In view of MOA （metal-oxide arrester） deterioration monitoring， a monitoring method based on an improved hidden Markov model （HMM） is proposed. An improved HMM model used for MOA deterioration monitoring is established by obtaining MOA leakage current data in a substation. According to the MOA leakage current harmonics content of the same bus in the substation， the HMM observation probability matrix is obtained using D-S （Dempster-Shafter） evidence fusion theory and particle swarm optimization. The influence of grid voltage harmonics on the model is studied by MATLAB simulation. The simulation results show that the maximum errors of MOA nonlinear characteristic coefficients a and k are 0.21% and 0.02%， respectively， under the interference of voltage harmonics. It has been proved that the model can counteract the influence of power grid voltage harmonics and improve the online monitoring accuracy of MOA deterioration.
