煤矿设备状态监测系统设计

Coal mine equipment condition monitoring system desig

  • 摘要: 针对煤矿井下设备群同时作业时设备监测数据高并发导致的低传输效率问题,提出了一种煤矿设备状态监测系统设计方案。该系统通过数据集成网关有效消除传感器网络的异构性:在数据集成网关中对不同传感器进行注册,利用传感器网络协议适配器调用不同传感器网络协议解析接口来消除传感器网络的异构性,生成统一格式的Java Script Object Notation(JSON)数据,并将数据发送到对应的消息推送服务中,通过ActiveMQ消息队列中的Queue通道进行点对点传输,数据传输服务将消息实时推送到网络传输模型中,实现设备状态数据的高并发传输,保证监测数据的实时性和可靠性。采用Netty网络传输模型避免空轮询导致的服务器负载增加,提高监测数据传输效率:在数据采集过程中,多个设备同时作业导致数据采样频率和传感器终端的并发请求数量升高,Netty模型中的Epoll模式优先处理已就绪的I/O连接,从而减少空轮询现象。测试结果表明,随着系统并发请求次数增加,采用Java Non-blocking I/O(NIO)模型比采用Netty模型的系统CPU使用率高28%;在系统并发请求次数相同的情况下,采用Java NIO模型的系统平均响应时间大于采用Netty模型的系统。采用Netty模型能有效提升煤矿设备状态监测系统的高并发处理能力,满足设备监测数据高效传输要求。

     

    Abstract: In order to solve the problem of low transmission efficiency caused by high concurrency of equipment monitoring data during the simultaneous operation of underground equipment groups in coal mines, a design scheme for coal mine equipment condition monitoring system is proposed. The system eliminates the heterogeneity of sensor networks effectively through the data integration gateway. Different sensors are registered in the data integration gateway, and the sensor network protocol adapters are used to call different sensor network protocol resolution interfaces to eliminate the heterogeneity of sensor networks, generate unified format of Java Script Object Notation (JSON) data, and send the data to the corresponding message push service. Through the point-to-point transmission through the Queue channel in the ActiveMQ message queue, the data transmission service pushes messages to the network transmission model in real time to obtain high concurrent transmission of equipment status data and ensure the real-time and reliability of monitoring data. The Netty network transmission model is used to avoid the increase in server load caused by empty polling and improve the efficiency of monitoring data transmission. In the process of data collection, multiple equipment operating at the same time lead to an increase in the frequency of data sampling and the number of concurrent requests from sensor terminals. The Epoll mode in Netty model prioritizes the ready I/O connections so as to reduce the empty polling. The test results show that as the number of concurrent requests in the system increases, the CPU usage of the system with the Java NIO model is 28% higher than that of the system with the Netty model. When the number of concurrent requests in the system is the same, the average response time of the system with the Java NIO model is longer than that of the system with the Netty model. The application of Netty model can improve the high concurrency processing capability of coal mine equipment condition monitoring system effectively and meet the requirements of high efficient transmission of equipment monitoring data.

     

/

返回文章
返回