As the key technology in smart grid

load disaggregation is important for the tasks such as load forecasting

demand side management and power security. The accuracy of the traditional methods subjects to the dimension of load signatures

the sampling frequency and the stability of load profile. In this paper

a Factorial Hidden Markov Model(FHMM)based load disaggregation method is proposed

which contains the load state disaggregation through extended Viterbi algorithm and the load allocation based on integer programming. The load is modeled as FHMM

and we extend the Viterbi algorithm to solve the FHMM directly. The proposed method is insensitive to the stability and accuracy of power data

so it is suitable for the residential and industrial devices. Meanwhile

through the sensitivity analysis of Viterbi algorithm

the relationship between the optimal states and the disturbance of the observation is established

which is significant for the reliability evaluation of the optimal states. This work is supported by National Natural Science Foundation of China(No. 61473218)and National Key Research and Development Program of China(No. 2016YFB0901904).