實現於STM32微處理器之超音波風速量測系統設計及時差演算法開發

Abstract

風速為大自然環境中舉足輕重的一個重要的物理數據，實際風速資料對於環境風場工程更為重要。而欲長時間記錄、監測某個地點之風速及風向，再將數據計算處理，精確和穩定度是首要考量。 相較於傳統風速計，(超)音波式風速計有著較快速、精確、耐用、及惡劣環境適應性之優點。因此，本論文以STM32微處理器作為架構核心，設計高精度的風速量測系統；並研製數種時差演算法，達到精密時差計算。其中較為不同的為「傅立葉相位法」，傅立葉相位法主要將時域訊號的包絡進行傅立葉轉換至頻域，透過計算頻域特定點的相位，回推時域的時間延遲關係，進而跳脫時域限制。 本論文分為三階段進行。第一階段為機構繪製及時間差演算法設計，機構設計主要目的為穩定架設換能器，以及避免波的近場效應及干涉。時間差計算部分，透過嘗試多種演算法以得到高精度的時間延遲資訊。在此階段以訊號產生器產生發射波型，示波器擷取發射和接收波形，將資料於電腦上以Matlab撰寫演算法並分析。 第二階段主要為自動發收系統設計，主體開發以STM32系統之微處理器為核心，盡可能地使用內建模塊控制發收運作，以及演算法實現，進而取代訊號產生器以及示波器，達到低成本、便利且穩定的系統實現。 第三階段為3D印表機設計機構及二維發收系統設計。機構部分，先以Solidworks繪製三維模型，後以3D印表機印出塑料機構，以便於傳感器架設。電路部分，加入額外的類比解多工IC，控制四通道的發收運作，最後，進行時延運算和向量合成，得到風速及相角資訊。

Wind speed is a significant physical data in the natural environment. The re-al-time wind speed data are very important for the wind field engineering. In order to record and monitor the wind speed and direction of a place, accuracy and stability are the primary considerations. (Ultra)sonic anemometer is suitable for harsh weather con-ditions and for long-time high-frequency wind speed measurement. Compared to tradi-tional anemometers, it can achieve higher accuracy and durability. Therefore, STM32 microprocessor was used as the core of whole structure to de-sign the high-precision wind speed measurement system. Furthermore, several algo-rithms were put forward to make precise time delay calculation. Among them, the 'FFT phase method' is relatively different one. FFT phase method mainly converts the envelope of the time domain signal to frequency domain. By calculating the phase of a particular point in the frequency domain, the time delay relationship in the time do-main is back-calculated and the time-domain restriction can be escaped. The paper is divided into three stages for research. The first phase is mechanism drawing and the design of time-delay calculation algorithm. The purpose of mecha-nism design is to avoid interference and near-field effects of waves. As for the time-delay calculation part, a variety of algorithms were tried to obtain high-precision time delay information. At this stage, the transmitted waveforms were generated by an arbitrary signal generator. And then the received waveforms were captured and sam-pled by an oscilloscope. After data storing, algorithm was written and analyzed with MATLAB. The second stage is mainly strived for automatic TR [transmit-receive] system design. With the modules on chip, microprocessor has abilities to generate / capture signal, control transmit-receive operation and implement algorithm. Replacing a func-tion generator and an oscilloscope by this system, a low-cost, convenient and stable system could be implemented. The third stage is mechanism printing with a 3D printer and two-dimensional TR [transmit-receive] system design. In the section of mechanism, first, three-dimensional model was drawn with Solidworks and printed by a 3D printer so as to facilitate sensors setup. When it comes to circuit section, an additional demultiplexer IC was added to control four-channel transmit-receive operations. Finally, time delay calculation and vector synthesis were performed to obtain the information of wind speed and phase an-gle.