In order to study the mechanical properties of fractured dolomite, this paper first uses the rubbing method to count the surface fracture of the dolomite rock samples, obtains the main parameters of the fracture characteristics, and then carries out the uniaxial compression test to obtain the uniaxial compressive strength of the dolomite rock. Then, the Pearson correlation coefficient and the maximum information coefficient (MIC) are used to analyze the fracture characteristics. The three-times polynomial is used to preprocess the original data, and the three-classification strategy is used to improve the logistic regression algorithm to predict the dolomite strength. Finally, the significant order of the influence of fracture characteristics on rock strength is judged by numerical simulation software PFC3D, and the dip angle effect of rock fracture is explained from the perspective of fracture propagation. The results show that: 1) When the regularization coefficient λ = 10000, the algorithm has the highest prediction accuracy and the strongest model generalization ability. 2) The numerical simulation analysis software PFC3D can accurately invert rock failure process and characteristics, and the order of influence of fracture characteristics on rock strength is inclination > length > position. 3) When α = 0° ~ 90°, the peak strength decreases first and then increases. The greater the peak strength of the rock, the more obvious the crushing phenomenon, and the worse the wing fracture expansion ability. When α = 60°, the peak strength is the lowest.