测控技术与仪器 外文翻译 外文文献 英文文献 高稳定性电容传感器系统及其性能评价

测控技术与仪器 外文翻译 外文文献 英文文献 高稳定性电容传感器系统及其性能 A High-St Capacitance Sensor System and Its Evaluation 译文标题 ability 高稳定性电容传感器系统及其性能评价 原文标题 Svetlana 作 者 译 名 斯维特拉娜 国 籍 美国 Avramov-Zamurovic IEEE TRANSACTIONS ON INSTRUMENTATION AND 原文出处 MEASUREMENT, VOL. 58, NO. 4, APRIL 2009 摘要 一种安装有微控制器，商用温湿传感器，电容容量数字化转换器和一个特制的电容传感器的新型传感器系统诞生了。我们通过仿真和对样机的测试对该系统的性能进行了评价。我们通过对周边环境条件的变化对系统的影响的分析制作了一个温湿补偿器。根据在不受控环境条件下，用1pF标准电容进行实验获得的数据，我们知道系统24小时稳定性是在30/106以内。高稳定性和高灵敏性使该系统在物体探测领域能够得到有效的应用。根据在不同状况下对不同的探测要求，我们设计出了几种传感器并对它们的性能进行了评价。 关键词—电容测量，电容传感器，电场仿真，湿度，温度变化。 I.导言 最近的技术进步提供了输入范围在几皮法之间的高分辨率(24位)Σ?Δ电容容量数字化转换器(CDC)，这些CDCs的潜在的应用领域是非常广泛的。电容传感器与人手操控的用来区分物品种类的机械手的一体化促进了我们的研究实验。在这个应用项目里设计要求的关键之处在于测量所要求的微小尺寸和高灵敏性和可靠性(稳定精度和重复精度)。最近研究出的一种简单的电容传感器用来测量介电常数的变化，以此来作为区分有界目标集内不同物体的标准。Avramov-Zamurovic 和 Lee的关注点是以下几个方面: 共 5 页 第 1 页 1) 使用一个充满氮气的1-pF标准电容来评测CDC。2)制造一个温湿 度补偿器。3)制造一个有效的传感器系统来探测不同状况下包括机 场安检在内的特定物品的探测。 II.电容传感器系统设计 一个先进的电容传感器系统的简图如图1所示。它是建立在控制CDC，环境温湿度传感器和收集测量数据的微控制器之上的。CDC与特制的电容传感器相连，电容传感器安装在目标物体之下。个人电脑用来处理数据和开发微控制器的软件。在这种应用领域里微控制器的扩展包是由美国海军军官学校武器和系统开发系研制开发的。所使用的微型电脑是8位机，频率为30MHz，具有卓越的计算系能。之所以选择这种硬件是因为它具有I2C的信息通讯能力，这对于CDC和传感器的使用是必要的。在本文中展示了微控制器的性能在电容传感器系统领域的应用中现出的特点，更多的微处理器和扩展包的详细信息可以在产品说明书中找到。 在我们的应用中CDC使用单极输入，其输入的变化范围为?4 pF，内部的激励信号为16 kHz。该CDC支持16- 和32-kHz的信号源。所选择的16-kHz的信号用来将CDC的读入与商用的经电容桥处理升高到20 kHz的信号进行比对。CDC是为移动电容传感器而设计，测量电容与地面相隔离。它的对地寄生电容达到了60 pF。使用一个集成了芯片的可编程的电容数模转换器可以消除电容测量过程中最大17pF的偏差。厂家申明直线性(0.01%),有效分辨率(21 bit)，和精确度? (4 fF)必须使用标准的电容来校验，此外还需要确定它的不确定度。这是未来研究的焦点内容。CDC各方面的详细解释请参阅厂家说明， 共 5 页 第 2 页 其与电容传感器系统相关的特点在本文中已给出。 CDC是安装在表面上的电路，安装板上设计有数字接口，用来与微控制器(I2C)连接。特弗纶隔热同轴电缆直接将CDC上的测量输入针脚与Bayonet–Neill–Concelman连接器连接起来。连接器安装在围绕着CDC的地面屏蔽盒上。CDC上有一个在自我校验程序上有应用的集成了芯片的温度传感器。当发现CDC受湿度变化，和在小范围内受温度变化影响显著时，温湿度传感器就被用来监测周围环境条件的变化。来自于温湿传感器的数据被用来评估环境条件的变化造成的电容修正，和用于交叉比对CDC集成芯片的温度读数，这些提高了测量程序的可靠性。 CDC计算出一系列电容读数的平均数。测量结果以c部分全刻度的形式被记录下来。测量中要求使用16 kHz的频率和2.5 V的电压。在每一次测量中应用的内部噪声过滤器和软件周期性的处理CDC的偏差和进行自我校核。厂家指定的最大的电容数据比率是90 Hz，但是将所有的过滤器，温湿传感器读数和自我校核程序的执行都考虑在内，系统的实际数据率为3 measurement/min。 III.电容传感器系统的稳定性和修正模型的评价 在实验室环境里(26 ? 1 ?C)，用一个充满氮气的标准的1-pF的标准电容在非主动热控状态下构建电容传感器系统的稳定性实验。通过应用高精度的商用电容桥测得标准电容的24小时稳定性为? 2/106。 环境湿度和温度的读数被用来建立一个弥补系统对环境条件变化依赖性的模型。通过对湿度对电容传感器的影响的观察发现在湿度 共 5 页 第 3 页 的变化和CDC的读数之间存在一个时滞。湿度时间常数τ被定义为在环境湿度变化的开始和电容读数开始漂移之间流逝的时间。一个动态的窗口用来显示电容和湿度的数据。优质的原型的时间常数据估计为30 min。除去随机噪声，组成电容，湿度，温度读数的数据集的平均数约为3分钟。迅速的突发型的相对于平均值的湿度偏差并不会显著影响电容的读数，约为30 min的增益平均值在湿度数据的处理中得到了应用。在电容修正之前的预处理进程(时间常数，噪声过滤，增益平均)要求大约1小时的时间。 IV.应用和测试结果 在本文中，我们叙述了应用电容来探测一定距离处不同目标物体的设计。探测器的设计根源于对交叉电容原理的更改。在测量中电容变化的高稳定性已经实现了。一些铜电极被固定在特弗纶支撑上，相隔0.47 cm。传感器的底部接地，同轴电缆用于测量电容。原型传感器被严格的建立了起来，类似的传感器可以用柔性的印制电路板来构造，使它适用于其他方面的应用。在设计阶段，运用基于筛分实验的aMAXWELL 3-D v4.1电气现场仿真软件对传感器的拓扑结构进行了分析。在仿真电容评估和实际测量之间，精确性方面存在的6%的差异要归因于仿真和原型之间存在的几处差异。测量是在一个巨大的接地盘上进行的以便将环境的影响降到最低。仿真模型有一个小一点的接地护罩。实际上，电极通过特弗纶绝缘材料直接连接在信号源上，仿真使用二维设计，其连接线路与电极平行。作为样品的塑料瓶里的自来水与棱柱型模型的测量结果不同，它的介电常数为81。仿真和测试显示出一个区域，当目标物体特别靠近传感器时(少于1cm)， 共 5 页 第 4 页 电容趋向于改变方向。为了解释靠近目标时传感器的变化，一个仿真建立了起来。在平行板电容几何学中预测到如果一种电介质插入电容器中，电容器的电容会增加。这种设想是基于平行电场线在电容计算中的相关原理提出的。根据使用1-pF的电容进行的稳定性测试实验中湿度和温度的变化，计算而得的修正值在电容传感器测量中获得了很好的应用。经过修正之后的残差仍然伴随着相对湿度值。在高湿度环境中(湿度范围为27% 到43%)，水分的吸收会对木头的介电常数产生影响，而上面提到的作用会加重这种影响。 关于系统的动态容量和它的校准的研究是最富有意义的。此篇文章关注的是稳定性分析，做了耗时很长的实验，稳定性结果揭示出了系统良好的性能。CDC可以配置两个电容输入端。这一特点提高了测量的精度。 V.结论 本文叙述了一种简单而有效的电容传感器系统，它能够辨别10cm以上距离的不同材料的目标物。其原型是由基于测量系统的微型电脑构成，包括一个CDC，环境传感器，和电容探测器。已经构建了一个模型来降低系统对环境温湿度变化的敏感性。温湿度变化的时间常数来源于测量数据。本系统经过实验室和野外条件的测试，并且相对于实验室标准进行了校准。当使用1-pF电容进行测量实验时，实验显示出了30/106以上的稳定性和对于材料介电常数变化的高灵敏性。 致谢 作者对于美国海军学院的J. Bradshaw 和 N. Tyson给予作者技术上的支持表示衷心的感谢。对联邦标准与技术协会的A. Koffman 共 5 页 第 5 页 的支持表示衷心的感谢。 