摘要
矿井生产具有高动态、快时变、混沌和随机特点,生产过程中的人、设备、环境、管理等海量数据存在多源异构特征使得安全风险评估不确定性增加。使用智能算法对空间生产场景进行模拟建模,对影响矿井安全的因素建立预警指标体系并对指标量化与预处理,通过BP神经网络模型对预处理后的指标数据模型训练与测试,测试结果表明,随着迭代次数的增加,模型的收敛性与准确率极大提高,可以最大限度地提供矿井安全风险评估决策所用信息、为提高煤矿安全监测的效率和监测深度提供技术支持。
Mine production is characterized by high dynamics,rapid time variation,chaos,and randomness.The massive data from human,equipment,environment,and management in the production process have multi-source heterogeneous characteristics,which increase the uncertainty of safety risk assessment.This paper uses intelligent algorithms to simulate and model the spatial production scenario,establishes an early warning index system for factors affecting mine safety,and quantifies and preprocesses the indicators.Through the training and testing of the preprocessed index data model using the BP neural network model,the test results show that with the increase of iteration times,the convergence and accuracy of the model are greatly improved,which can provide information for mine safety risk assessment decisions to the greatest extent.It provides technical support for improving the efficiency and depth of coal mine safety monitoring.
作者
李瑞华
刘汉烨
冯治东
康亚明
安强强
LI Rui-hua;LIU Han-ye;Feng Zhi-dong;KANG Ya-ming;AN Qiang-qiang(School of Information Engineering,Yulin University,Yulin 719000,China)
出处
《榆林学院学报》
2024年第5期92-95,共4页
Journal of Yulin University
基金
陕西省科技厅2023年重点研发计划(2023-YBSF-014)。
关键词
大数据
BP神经网络
矿井生产
预警技术
big data
BP neural network
mine production
early warning technology