Outlier detection for control process data based on wavelet-HMM methods

According to the limitation of the principle of outlier detection based on wavelet,this paper proposes an outlier detection method called wavelet-hidden Markov model(W-HMM) algorithm.In this algorithm,the signal is decomposed under some scale,and when the wavelet decompositions of the signal are different from the most other wavelet decompositions,the signal can be seen as potential outlier.Aiming to make further accurate judgement,and by calculating the similarity probability between the wavelet coefficient of this signal and that of normal signal,the final confirming is obtained by using Viterbi algorithm which is applied to HMM.Finally,experimentation and application show the effectiveness and practicality of the proposed detection method.