高级检索

基于粗糙集和改进胶囊网络的煤矿智能通风管理方法

苏淑娴, 欧阳名三

苏淑娴, 欧阳名三. 基于粗糙集和改进胶囊网络的煤矿智能通风管理方法[J]. 煤炭科学技术, 2021, 49(7): 124-132.
引用本文: 苏淑娴, 欧阳名三. 基于粗糙集和改进胶囊网络的煤矿智能通风管理方法[J]. 煤炭科学技术, 2021, 49(7): 124-132.
SU Shuxian, OUYANG Mingsan. Intelligent ventilation management method of coal mine based on rough set and improved capsule network[J]. COAL SCIENCE AND TECHNOLOGY, 2021, 49(7): 124-132.
Citation: SU Shuxian, OUYANG Mingsan. Intelligent ventilation management method of coal mine based on rough set and improved capsule network[J]. COAL SCIENCE AND TECHNOLOGY, 2021, 49(7): 124-132.

基于粗糙集和改进胶囊网络的煤矿智能通风管理方法

Intelligent ventilation management method of coal mine based on rough set and improved capsule network

  • 摘要: 煤矿智能通风管理是煤炭开采过程中实现安全保障的关键一环,由于煤矿类型及其开采技术条件复杂多变,矿井通风过程存在众多通风安全隐患,通风系统的可靠性得不到保障,严重影响矿井通风系统的稳定性。为了解决煤矿通风系统存在的通风异常情况下应急决策水平与智能调控水平不高的问题,研究了一种基于粗糙集算法和改进胶囊网络的煤矿智能通风管理系统。该系统采用基于粗糙集的信息约简模型,利用粗糙集算法进行影响指标约简,筛选掉无关影响因素,降低样本数据冗余性,提高数据样本的可靠性与模型的泛化能力;采用基于改进胶囊网络的煤矿通风环境胶囊感知模型,通过卷积重构胶囊神经元组,对数据特征进行采集,建立设备状态感知胶囊和通风环境感知胶囊等模型,由此构建了煤矿井下区域胶囊网络,来实现对矿山通风环境全面感知、智能监控与智慧决策。试验仿真结果表明:基于粗糙集与改进胶囊网络的煤矿智能通风管理系统对煤矿通风安全决策的准确率为89.5%,召回率为83.7%,F值为86.5%,较其他系统准确率提升了4.4%,召回率提升了8%,F提升了6.4%,大幅提高了煤矿通风安全决策准确率,对出现的安全隐患及时预警效果显著,具有信息感知能力强、决策准确等特点,为煤矿井下通风安全提供重要保障。
    Abstract: Intelligent ventilation management of coal mine is a key link to ensure safety in the process of coal mining. Due to the complexity and changeableness of coal mine types and mining technical conditions, there are many hidden safety hazards in the process of mine ventilation, and the reliability of ventilation system cannot be guaranteed, which seriously affects the stability of mine ventilation system. In order to solve the problem that the emergency decision level and intelligent control level of coal mine ventilation system are not high under the condition of abnormal ventilation, this paper studies a coal mine intelligent ventilation management system based on rough set algorithm and improved capsule network. In this system, the information reduction model based on rough set is adopted, and the rough set algorithm is used to reduce the influence index, so as to screen out irrelevant influencing factors, reduce the redundancy of sample data, and improve the reliability of data samples and the generalization ability of the model. Network based on improved capsule of coal mine ventilation environment perception model, through convolution reconstruction capsules group of neurons, the characteristic of data acquisition, perception capsule and ventilation equipment state environmental perception capsule model, such as the capsule network constructed in the coal mine area, to realize the mine ventilation environment comprehensive perception, intelligent monitoring and intelligent decision-making. The experimental simulation results show that based on rough set and improve capsule network intelligent ventilation management system of coal mine ventilation safety in coal mine decision accuracy is 89.5%, the recall rate was 83.7%, the F value of 86.5%, compared with other system accuracy of 4.4%, the recall rate of 8%, the F value increased by 6.4%, greatly improving the mine ventilation safety decision-making accuracy, for the existing security hidden dangers timely warning effect is remarkable, has the characteristics of information perception ability, decision-making accurate, provide important guarantee for the coal mine ventilation safety.
  •   基于粗糙集和改进胶囊网络的矿井智能通风管理系统体系结构

      胶囊神经网络误差曲线

      无线传输与有线传输结合原理

      模型描述层结构

      智慧决策层结构

      粗糙集-胶囊网络煤矿通风安全决策模型

      感知胶囊元模型

      胶囊网络结构

      胶囊神经元向量示意

      粗糙集神经网络误差曲线

  • 期刊类型引用(4)

    1. 翟新献,郭钊洋,方建厂,张志勇,赵晓凡. 综放开采覆岩变形分区和矿压显现模拟研究. 煤炭科学技术. 2025(01): 96-106 . 本站查看
    2. 龚惠春,夏广宁,杨生龙,刘鹏,杨龙. 深埋工作面回采巷道复合顶板水压致裂技术应用. 陕西煤炭. 2024(08): 90-94+140 . 百度学术
    3. 宋高峰,黄鹏,李鹤鹤,昝明惠,孔德中,张鹏飞. 基于能量法的工作面端面冒顶机理及“支架-围岩”耦合关系试验研究. 煤炭科学技术. 2024(07): 11-22 . 本站查看
    4. 巴玉龙,曹民远,刘明,徐源山. 双低位亚关键层破断能量聚散及控制研究. 煤炭工程. 2023(07): 89-93 . 百度学术

    其他类型引用(4)

图(10)
计量
  • 文章访问数:  82
  • HTML全文浏览量:  0
  • PDF下载量:  290
  • 被引次数: 8
出版历程
  • 网络出版日期:  2023-04-02
  • 发布日期:  2021-07-24

目录

    /

    返回文章
    返回