Index Uncertainty Modeling in Grid Planning Based on Probability Box Theory

As grid planning requirements are increasingly refined, the establishment of scientific and reliable grid evaluation indicators has become a matter of urgency. Previous research has mainly focused on the optimization problem of decision indicator systems after the indicators have been obtained, and has ignored the uncertainty of the indicators, which affect the reliability and economy of the entire grid plan. To consider the uncertainty of the indicators, we introduce the superiority of the probability box in dealing with uncertain information, and then use the probability box to model different types of indicators. Finally, using the uncertainty associated with obtaining load indicators as an example, we process sample data by constructing a probability box and by the traditional method, and compare our results with those of the support-vector-machine modeling method. The result shows that the modeling method based on probability theory has great significance for improving decision-making planning.