Research on prediction system for rockburst based on artificial intelligence application methods

Based on theoretical analysis and on-the-spot monitoring methods, a prediction system for rockburst consisting of long-term and short-term predicting models was proposed. The long-term predicting model adopted a wavelet neural network predicting model by using the rockburst materials of underground projects at home and abroad, so as to forcast the trend of rockburst. In the short-term prediction model, a wavelet neural network model based on the Acoustic Emission(AE) monitored was established to forecast the future AE firstly, and then a catastrophe prediction model for rockburst was founded based on AE forecasted in order to forcast the rockburst near the monitoring site accurately. The two models both used wavelet neural network theory, and can enhance the rate of convergence and fault-tolerant capability, and assure the effects of prediction. A practical example showed that the prediction system has high accuracy, and the prediction results accord with the field performances.