Contamination grades assessment is very important content for the online monitoring system of insulator leakage current（LC）.However

the assessment is difficult due to the nonlinear relationship between the electric characteristic variables of the LC

the environment factors and the contamination condition of insulator surface.In this paper

based on the simulation experiments in laboratory and the field data

the parameters of support vector machine are optimized by using particle swarm optimization algorithm to establish an SVM pattern recognition model for assessment of the contamination grades.The method takes advantages of the minimum structure risk of SVM and the quick global optimization ability of particle swarm in quickly setting up the mapping relation of contamination grade to R.M.S.of LC

peak value of LC

amplitude and number of LC pulse

and environmental temperature and humidity by learning from sample data.Experimental results show that the contamination grades assessment method is effective.