Impacts of observation line layout morphology and survey point missing on the inversion of predicted parameters of mining subsidence

Objective Accurate prediction parameters for mining subsidence are important basis for full cycle green coal mining, while inversion based on the subsidence data of observation lines serves as the main method to obtain these parameters. Methods To quantitatively analyze the impacts of the observation line layout morphology and survey point missing on the inversion of prediction parameters for mining subsidence, this study developed a probability integral model-based inversion method for these parameters using the chimpanzee optimization algorithm (ChOA). Using this method combined with digital simulation experiments, this study predicted subsidence parameters under six observation line morphologies and survey point missing in three different areas: the maximum subsidence area, the boundary area, and the inflection point area. Furthermore, this study revealed mechanisms for the impacts of these observation line morphologies and survey point missing on the parameter inversion results. Results and Conclusions The results indicate that the inversion using ChOA yielded accurate subsidence prediction parameters, with the mean square errors of subsidence coefficient q, influence angle tangent tanβ, mining influence propagation angle θ₀, and average inflection point offset s₀ being below 0.01, below 0.04, about 1.0, and above 10 m respectively. The changes in the observation line layout morphology produced minor impacts on θ₀ but greatly influenced q, tanβ, and s₀. In the case of non-standard observation line morphologies, relying solely on the parameter inversion method may lead to the distortion of inversion results. In the case of the insufficient mining of a mining face, the survey point missing in the maximum subsidence area had slight effects on tanβ and θ₀. However, with an increase in the number of missing survey points in the maximum subsidence area, information on the maximum subsidence gradually shrank, which led to gradually decreasing q and s₀. The survey point missing in the boundary area had small impacts on parameter inversion results but influenced the determination of the movement range of subsided basins, along with angular parameters such as boundary and movement angles. In the case where missing survey points in the inflection point area accounted for less than 40%, the survey point missing posed minor impacts on the parameter inversion. Otherwise, with an increase in the number of missing survey points, the subsidence curve morphology was out of control, q and s₀ gradually decreased, while tanβ gradually increased. Given the strong correlations among these parameters, when the minimum sum of squares of prediction residuals acts as the criterion for the fitness function, it is feasible to reduce the impacts of changes in the observation line layout morphology or survey point missing on the parameter inversion results by narrowing the parameter optimization ranges and using interpolation.