Based on the well-logging data of typical wells of Zhijin,Panxian and Weining areas in western Guizhou,the well-logging data GR of late Permian coal-bearing strata were processed and wavelet transform technique was used to carry out the sequence stratigraphy division and correlation.The study mainly focuses on the controlling effects which Milankovitch had on high frequency sequence,Milankovitch cycle can be used as a ruler of sequence stratigraphy division and correlation to ensure the scientifcity and the unity of sequence stratigraphy division.According to well-logging signal of the ideal Milankovitch cycle,the corresponding relation between the wavelet scales and the cycles is determined by wavelet analysis.Through analyzing analog signals of subsequence sets to search the corresponding relation between various system tracts and the features of time-frequency,the internal features of wavelet transform scalogram could be made clearly.According to ideal model research,features of Milankovitch curves and wavelet spectrum can be seen clearly and each well can be classifed into four third-order sequences and two system tracts.At the same time Milankovitch cycle can realize the division and correlation of stratigraphic sequence in a quick and convenient way. Based on the well-logging data of typical wells of Zhijin, Panxian and Weining areas in western Guizhou, the well-logging data GR of late Permian coal-bearing strata were processed and wavelet transform technique was used to carry out the sequence stratigraphy division and correlation. The main focus of the controlling effects which Milankovitch had on high frequency sequence, Milankovitch cycle can be used as a ruler of sequence stratigraphy division and correlation to ensure the scientifcity and the unity of sequence stratigraphy division. According to well-logging signal of the ideal Milankovitch cycle, the corresponding relation between the wavelet scales and the cycles is determined by wavelet analysis. Based on analog signals of subsequence sets to search the corresponding relation between various system tracts and the features of time-frequency, the internal features of wavelet transform scalogram could be made clearly. According to ideal model research, features of Mil ankovitch curves and wavelet spectrum can be clearly identified and each well can be classifed into four third-order sequences and two system tracts. At the same time Milankovitch cycle can realize the division and correlation of stratigraphic sequence in a quick and convenient way.