Abstract: There are problems in inversion of mining subsidence prediction parameters such as low accuracy, large calculation amount, and complicated process.Therefore, how to accurately and quickly invert the probability integration prediction parameters has become a hot and difficult problem in mining subsidence prediction.Bat algorithm (BA), as a meta-enlightenment optimization algorithm, has the advantages of simple algorithm, fewer parameters, fast calculation speed, high accuracy of parameter seeking, and easy programming.Therefore, this paper applies BA to the inversion of mining subsidence prediction parameters, constructs the minimum cumulative sum of the absolute value of the difference between the actual measured value and the predicted value of subsidence and horizontal movement as the standard fitness evaluation function, and constructs the probability based on BA Integral expected parameter inversion model.Firstly, verify the accuracy and stability of the BA-based probability integral method (PIM) model through simulation experiment analysis, and then use the constructed BA-PIM parameter inversion model to solve the PIM parameters of the 1414 work face of a coal mine in Huainan city.The solution result are q=1.03, tanβ=1.97, b=0.38, θ=86.66°, S₁=-2.51 m, S₂=-13.42 m, S₃=60.36 m, S₄=42.37 m, sinking and horizontal movement fitting median error is 116.34 mm, which is in line with engineering standards.The research results show that the BA-PIM parameter inversion model constructed in this paper has a certain reference value for improving the accuracy of mining subsidence prediction parameters.