Abstract: In order to solve the problems of complexity，large amount of calculating the mining subsidence prediction paramters inversion of the probability integral method,the fruit flies algorithm with the characteristics of simple,low computational complexity and high precision is introduced to the mining subsidence prediction parameters inversion of the probability integral method.The basic principle of the mining subsidence prediction parameters inversion of the probability integral method based on fruit flies algorithm is studied in depth.The fitness function model of minimum mean square of the subsidence fitting values and measured values is established.Based on the mining subsidence measured data of a coal mine in Anhui province,the mining subsidence prediction parameters inversion of the probability integral method are conducted by using the fruit flies algorithm,genetic algorithm and particle swarm algorithm respectively,the mean square error of the subsidence fitting values and measured values is taken as the evaluation criteria of the above three algorithms,the comparison results show that the mean square error of the subsidence fitting values and measured values is 33.7 mm,the relative mean error is 1.4%,which are lower than that of the genetic algorithm and particle swarm algorithm,therefore,it is further indicated that the fruit flies algorithm is suitable to conduct mining subsidence prediction parameters inversion of the probability integral method,it has some reference for improving the mining subsidence prediction precision of the probability integral method.

Key words: Mining subsidence, Fruit flies algorithm, Probability integral method, Parameter inversion, Genetic algorithm, Partical swarm algorithm