The invention provides a seismic wave phase spectrum disturbance method based on real number wavelet transform. The seismic wave phase spectrum disturbance method specifically comprises the followingsteps of S1, selecting a target seismic wave, a time shift discrete sampling point and a scale discrete sampling point, S2, conducting wavelet transform on the target seismic waves, so that wavelet coefficients can be obtained, S3, grouping according to an adjacent relationship on a scale domain to obtain a secondary scale and a corresponding secondary wavelet coefficient, S4, defining the phase of the secondary wavelet coefficient, S5, modifying the phase spectrum of the secondary wavelet coefficient to generate a new phase spectrum, and inversely deriving a newly generated wavelet coefficient, and S6, performing inverse transformation by using the newly generated wavelet coefficient to generate random artificial seismic waves. According to the method, the random artificial seismic wavesare generated by disturbing the natural seismic waves, so that the generated artificial waves are high in randomness, and the energy distribution characteristics of the natural waves in the frequencydomain are reserved.