A dynamic prediction method for surface mining subsidence based on dynamic probability integral model and Logistic model

In order to accurately describe the dynamic evolution of surface deformation over time caused by underground coal mining, a dynamic probability integral method model (DPIM) was first constructed by combining the probability integral method model (PIM) and the Knothe time function model. Then, taking into account the changing rules of the predicted parameters during mining process, the Logistic model was used to fit the dynamic predicted parameters, and a DPIM-Logistic dynamic fusion subsidence prediction model was constructed. Finally, in view of the highly nonlinear function of the predicted model, the fireworks algorithm was introduced to solve the parameters of the model. Simulation experiments show that the error in subsidence fitting is 2.37 mm and the error in prediction results is 5.74 mm in inverting parameters. This method was applied to a working face in Huainan Mining Area. The error in the inversion parameter fitting was 59.67 mm, and the error in the prediction results was 73.10 mm. It effectively verified the accuracy and reliability of the predicted results.