超声波传感器
频道(广播)
声学
结构健康监测
计算机科学
数据采集
电子工程
电气工程
电信
工程类
物理
操作系统
作者
Zehua Li,Wendong Xue,Zhenhai Yang,Xin Huang,Xiaofeng Yang,Yishou Wang
标识
DOI:10.1088/1361-6501/adbb0e
摘要
Abstract Structural health monitoring (SHM) technology based on ultrasonic guided waves (UGWs) has become a prominent area of research and application, with SHM instrument development emerging as a key focus. Existing UGW SHM systems typically use the single-actuation-single-reception (SASR) mode and require an external computer for data processing and analysis. This results in a large system size and low scanning efficiency, making it difficult to meet the requirements for airborne applications. In response to the SHM requirements for airborne applications, this paper designs an architecture with field-programmable gate array (FPGA) for synchronous high-speed acquisition and an advanced RISC machine (ARM) embedded system for asynchronous analysis and processing. A core control module for real-time UGW acquisition was developed, featuring four-channel 14-bit 100 MSPS high-speed synchronous sampling. At the FPGA level, a high-throughput first in first out data buffer was designed. It achieves average read and write speeds of 330 MByte s −1 and 1.6 GByte s −1 , respectively. Data is asynchronously transferred to the ARM embedded system via the peripheral component interconnect express bus for analysis and processing, thereby implementing functions with a low-power CPU. The acquisition control module uses the single-actuation-multiple-simultaneous-reception mode, which improves scanning efficiency by up to 4 times compared to the SASR mode used in traditional monitoring systems. While improving system integration and reducing weight, the system’s power consumption is also reduced, achieving a standby power consumption of less than 20 W. Experiments based on the core control module developed in this paper involved constructing an UGW SHM integrated system with 64 piezoelectric sensor channels. System performance tests and damage localization experiments on composite materials were conducted, achieving an error within 2 cm.
科研通智能强力驱动
Strongly Powered by AbleSci AI