Dominant Instability Mode Identification Method of Power System Based on Hierarchical-self-adaptive Oblique Decision Tree
Under the background of the development of new type power systems,post-fault transient power angle instability and transient voltage instability interact with each other.The current study on identifying dominant instability modes using artificial intelligence faces a"black box"issue,as this field is still in its nascent stages of development.Therefore,a new method of dominant instability mode identification based on hierarchical-self-adaptive oblique decision tree is proposed.The feature construction method of sliding window trajectory cluster is used to capture the time series response information of electrical quantities,which improves the training speed,generalization ability and robustness of the model.In the construction of the identification model,the oblique decision tree classification model fused with the decision surface of the support vector machine is used to establish the mapping relationship between the dominant instability mode and the feature quantity,and the nonlinear combination of multi-dimensional features is used to enhance the branch ability of the internal nodes of the traditional decision tree and improve the identification accuracy.In the first stage of the method,the stability evaluation is carried out,and the cost-sensitive mechanism is introduced to reduce the risk of misjudgment of the instability condition.In the second stage,the dominant instability mode is then identified.In addition,a hierarchical-self-adaptive strategy is used to improve the speed of sample identification and further reduce misjudgment.The generated identification rules are characterized by polynomials,which can mine the correlation between multi-combination features and dominant instability modes.The effectiveness of the proposed method is verified by the transient power angle instability-voltage collapse system example and the ten-thousand-node standard example,which provides a new idea for the identification and analysis of the dominant instability mode based on artificial intelligence.
万方数据
维普