共 5 页 第 6 页 Abstract A new capacitance sensor system was developed with a microcontroller, commercial humidity and temperature sensors, a capacitance-to-digital converter, and a custom-built capacitance sensor. The performance of the system was evaluated by simulation and testing of the prototype. The impact of variations of ambient conditions on the system performance was analyzed, and a model for correcting the humidity and temperature influence was developed. Based on the experimental results obtained in an uncontrolled environment using a standard capacitor of 1 pF, the 24-h stability of the system was estimated to be within 30 parts in 106. The high stability and sensitivity of the system allow its effective use in object-detection applications. Several sensors were constructed and evaluated while sensing various materials under different scenarios. Index Terms—Capacitance measurement, capacitance sensor, electric field simulation, humidity, temperature variations. I. INTRODUCTION RECENT technological advancements have made available a high-resolution (24-bit) Σ?Δ capacitance-to-digital converter (CDC) with a capacitance input range on the order of a couple of picofarads [1]. Potential applications of these CDCs are numerous. The project that motivated our research involves a capacitance sensor incorporated on a robotic arm to classify the objects to be manipulated by the arm [2]. The critical design requirements in this application were small physical size and high sensitivity and reliability (stability and repeatability) of measurements. A simple capacitive sensor was developed to measure the variations in the dielectric constant as a classification criterion for a limited object set [3]. Avramov-Zamurovic and Lee focusedon the following: 1) evaluating the CDC using a nitrogen-filled standard 1-pF capacitor; 2) developing a correction model for the humidity and temperature influences on measurements; and 3) building useful sensors for detecting the presence of selected materials under different scenarios related to airport security applications. II. CAPACITANCE SENSOR SYSTEM DESIGN A block diagram of the developed capacitance sensor system is shown in Fig. 1. It is based on a microcomputer that controls the CDC and ambient humidity and 共 5 页 第 7 页 temperature sensors and collects measurement data. The CDC is connected to a custom made capacitive sensor placed under a target object. A personal computer is used to display the data and develop the microcontroller software. The microcontroller development kit used in this application was developed at the U.S. Naval Academy in the Weapons and Systems Department. The microcomputer used is an 8-bit processor operating at 30 MHz, with excellent math performance. This hardware was selected because it has I2C communication capability, which is necessary in the use of CDC and sensors. The features critical to the performance of the microcontroller in the capacitive sensor system application are presented in this paper, and more detailed information on the microprocessor and the development kit can be found in the product manuals in [5] and [6]. The CDC in our application uses a single-ended input in the range of ?4 pF with an internal excitation signal at 16 kHz. The CDC supports 16- and 32-kHz sources. The 16-kHz signal was chosen to compare the CDC readings with those of a commercially available capacitance bridge that operates up to 20 kHz. The CDC is designed for floating capacitance sensors, and the measured capacitance is isolated from the ground. It is tolerant of parasitic capacitance to ground of up to 60 pF. There is an option to eliminate the offset in measured capacitance of up to 17 pF using an on-chip programmable capacitance digital-to-analog