壓電式力量感測器之適應性低頻特性補償

Abstract

本研究之目的是藉由適應性估測方式補償壓電式 (piezoelectric) 力量感測器的低頻失真，以得到系統的實際低頻受力。相較於先前的研究，本研究的特點在於不改感測器結構設計的前提下，藉由適應性法則估測系統之低頻受力和干擾，其中包含了系統的未知和不確定性部分，並進一步藉由低通濾波的方式，以減少加速度和速度訊號的使用，估測出系統的受力。 實驗平台是由實驗室成員合力設計和組裝之一維線性馬達實驗平台，控制核心使用美國德州儀器公司(Texas Instruments, TI)生產之TMS320C6713 DSP開發板，搭配實驗室成員所自行研發、具備FPGA等IC之擴充子板。於FPGA方面，以VHSIC (Very High Speed Integrated Circuit) 硬體描述語言(VHDL)撰寫編碼器、ADC與DAC等周邊界面訊息處理函式；而在控制法則實現上，以C/C++撰寫控制器程式並由TI提供的Code Composer Studio (CCS)發展控制程式。實驗結果顯示，本研究提出之方法能有效改善壓電式感測器低頻量測失準現象，準確地量出其正確的受力，並且能減少系統複雜度、降低系統所需的成本和空間。

This research presents a scheme to adaptively estimate the applied force on a linear experimental platform. Strain gauge force sensors and piezoelectric force sensors are two commonly used force sensors in modern industry. The piezoelectric force sensors are widely used to measure forces because of their high sensitivity, high stiffness, and a wide dynamic range. However, if the applied force is static or has a very low-frequency, the charge generated by the piezoelectric quartz will gradually decay to zero. So it is unable to measure low-frequency components of an external force. Compared to the piezoelectric force sensors, the strain gauge force sensors, commonly used to measure static forces, are based on the deformation by themselves, so that they can measure both static and dynamic forces. But the strain gauge force sensors also have some disadvantages, such as large size, low stiffness, and tedious calibration. This research proposes an adaptive algorithm for estimating low-frequency force to compensate for the low-frequency deficiency of the piezoelectric force sensors and the disturbances of the system. Different from the traditional methods that modify sensors' mechanical structures, a commercially available piezoelectric force sensor provides force data for an adaptive algorithm which can estimate force precisely, including low-frequency range. Furthermore, by using the low-pass filtering, the requirements of the acceleration and speed signals can be relaxed, so it can reduce the system’s cost and the required space. The experimentation is carried out using TSM320C6713 DSP with a daughter board, that includes FPGA, DAC, and ADC. The FPGA is configured using the VHSIC hardware description languages to implement data acquisition and storage. The Code Composer Studio programs developed by Texas Instruments is an effective tool for developing C/C++ programs for the control algorithm. The sampling rate of DSP is equal to 11 kHz, and a real-time control system is achieved. Experimental results show the effectiveness of the proposed scheme.