Kernel Feature Identification Algorithm based on Improved Ant Colony Optimization and its Application in Transient Stability Assessment

L. Guan, X. Zhang, and T. Wang (PRC)


Feature selection, Ant colony optimization algorithm, Local search loop, k-nearest neighbor classifier, Security assessment


An improved ant colony optimization algorithm is proposed and combined with the k-nearest neighbor (k-NN) classifier to realize feature selection. Weighted sum of the k-NN classification error and the selected feature dimension constructs the fitness function. A local search loop is designed to remove the redundant or strong correlated features and improves the performance of the ant colony algorithm. Discussions on the adjustment of the optimization parameters are also presented. The feasibility and effectiveness of the proposed algorithm is first verified by a set of artificial test data and then applied to tackle the power system stability assessment problem. In IEEE 10-unit-39-bus system, the proposed scheme obtains well-behaved security-related kernel feature and provides good transient stability assessment performance.

Important Links:

Go Back