基于GA-BP神经网络的矿井突水判别数学模型研究

Research on Mine Water Inrush Discriminating Mathematical Model Based on GA-BP Neural Network

针对传统矿井突水判别方法存在错误率高、精度低以及稳定性差等问题,提出了一种基于GA-BP神经网络的矿井突水判别数学模型。首先矿井突水判别数学模型采用6种特征离子作为判别因子,然后采用遗传算法(GA)对BP神经网络的权值和阈值进行优化,进而构建GA-BP神经网络矿井突水判别数学模型。以某实际矿区的多种水源样品为例,将其作为判别模型中的训练样本,GA-BP神经网络模型对其进行训练,实现待测水源样本的判别。结果表明,判别模型具有更高的突水判别精度和可靠性。

In view of the problems of high error rate,low accuracy and poor stability in the traditional mine water inrush discrimination method,a mathematical model of mine water inrush discrimination based on GA-BP neural network is proposed.First,six characteristic ions are used as discriminant factors in the mathematical model of mine water inrush discrimination.Then genetic algorithm(GA)is used to optimize the weights and thresholds of BP neural network,and then a mathematical model for identifying mine water inrush from the GA-BP neural network is constructed.Taking a variety of water source samples in a practical mining area as an example,it is used as a training sample in the discriminant model,and the GA-BP neural network model is trained to realize the discrimination of the water source samples to be measured.The results show that the discriminant model proposed has higher accuracy and reliability of water inrush detection.