converter. Manufacturer statements on linearity (?0.01%), effective resolution (21 bit), and accuracy (4 fF) have to be verified using standard capacitors, and the uncertainty levels need to be established. This is a focus of future research. Detailed explanation of all aspects of the CDC is given in the manufacturer manual [1], and the characteristics relevant to the capacitance sensor system are given in this paper. The CDC is a surface-mounted circuit, and the mounting board is designed to have digital connections to the microcontroller (I2C). A Teflon-insulated coaxial cable directly connects the measuring input pins on the CDC to the Bayonet–Neill–Concelman connectors. The connectors are mounted on the grounded shielded box that encloses the CDC. The CDC has an on-chip temperature sensor used in a selfcalibration process. Since it was observed that the CDC is significantly impacted by humidity variations and, to a lesser extent, by temperature variations, humidity and temperature sensors are used to measure ambient conditions. The data from humidity and temperature sensors are used 共 5 页 第 8 页 to estimate capacitance corrections due to ambient variations and to cross check the CDC on-chip temperature readings, contributing to an improved measurement process reliability. The CDC averages a preset number of capacitance readings. The measurements are recorded as a fraction of the full scale. The measured frequency of 16 kHz and an applied voltage of 2.5 V are specified. Internal noise filters are applied to each measurement, and the software periodically executes the CDC offset and gain self-calibration. The manufacturer-specified maximum capacitance data rate is 90 Hz, but with all of the filtering, temperature and humidity sensor readings, and self calibration procedures implemented, the practical system data rate is 3 measurement/min. III. EVALUATION OF THE CAPACITANCE SENSOR SYSTEM STABILITY AND CORRECTION MODEL A nitrogen-filled standard 1-pF capacitor with no active thermal control was used to establish the stability of the capacitance sensor system in the laboratory environment (26 ? 1 ?C). The 24-h stability of the standard capacitor, which was measured using a high-accuracy commercial capacitance bridge, is ?2 parts in 106. The ambient humidity and temperature readings were used in developing a model for correcting the capacitance system dependence on environmental condition changes. Observation of the effect of humidity on the capacitance sensor system showed that there was a time lag between the change in humidity and that in CDC readings. Humidity time constant τ is defined as the time elapsed between the onset of the ambient humidity change and the start of the capacitance reading drift. A sliding window was used to match the capacitance and humidity data. The time constant for the developed prototype was estimated to be 30 min. To eliminate random noise, the data sets consisting of capacitance, humidity, and temperature readings were averaged over 3 min. Since the rapid burst-type deviations of humidity from the average value do not significantly influence the capacitance readings, a running average AH over 30 min was applied on the humidity data. The preprocessing (time constant, noise filtering, and running average) requires about 1 h of lead time before the capacitance corrections can be applied. IV. APPLICATION AND TEST RESULT 共 5 页 第 9 页 In this paper, we describe the design of capacitance probes for possible use in sensing different target materials at a distance. The probe design evolved from modifications on the cross-capacitance principle. A high level of stability