Abstract: he acoustic and density logging values have been influenced by wellbore collapse and mudstone alteration in Junggar basin. To solve this problem, firstly, the concept of collapse and alteration exponent has been put forward. In terms of the influence extent, the density and sonic logging data are corrected with empirical formula, forward modeling and neural net. Through the application of the above correction method to seismic calibration and wave impedance inversion, it is more accurate in forecasting the phase and energy corresponding relationship between synthetic seismogram and seismic trace near a well, and more accurate in forecasting sand body.