In China, as a major resource, coal has made great contributions to national energy security and social development. The mining of coal resources can cause surface subsidence damage, and in particular, the mining of coal resources in thick loose layer mines is the most serious. How to accurately predict the surface subsidence caused by coal mining in thick loose layer mines has become an urgent problem to be solved. To solve this problem, numerical simulations based on the measured data were used to reveal that the thickness of the loose layer is the intrinsic mechanism that affects the value of the surface subsidence and the large range of subsidence. On this basis, the hyperbolic secant function is used as the influence function of unit mining to derive the expected model of subsidence under thick loose layer conditions: the hyperbolic secant subsidence prediction model. Compared with the probability integral method, the hyperbolic secant subsidence prediction model’s prediction accuracy RMSE value is improved by 38%. The hyperbolic secant subsidence prediction model can realize accurate estimation of the subsidence value in the thick loose layer mine area. This greatly enriches the mining subsidence prediction theory and provides a scientific basis for the assessment of surface damage and ecological environment restoration after coal seam mining under a thick loose seam mining area.

The mining of underground coal resources can trigger geological hazards such as subsidence basins, cave-in pits, and step cracks. In China, the probability integral method (PIM), the most popular method for predicting surface movement deformation caused by coal resource mining, has a prediction accuracy that is mainly influenced by both the measurement data (i.e., quantity and quality) from ground movement observatories and the parameter inversion method. To obtain more accurate PIM parameters in the absence of observational data, we propose a combined machine learning model (RF-AGA-ENN)—random forest (RF) extracts the best combination of features as the input layer of Elman neural network (ENN); ant colony algorithm (ACO) and genetic algorithm (GA) are combined (called AGA) for the weights and thresholds of ENN optimization. The results of the study show that (1) the RF-AGA-ENN model is used to obtain PIM values with MAXRE values between 1.94% and 9.18%, AVERY values between 0.98% and 3.98%, and RMSE values between 0.0050 and 0.9632. (2) Compared with the PIM parameters obtained from BP neural network, RF-ENN, RF-ACO-ENN, and RF-GA-ENN models, the PIM parameters obtained from the RF-AGA-ENN model have better stability and accuracy. (3) According to the PIM parameters obtained by the RF-AGA-ENN model, the predicted and measured values of surface settlement at the 11111 working face have a high degree of agreement. In summary, the RF-AGA-ENN model to obtain the PIM parameters has good application value.

In China, gas and oil reserves are very scarce, but coal resources are abundant in the energy architecture, which decides that coal will remain the dominant energy source for a long time in the future. The accurate prediction of the size and extent of surface movement after coal seam mining is of great significance for the safe promotion of production activities in the mine area and the safety of people’s lives and properties in the mine area. The surface movement deformation under thick loose seam conditions indicates the phenomenon of a large subsidence value and influence range. To predict the size and range of surface movement deformation under thick loose layer conditions accurately, a hyperbolic secant model is constructed based on the hyperbolic secant function. For high nonlinearity of the model parameters, the adaptive step fruit fly algorithm (ASFOA) is introduced into the process of solving the model parameters. Simulation experiments are conducted in three aspects: monitoring point antideficiency, antigross error, and parameter stability. The simulation results show that the ASFOA algorithm achieves high accuracy in finding the parameters of the hyperbolic secant model. The hyperbolic secant model was applied to the 11111 working face under the mining conditions of thick loose layer geology in the Huainan mine. The engineering application results indicate that the hyperbolic secant model performs well on the prediction of surface movement deformation under thick loose layer conditions.