Parameter inversion of probability integral method based on improved crow search algorithm

Coal mining can cause surface movement and deformation, which can lead to serious geological disasters. The prediction of mining subsidence is not only the basis of evaluation of mining to damage, but also the prerequisites for disaster prevention and reduction. In China, the most widely used subsidence prediction method is the probability integration method. Whether the prediction parameters can be obtained accurately or not is the key to the application of probability integration method in subsidence prediction. In this paper, a probability integral parameter inversion model based on improved crow search algorithm (ICSA) is constructed. The research results show that the probability integral parameters obtained based on the ICSA inversion model have high accuracy, and the relative error of the parameters is less than 1.9%. The inversion model based on ICSA has the ability to resist gross error, Gaussian noise, and missing points. In addition, this paper creatively analyzes the sensitivity of probability integral parameters and the influence of prior data on the model and provides constructive suggestions for its application in engineering practice. The probability integral parameters of 1414 (1) working face in Guqiao Mine, Huainan, China, are inversed by using the probability integral parameter inversion model based on ICAS.