Method for predicting burst pressure of supporting cylinder in expansion type drilling liner hanger based on machine learning

A method for predicting burst pressure of a supporting cylinder in an expansion type drilling liner hanger based on machine learning is characterized in that a supporting cylinder burst pressure prediction model considering toughness burst failure caused by insufficient material strength under internal pressure and environmental fracture failure caused by insufficient fracture toughness is established based on the unified strength theory and the fracture mechanics principle; by means of a machine learning model, corresponding mechanical parameters (yield strength, tensile strength, compressive strength and impact toughness) of the supporting cylinder are obtained by inputting geometrical parameters (outer diameter and wall thickness), measured on site, of the supporting cylinder under different expansion rates, and the cost for obtaining the mechanical parameters of the supporting cylinder under the real working condition is reduced; the method can be used for predicting the bursting pressure before and after