Abstract:

In order to improve the prediction accuracy of mining subsidence, a method based on improved genetic algorithm and Holt-Winters model is proposed. Improved genetic algorithm (IGA) is put forward due to the defects of genetic algorithm. With the aid of the improved genetic algorithm, parameters of Holt-Winters model can be greatly optimized. Then the IGA-Holt-Winters model is applied in a mining subsidence forecast of Changchun-Siping highway. The result shows that the improved genetic algorithm enhances convergence speed and precision of the algorithm. Furthermore, the convergence of method improve forecast accuracy with percentage error less than 2% and mean errors less than 0.79% for longterm prediction of the mining subsidence. The model has better prediction accuracy and can be used for long-term prediction of mining subsidence.

Key words: improved genetic algorithm, niche selection technology, Holt-Winters model, mining subsidence forecast, ground subsidence