A Transient Power Angle Stability Fault Screening System and Method Based on Support Vector Machine

Abstract

The invention discloses a transient power angle stability fault screening system and method based on a support vector machine. Based on the quantitative evaluation results of the transient power angle stability of the samples in the historical sample set, the feature quantities are selected to avoid the problem of cumbersome calculation using the feature selection algorithm and the problem of misselecting and missing key feature quantities; Conduct supervised learning training to improve the accuracy of the prediction model; use the regression prediction model with the transient power angle stability margin as a variable for the sample set with only stable samples, to avoid the limitations of only using the classification prediction model; based on the characteristics of the current operating mode Quantitatively judge the applicability of the transient power angle stability prediction model to avoid the "misjudgment" problem of the machine learning algorithm; based on the difference between the detailed transient power angle stability quantitative evaluation results and the predicted value, determine whether the expected fault and its corresponding power flow mode are included in the sample. The number of unstable samples in the sample set is appropriately increased to ensure the rationality of the samples in the sample set.

Description

technical field [0001] The invention belongs to the technical field of power system automation, and in particular relates to a system and method for screening transient power angle stability faults based on a support vector machine. Background technique [0002] Online transient safety and stability assessment has now become an urgent need for dispatching operation of large power grids, and its calculation cycle is generally required to be within 5 minutes. As the scale of the power system increases, the expected number of faults that need to be evaluated online for transient safety and stability will be large, and the evaluation time of a single fault will also become longer. For a large power grid with thousands of generators and tens of thousands of computing nodes, if the expected faults are not screened and the transient safety and stability assessment of tens of thousands of expected faults is completed within 5 minutes, a large number of computing resources need to be...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the deficiencies in the prior art, the present invention provides a support vector machine-based online transient power angle stability predictive fault screening method for power systems, which solves the incompatibility between the correctness and screening efficiency of fault screening in the existing Safety and stability analysis of technical problems with low calculation speed and accuracy

Images