Sensitivity Analysis Of Power Network Planning Based On Probability Box

With the power system construction and power system operating environment have been expanding and increasingly complex, network planning programs need to consider more factors, such as geography, culture, and environment. Increasing influence factors makes network planning programs to include more and more of the indicators, and some indicators includes many uncertainty factors so that network planning becomes uncertain attribute decision making problems. More and more uncertain data generated in the environment of uncertainty indicators, in order to prevent network planning decision-making process deviation, it is necessary to make the identification of key indicators, the need for more indicators to make screening. Filtering a program of minimal or no effect on the calculation of indicators and impact indicators, in this case, the use of only certain models and methods of decision-making program is not enough. Study the sensitivity of the index to improve the plan of the power grid becomes very necessary.Existing relevant research areas already have some engineering decisions related to methods and network planning, most of which is to improve the methods and did not fully consider the uncertainty indicators, the indicators expressed uncertainty will result in inaccurate results greater impact on the value of the program indicators to determine the error, so that the final plan is not easy to detect the presence of hidden dangers. To solve these problems, the paper with uncertain characteristics of the model parameters as a starting point, it has the characteristics of in-depth analysis and study of the appropriate sensitivity analysis algorithm. During the algorithm research, according to the physical meaning of indicators and data has the characteristics of a proposed uncertainty species sensitivity analysis model based on the probability box, and then punching the p-box to get the result.Firstly, the power system planning decisions mainstream method of in-depth inquiry to further understanding of various algorithms for solving computing process based on their ideas, clarify the analysis and calculation of these programs is a necessary condition for sensitivity analysis, and are described from the viewpoint of probability applied decision analysis project sensitivity analysis and the importance of sensitivity analyzes. Next, the paper generally uncertain data sensitivity analysis method was studied by analysis of traditional uncertain data and sensitive analytical methods, such as derivation, linear programming, calculus and other principles, the traditional sensitivity analysis of the data is not random certainty and uncertainty of different cognitive processing methods summarized. Traditional sensitivity analysis indicates the presence of data uncertainty imperfect, incomplete and other issues, proposed using probabilistic representation of uncertainty data box, and sensitivity analysis for new ideas. Probabilistic box data modeling uncertainty, not only can solve the problem of data having random uncertainty can also handle data having cognitive uncertainty, combined with the nature of the case and the probability of project sensitivity analysis physical meaning analysis, we propose a method based on sensitivity analysis reduce the probability cartridge. Finally, a numerical example, the paper presents a method used in the sensitivity analysis network planning, the establishment of index probability boxes, sensitivity index obtained by the reduction of the probability of the cartridge, and the sensitivity of conventional experimental comparison algorithm, the algorithm effective verification papers sex.