The phase division of the degradation state of the AC contactor is an important prerequisite for appliance self-perception. Therefore

the phase division problem is explored considering its features of full-life multi-state and multi-attribute interactions. Based on the original multiple feature parameters

the Wasserstein distance is used for the feature parameter trend enhancement processing. The enhanced features are weighted and fused to establish the degradation state time series model based on the improved weighted TOPSIS method. On this basis

the degradation state time series are analyzed in terms of stage division and whole-life state distribution by using the Pelt algorithm

taking into account the influence of external data interference. The experimental validation shows that the method of this paper is suitable for the research of AC contactor state degradation and division. Compared with the other two methods

the proposed method exhibits an obvious advantage in state classification accuracy

which lays a theoretical foundation for the precise assessment and intelligent application of degradation state.