Study On Fragmentation Forecasting For Length Hole Bench Blasting

The fragment-size distribution of rock blasting is an important index to evaluate the blasting quality quantitatively. It affects the efficiency of each follow-up producing process such as loading, transport, fragmentation and the total cost in mining. Therefore, it has momentous practical significance in making a thorough study on the fragment-size distribution of rock blasting. The fragment-size distribution of rock blasting is determined by many uncertain factors, it is difficult to use simple mathematical formula to express the relationship between the factors and fragment-size distribute. It not enough only relies on experiential theory and formula generalized from practice to resolving the uncertain problem in blasting.Artificial neural network technology has a long history, in recent years it has been very rapid. In many fields it has broad potential applications, such as forecast, pattern recognition, automatic control, other areas intelligence simulation and information processing. Artificial Neural network technology has notable merit, such as large-scale parallel processing, distributed storage, automatic adaptability and fault-tolerant, it can resolve effectively the hard precise modeling, highly nonlinear and uncertainties between influence factors and fragment-size distribution.Analyze and study the actual date in Shenjiahu highway H8 slope-cut blasting, first of all, get out the main influence factors: explosive specific charge, characteristic fragment of natural rock, spacing, toe burden, burden to spacing ratio, homogenize index of natural rock before blasting;Then, sum up hidden law between the main influence factors and fragment-size distribution by artificial neural network analysis, guide and forecast the fragment-size distribution more precise. The results showed that the largest error was only 0.47 between forecast value and factual true value;the other errors are 0.2%, which fully meet the application requirements of the project.Artificial neural network can build the dynamic and nonlinear relationship between the main influence factors and fragment-size distribution, which has betterfault tolerant. If the collected study sample form practical project is reprehensive, precise, then select the right network calculate parameters;it is a feasible and accurate study method to forecast the fragment-size distribution by artificial neural network.