The reasons for compressing and transmitting the fault recorded data in extra-high voltage（EHV） substations in time are given. The advantages and disadvantages of the traditional data compressing methods are analyzed. The novel self-adaptive data compressing method based on signal singularity is proposed. The matching search process of wavelet coefficient singularity in every wavelet scale and the self-adaptive threshold method of the maximum scale number are analyzed. The validity of the compressing method is proved by actual fault recorded data from the East China power network.