基于BP网络的煤层注水效果的预测

Forecasting the Effect of Water Infusion for Coal Seam Based on BP Neural Network

对煤层注水效果的科学预测是调整注水工艺参数的重要依据。由于煤层注水效果受多种因素影响,属于非常复杂的非线性问题,很难用传统的数学方法进行定量评价。针对影响煤层注水效果的诸多地质因素及注水工艺参数,采用具有非线性映射功能的神经网络预测煤层注水效果,通过比较和实例验证,其误差为6.25%,较其它方法精度有明显提高,对于实际煤层注水工程有指导意义。

The scientific forecasting of water infusion for coal seam effect is the important basis adjusting water infusion technological parameter. Water infusion for coal seam effect is affected by many factors, and it is a very complicated non - linear problem, so it is very difficult to be solved by traditional mathematical methods. Specifically for the geology factor and water infusion technological parameter affecting the coal seam water infusion effect, the neural network with the non - linear mapping function is adopt to forecast coal seam water infusion effect , whose error is 6. 25%. With comparison and the examples, the method using in this paper has higher accuracy than others , and can be used for guiding the actual water infusion for coal seam engineering.