???????????????GM-Markov???????о?

?о???????????????????????????????????????GM-arkov????????????????????PSO?????GM??1??1???????????????????????????????????е???????????н??????????????????????????????????????????????GM-arkov?????GM(1,1????????????б????????????????GM-arkov????????????????????GM??1??1????

STUDY ON GREY MARKOV PREDICTION MODEL FOR OLD GOAF RESIDUAL SUBSIDENCE

On the basis of analyzing the shortcomings of GM(1,1) in predicting fluctuant residual subsidence series of old goaf, a new model for residual subsidence prediction named grey Markov prediction model (GM Markov) was proposed. In this model, particle swarm optimization (PSO) is used to optimize the parameters of background and initial disturbance values of GM(1,1) so that the trend could be predicted and separated fully from the subsidence series; then, Markov chains is selected to correct the prediction value. The model of GM Markov was applied to predict the residual subsidence of an old goaf and the prediction values were compared with that of GM(1,1). Compared with GM(1,1), the result of GM Markov model show good qualities in terms of prediction accuracy and stability.