in measuring the capacitance changes was achieved. The copper electrodes are fixed on Teflon support and separated by 0.47 cm. The bottom of the sensor is grounded, and coaxial cabling is provided to measure the capacitance. The prototype sensor is rigidly built, but similar sensors could be constructed using flexible printed board circuit, making it suitable for other applications. In the design phase, the sensor topology was analyzed using a MAXWELL 3-D v4.1 electric field simulation software based on mesh analysis. The 6% difference in accuracy between the simulation capacitance values and the measurements is attributed to several discrepancies between the simulation and the prototype realization. Measurements were made on a large grounded plate to minimize the effect of the surroundings, where the simulation has a smaller grounded shield. Practically, the electrodes are directly connected to the source through the Teflon insulation, and the simulation used 2-D layouts and had the connection traces parallel to the electrodes. Tap water in a plastic bottle measured as a sample differed from the prism shared model in size and had a dielectric constant of 81. Note that both the simulation and measurements show a zone when the target sample is very close (less than 1 cm) to the sensor where the capacitance trend changes direction. To explain the sensor behavior at close proximity to the target, a simulation was performed. In the parallel-plate capacitor geometry, it is expected that the capacitance will increase if a dielectric sample is inserted. This assumption is based on the parallel electric field lines used in capacitance calculation. The corrections were applied using humidity and temperature factors calculated from the stability tests using a 1-pF capacitor. Note that the residual pattern after corrections still follows the relative humidity. This effect could be attributed to the dielectric constant of the wood sample changing due to the water absorption in a high level of humidity (ranging from 27% to 43% humidity). The most significant is the research on the system’s dynamic capabilities and its calibration. Since this paper was focused on stability analysis, very long tests were performed, and the stability results revealed very good performance. The CDC could be configured to have two capacitance inputs. 共 5 页 第 10 页 This feature improves the measurement accuracy. V. CONCLUSION This paper has described the development of a simple yet effective field-deployable capacitance sensor system for the discrimination of materials at distances of up to 10 cm from the target. The prototype consists of a microcomputer-based measurement system, including a CDC, ambient sensors, and capacitance probes. A model was developed to reduce the sensitivity of the system to ambient temperature and humidity variations. The time constants for temperature and humidity variations were determined from measurement data. The system was tested under laboratory and field conditions and calibrated against laboratory standards. It showed a stability that is better than 30 parts in 106 while measuring 1-pF capacitance and high sensitivity to variations in material dielectric constants. ACKNOWLEDGMENT The authors would like to thank J. Bradshaw and N. Tyson of the U.S. Naval Academy for their technical assistance and A. Koffman of the National Institute of Standards and Technology for his support. 共 5 页 第 11 页 指 导 教 师 评 语 外文翻译成绩: 指导教师签字: 年 月 日 注:1. 指导教师对译文进行评阅时应注意以下几个方面:?翻译的外文文献与毕业设计(论文)的主题是 否高度相关，并作为外文参考文献列入毕业设计(论文)的参考文献;?翻译的外文文献字数是否 达到规定数量(3 000字以上);?译文语言是否准确、通顺、具有参考价值。 2. 外文原文应以附件的方式置于译文之后。 共 5 页 第 12 页