Prediction on characteristic parameters of blasting vibration based genetic BP neural network in coal mine

In order to solve a high danger, many influence factors, discrete measured results of the characteristic parameters and calculation nonlinear problems occurred by the blasting vibrations in the mining area, the establishment on the optimized BP neural network prediction model based on the genetic algorithm(GA) was applied to fit the nonlinear relationship between the mine blasting vibration parameters and characteristic parameters. The model was also applied to the accurate prediction on the characteristic parameters of the mine blasting vibration. The study results showed that in the actual engineering, a GA-BP neural network model could have an important applied value to determine the blasting vibration characteristic parameters and also could save great labor and financial resources. In comparison with the experience formula and BP neural network prediction model, the GA-BP neural network prediction model could have the more strong capacity to solve the complicated nonlinear problem. The predicted value would have a relative error within 10% to the actual value and would not be a minimum value locally. The stability would be better and the average predicted accuracy would be high. From the applied sample number, the model would provide a practical and effective method to the small sample and multi factor parameter prediction and the work load would be small, flexible and wide suitable.