Shale reservoir CO2 energization composite fracturing simulation fracture propagation prediction method and system based on machine learning algorithm

The invention relates to a shale reservoir CO2 energization composite fracturing simulation fracture propagation prediction method and system based on a machine learning algorithm. The method comprises the steps that 1, a natural fracture model is generated through Python; 2) establishing a shale reservoir geometric model based on the distribution condition of the natural fractures; carrying out unstructured grid division on the shale reservoir geometric model containing the natural fractures; inserting a global cohesion unit; establishing a fluid-solid coupling model of CO2 fluid based on a Cohesive module; flow of fracturing fluid is controlled; (3) CO2 energy increasing composite fracturing; the problem that a conventional hydraulic fracturing model does not involve phase change, mixed-phase heat transfer and the like is solved, and the conventional hydraulic fracturing model is not suitable for simulating the crack propagation process of CO2 energizing composite fracturing; the problems of long time consu