【目的】准确的采煤沉陷预计参数是实现全周期绿色开采的重要基础依据，基于测线沉陷数据进行反演是获取上述参数的主要手段。【方法】为定量分析测线布设形态与测点缺失对采煤沉陷预计反演的影响，在基于黑猩猩优化算法构建概率积分模型沉陷预计参数反演方法的基础上，结合数值模拟实验反演得到6种测线形态和3个不同位置(最大下沉区域、边界区域和拐点区域)测点缺失时的采煤沉陷预计参数，并揭示其对参数反演结果的影响机理。【结果和结论】结果表明：采用黑猩猩优化算法反演的参数精度较高，下沉系数q的中误差均不超过0.01，影响角正切值tanβ的中误差不超过0.04，开采影响传播角θ₀的中误差约为1.1，平均拐点偏移距s₀的中误差优于10 m。观测线形态改变对θ₀影响较小，但对q、tanβ和s₀影响较大，当观测线布设成非标准形态时，单纯依赖参数反演方法可能导致反演结果的失真。当工作面为非充分采动时，最大下沉区域测点缺失对tanβ和θ₀影响不大，但随着最大下沉区域缺失测点的增多，最大下沉信息含量逐渐减小，q和s₀会逐渐减小。边界区域测点缺失对参数反演影响较小，但会影响下沉盆地移动范围及边界角、移动角等角量参数的确定。拐点区域测点缺失占比不超过40%时，测点缺失对参数反演影响较小，但拐点区域测点缺失占比超过40%时，随着缺失测点的增多，曲线形态失去控制，q和s₀会逐渐减小，而tanβ逐渐增大。由于采煤沉陷预计参数间具有强相关性，在适应度函数准则为预测残差平方和最小的约束下，当测线形态改变或测点缺失时，可通过缩小参数寻优范围或插值方法削弱其对参数反演结果的影响。

[Objective] Accurate subsidence prediction parameters prove to be an 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 subsidence prediction parameters, 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 obtained the subsidence prediction 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.