Abstract: Because of the non-linear representation of the function model of probability integral method and the correlation between the parameters,there are some problems in mining subsidence prediction parameters inversion of probability integral method such as complicated calculation, easy divergence, easy to fall into local optimum and so on.Fruit flies algorithm with low parameter setting, high calculation efficiency, easy implementation, strong global optimization ability and high precision is used for inversion of mining subsidence prediction parameters based on probability integral method, the algorithm is also improved.With the measured sinking value as the reference, the predicted results are compared with the measured values.The results show that the improved fruit flies algorithm is applicable to the prediction parameters inversion of mining subsidence by probability integral method.It has positive effect on improving the prediction precision of mining subsidence by probability integral method.