基于形心法的叶尖到达时刻高精度提取方法研究  

Research on High-Precision Extraction Method of Blade Tip Arrival Time Based on Centroid Algorithm

作　　者：周聪 段发阶[1] 刘志博 李健[1] ZHOU Cong;DUAN Fajie;LIU Zhibo;LI Jian(State Key Lab of Precision Measuring Technology&Instruments,School of Precision Instrumeres and Optoelectronics Engineering,Tianjin University,Tianjin 300072,China)

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072

出　　处：《传感技术学报》2022年第7期952-959,共8页Chinese Journal of Sensors and Actuators

基　　金：国家重点研发计划项目(2020YFB2010800);国家自然科学基金项目(51775377);中国航发四川燃气涡轮研究院外委课题(GJCZ-2020-0040,GJCZ-2020-0041);国家自然科学基金项目(61971307,61905175);霍英东教育基金会资助(171055);中国博士后科学基金(2020M680878);广东省重点研发计划项目(2020B0404030001);天津市科技计划项目(20YDTPJC01660)。

摘　　要：基于叶尖定时的叶片振动参数实时监测已成为现代大型旋转机械健康监测领域的主要技术手段,叶尖到达时刻高精度提取是实现叶片振动参数准确测量的关键,由于传统的时刻提取方法不能满足日益增长的振动参数辨识精度需求,提出了基于形心法的提取方法。理论证明形心具备信号强度不敏感及抗噪性能强的优势,可实现叶片到达时刻的高精度提取,仿真实验表明该方法显著优于传统方法。形心法通过高速ADC采集信号,FPGA实时计算其形心,提取叶尖到达时刻。通过DG2052信号源生成叠加白噪声的高斯脉冲信号进行实验,在噪声叠加比例为30%,转速为12000 rpm的条件下,形心时刻鉴别标准差为19 ns,相对传统前沿时刻鉴别73 ns的标准差,减少了74.0%,显著提高了叶尖到达时刻的提取精度。Real-time monitoring of blade vibration parameters based on blade tip timing has become the main technical means in the modern large-scale rotating machinery health monitoring field.The high-precision extraction of blade tip arrival time is the key to the accurate measurement of blade vibration parameters.The tradition method cannot fulfill the increasing demand for the high identification accuracy of vibration parameters,therefore,an extraction method based on centroid is proposed.The centroid has the advantages of insensitive signal strength and strong anti-noise performance advantages,and it can achieve the high-precision extraction of TOA.Simulation result shows that this method is significantly better than the traditional method.During the application of the centroid method signals are captured through high-speed ADC,the centroid is calculated through FPGA in real time to extract the TOA.Gaussian pulse signal superimposed with white noise is generated by DG2052.When the noise superimposition ratio is 30%,the rotation speed is 12000 rpm,the standard deviation of centroid time identification is 19 ns,which is reduced by 74.0%compared with the 73 ns of traditional leading edge identification method,indicating that the extraction accuracy of TOA is significantly improved.

关 键 词：叶尖定时 形心法 FPGA 

分 类 号：V232.4[航空宇航科学与技术—航空宇航推进理论与工程]