case, the use of AHP, fuzzy evaluation method of Blind Number and the evaluation method of probability box to make decisions on the program are to verify the probability box credibility and superiority, At the same time, propose a new probabilistic box method in decision-making at the time complexity were optimized.Thesis indicator data processing, instead of doing a simple estimate process, but the power system data underlying the results are collected and after use modeling approach to deal with come, this is closer to the real situation, and can accurately obtain the indicators, so that the network planning assessment rough without scientific issues have been solved.For the problem of indicators uncertainty information in the network planning, Thesis on the grid indicators uncertainty modeling based on probability box theory, and numerical example proved the theory of probability box modeling planning program indicators uncertainties more scientific and reasonable.