基于嵌入式技术的汽轮机振动信号分析仪

Turbine's Vibration Signal Analyzer Based on Embedded Technology

嵌入式技术在信息采集及处理领域有广泛的应用。论文以嵌入式技术为核心进行了汽轮机振动信号分析仪的设计与开发。硬件组成上,采用基于嵌入式微处理器的多CPU并行结构,分为以ARM处理器LPC2292为核心的嵌入式数据采集单元和以ARM处理器LPC1768为核心的嵌入式数据处理单元,二者之间通过基于双端口RAM的共享内存进行高效数据通信。汽轮机振动信号处理算法以Welch功率谱估计算法为核心,兼顾谱估计的分辨力和方差性能。软件设计采用层次化结构,在嵌入式操作系统μCLinux和嵌入式数据库管理系统SQLite的基础上进行数据采集、处理及分析程序的开发。所设计开发的汽轮机振动信号分析仪在工业现场得到了很好的应用,实时性和谱估计质量均能满足工业现场对汽轮机组进行故障检测与诊断的需求。

Embedded technology has been widely used in the field of information acquisition and processing. In this paper,turbine＇s vibration signal analyzer was designed by using embedded technology. In hardware design,the parallel structure of multiple CPU was used and the whole analyzer was divided into two parts：embedded data acquisition unit which takes ARM processor LPC2292 as the core and embedded data processing unit which takes ARM processor LPC1768 as the core. The efficient communication between the two parts was achieved by shared memory based on dual port RAM. Welch method was use as the core algorithm of the power spectrum estimation of turbine＇s vibration signal,in which both resolution performance and variance performance can be ensued. Hierarchical structure was used in software designing. On the basis of embedded operating system μCLinux and embedded database management system SQLite,the data acquisition,processing and analyzing program was developed. The designed vibration signal analyzer has been used in industrial site and the real time capability and power spectrum estimation quality can meet the demand of fault detection and diagnosis of turbine unit in industrial site.