| 摘要: | 現今橋梁健康安全檢測大多仍屬於有線傳輸，並且同一時間僅對
單一或少量幾個構件進行檢測，檢測所有橋梁構件，需要伴隨大量器
材設備的移動，及多次有線檢測儀器的安裝，相當的費時費力。 因此
本研究研發出一種無線振動量測裝置，裝置上包含了一顆微控制器
(Micro Control Unit,MCU)、 一顆高精度的 ADXL355 加速度計量測鋼
纜加速度、 一個 LoRa(Long Range)做為無線傳輸模組、 SD 卡模組儲
存實驗資料，本儀器體積大約為 15cm*10cm*5cm， 方便攜帶、無須
架設有線線材、 且易安裝於鋼纜上。本研究將儀器架設在真實橋梁鋼
纜上進行現地實驗，其運作模式為平常系統會維持待機狀態，當使用
者發送檢測命令後，系統即開始對鋼索進行檢測，透過量測鋼纜的歷
時加速度資料，並進行快速傅立葉轉換(Fast Fourier Transform, FFT)，並運行自行研發之演算法計算出鋼纜的振態頻率， 透過自動化作業及無線傳輸將特徵頻率回傳給使用者，使用者即可根據弦理論推算鋼纜
張力。在軟體方面，本研究克服了微控制器內存不夠影響快速傅立葉
轉換資料量、與在單核心 CPU 需同時進行加速度資料採樣與儲存的
問題， 在特徵頻率演算法部分也提升了特徵頻率輸出的準確度。 實驗
結果證明了系統的可行性，成功在感測節點完成測量與運算，並成功ii
以無線傳輸方式將特徵頻率回傳給使用者，成功開發出一種可在邊緣
運算且省時省力的鋼纜健康監測系統。;Nowadays, most bridge health and safety inspections are still wired
transmissions, and only a single or a small number of bridge components
are inspected simultaneously. The detection of all bridge components
requires the movement of a large number of equipment, the installation of
several wired detection instruments, and the structure of multiple wired
inspection instruments, which is quite time-consuming. Therefore, this
research has developed a low-power wireless vibration sensing device,
which includes a micro control unit (MCU), a high-precision adxl355
acceleration meter to measure the acceleration of steel cable, and a Lora
(Long range) as a wireless transmission module and SD card module to
store experimental data. The volume of this instrument is about
15cm*10cm*5cm, which is convenient to carry and easy to install on steel
cables. The operating mode is that the system will maintain a standby state.
When the user sends the detection command, the system starts to detect the
vibration on the steel cable. Through measuring the diachronic acceleration
data of the steel cable, performing a fast Fourier transform (FFT), and
running the self-developed algorithm to calculate the vibration frequency
of the steel cable, the eigenfrequency is transmitted back to the user
through automatic operation and wireless transmission, and the user can
calculate the cable tension based on string theory. In terms of software, this
study overcomes the problems of insufficient memory in the
microcontroller, which affects the amount of fast Fourier transform data,
and the need to sample and store acceleration data simultaneously in a
single core CPU. In the eigenfrequency algorithm part, the accuracy of theiv
eigenfrequency output is simultaneously experimental results that prove
the system’s feasibility. The measurement and calculation are completed at
the sensing node, and the eigenfrequency is transmitted back to the user by
a wireless transmission system that saves time and effort. |
