It is important for accomplishing online monitoring and risk prediction of flashover from contaminated insulators to develop efficient algorithm for identifying surface charge state of insulators.Following the analysis of surface discharge character

a novel discernment algorithm is proposed

where the discernment process contains three steps: the discharge states are discerned based on leakage current via Bayes discernment algorithm; the stages are discerned based on humidity through fuzzy theory; the results from the first and two steps are combined by an arbitration operation and the final result about the discharge state is achieved. The validity testing results show that the novel algorithm is endowed with a good universality and the algorithm trained out by one group data enables to identify the discharge state of the other artificial test with good accuracy and the mistake happens within the adjacent states. So the algorithm is suitable for discerning surface discharge state of HV insulators through LC and humidity signal.