一种测震仪器数据实时接入中间件设计与实现

分析了几种地震数据采集器的网络协议和数据格式,开发了一套独立的仪器数据访问中间件,为实现上位机应用程序与多种仪器的网络通信、各种数据流的统一接收、处理、发布和共享提供了方法和工具。将中间件应用于测震台站数据监控系统中,为运维人员对台站运行质量的跟踪和数据应用提供了方法和工具,为面向仪器的软件工程实施提供了典型示范。

Design and Implementation of the Middleware to Access Realtime Stream of Digitizers

A digitizer is a key equipment in conducting studies on seismic networks; it acquires analogue signals from the seismometer and converts them into digital signals with the AD module. These signals are outputted via the Ethernet interface in the form of TCP/IP flow. Compared with the relational data set, TCP/IP flow has the following characteristics:continuous production, real-time arrival, and unlimited potential amounts. Based on these characteristics, researchers have proposed a method to access the layer between the data source and application, which can continuously and asynchronously collect data from the underlying part and send them to the upper layer. This hierarchical structure decouples the data producer process and consumer process, which improves the robustness and throughput of the whole system. However, the maximum number of concurrent connections that can be supported by the digitizer is usually limited; hence there is a need to develop an integrated data access service, which can provide a unified interface to access all kinds of data. Therefore, in this paper, the architecture design of a data-accessing-middleware is first introduced, and the implementation of each component is then proposed. In the end, the middleware is applied in the seismic station data monitoring system. The producer/consumer pattern is used to support concurrent and asynchronous communication. The whole system is divided into five components:digitizer adapter, stream buffer, stream adapter, data buffer, and data publisher. The instrument adapter establishes socket connections with the digitizer and obtains streams; it then writes the streams into the stream buffer without decoding. The stream buffer is a synchronized FIFO (First In First Out) queue. The stream adapter reads streams from the queue, extracts data, and writes them into the data buffer. The data buffer is also a synchronized FIFO queue, in which the data publisher reads data from the queue and publishes them in two ways:by storing them into the database or by sending them to a Java Message Service (JMS) provider. With this middleware, a system is developed to monitor the data quality of seismic stations. The system can collect real-time seismic wave data and health data from all seismic stations, and then provide these data to users through Web services. The health data include several parameters collected by the sensor inside the digitizer, including power input, power output, power battery, DSP temperature, clock difference, frequency difference, drive free space, and mass positions. When the value of the parameter is beyond the normal range, it can affect the availability of data, thus affecting the effectiveness of the station. The system provides methods and tools to track the data quality of stations for the engineers. Ultimately, this method can improve the efficiency of the maintenance work and promote further use of the accumulated health data.