光电传感器英文和译文.doc

Progress in Materials ScienceVolume 46, Issues 34, 2001, Pages 461504The selection of sensorsJ Shieh, J.E Huber, N.A Fleck, , M.F AshbyDepartment of Engineering, Cambridge University, Trumpington Street, Cambridge CB2 1PZ, UKAvailable online 14 March 2001.http:/dx.doi.org/10.1016/S0079-6425(00)00011-6, How to Cite or Link Using DOIPermissions & ReprintsAbstractA systematic method is developed to select the most appropriate sensor for a particular application. A wide range of candidate sensors exist, and many are based on coupled electrical and mechanical phenomena, such as the piezoelectric, magnetostrictive and the pyro-electric effects. Performance charts for sensors are constructed from suppliers data for commercially available devices. The selection of an appropriate sensor is based on matching the operating characteristics of sensors to the requirements of an application. The final selection is aided by additional considerations such as cost, and impedance matching. Case studies illustrate the selection procedure.KeywordsSensors; Selection; Sensing range; Sensing resolution; Sensing frequency1. IntroductionThe Oxford English Dictionary defines a sensor as “a device which detects or measures some condition or property, and records, indicates, or otherwise responds to the information received”. Thus, sensors have the function of converting a stimulus into a measured signal. The stimulus can be mechanical, thermal, electromagnetic, acoustic, or chemical in origin (and so on), while the measured signal is typically electrical in nature, although pneumatic, hydraulic and optical signals may be employed. Sensors are an essential component in the operation of engineering devices, and are based upon a very wide range of underlying physical principles of operation.Given the large number of sensors on the market, the selection of a suitable sensor for a new application is a daunting task for the Design Engineer: the purpose of this article is to provide a straightforward selection procedure. The study extends that of Huber et al. 1 for the complementary problem of actuator selection. It will become apparent that a much wider choice of sensor than actuator is available: the underlying reason appears to be that power-matching is required for an efficient actuator, whereas for sensors the achievable high stability and gain of modern-day electronics obviates a need to convert efficiently the power of a stimulus into the power of an electrical signal. The classes of sensor studied here are detailed in the Appendices.2. Sensor performance chartsIn this section, sensor performance data are presented in the form of 2D charts with performance indices of the sensor as axes. The data are based on sensing systems which are currently available on the market. Therefore, the limits shown on each chart are practical limits for readily available systems, rather than theoretical performance limits for each technology. Issues such as cost, practicality (such as impedance matching) and reliability also need to be considered when making a final selection from a list of candidate sensors.Before displaying the charts we need to introduce some definitions of sensor characteristics; these are summarised in Table 1.1 Most of these characteristics are quoted in manufacturers' data sheets. However, information on the reliability and robustness of a sensor are rarely given in a quantitative manner.Table 1. Summary of the main sensor characteristicsRangemaximum minus minimum value of the measured stimulusResolutionsmallest measurable increment in measured stimulusSensing frequencymaximum frequency of the stimulus which can be detectedAccuracyerror of measurement, in% full scale deflectionSizeleading dimension or mass of sensorOpt environmentoperating temperature and environmental conditionsReliabilityservice life in hours or number of cycles of operationDriftlong term stability (deviation of measurement over a time period)Costpurchase cost of the sensor ($ in year 2000)Full-size tableIn the following, we shall present selection charts using a sub-set of sensor characteristics: range, resolution and frequency limits. Further, we shall limit our attention to sensors which can detect displacement, acceleration, force, and temperature.2 Each performance chart maps the domain of existence of practical sensors. By adding to the chart the required characteristics for a particular application, a subset of potential sensors can be identified. The optimal sensor is obtained by making use of several charts and by considering additional tabular information such as cost. The utility of the approach is demonstrated in Section 3, by a series of case studies.2.1. Displacement sensorsConsider first the performance charts for displacement sensors, with axes of resolution versus range R, and sensing frequency f versus range R, as shown in Fig. 1 and Fig. 2, respectively.Fig. 1. Resolution versus sensing range for displacement sensors.View thumbnail imagesFig. 2. Sensing frequency versus sensing range for displacement sensors.View thumbnail images2.1.1. Resolution  sensing range chart (Fig. 1)The performance regime of resolution versus range R for each class of sensor is marked by a closed domain with boundaries given by heavy lines (see Fig. 1). The upper limit of operation is met when the coarsest achievable resolution equals the operating range =R. Sensors of largest sensing range lie towards the right of the figure, while sensors of finest resolution lie towards the bottom. It is striking that the range of displacement sensor spans 13 orders of magnitude in both range and resolution, with a large number of competing technologies available. On these logarithmic axes, lines of slope +1 link classes of sensors with the same number of distinct measurable positions, . Sensors close to the single position line =R are suitable as simple proximity (on/off) switches, or where few discrete positions are required. Proximity sensors are marked by a single thick band in Fig. 1: more detailed information on the sensing range and maximum switching frequency of proximity switches are summarised in Table 2. Sensors located towards the lower right of Fig. 1 allow for continuous displacement measurement, with high information content. Displacement sensors other than the proximity switches are able to provide a continuous output response that is proportional to the target's position within the sensing range. Fig. 1 shows that the majority of sensors have a resolving power of 103106 positions; this corresponds to approximately 1020 bits for sensors with a digital output.Table 2. Specification of proximity switchesProximity switch typeMaximum switching distance (m)Maximum switching frequency (Hz)Inductive6×1041×10155000Capacitive1×1036×1021200Magnetic3×1038.5×1024005000Pneumatic cylinder sensors (magnetic)Piston diameter 8×1033.2×1013005000Ultrasonic1.2×1015.2150Photoelectric3×1033002020,000Full-size tableIt is clear from Fig. 1 that the sensing range of displacement sensors cluster in the region 105101 m. To the left of this cluster, the displacement sensors of AFM and STM, which operate on the principles of atomic forces and current tunnelling, have z-axis-sensing ranges on the order of microns or less. For sensing tasks of 10 m or above, sensors based on the non-contacting technologies of linear encoding, ultrasonics and photoelectrics become viable. Optical linear encoders adopting interferometric techniques can achieve a much higher resolution than conventional encoders; however, their sensing range is limited by the lithographed carrier (scale). A switch in technology accounts for the jump in resolution of optical linear encoders around the sensing range of 0.7 m in Fig. 1.Note that “radar”, which is capable of locating objects at distances of several thousand kilometres,3 is not included in Fig. 1. Radar systems operate by transmitting high-frequency radio waves and utilise the echo and Doppler shift principles to determine the position and speed of the target. Generally speaking, as the required sensing range increases, sensors based on non-contact techniques become the most practicable choice due to their flexibility, fast sensing speed and small physical size in relation to the length scale detected. Fig. 1 shows that sensors based on optical techniques, such as fibre-optic, photoelectric and laser triangulation, cover the widest span in sensing range with reasonably high resolution.For displacement sensors, the sensing range is governed by factors such as technology limitation, probe (or sensing face) size and the material properties of the target. For example, the sensing distance of ultrasonic sensors is inversely proportional to the operating frequency; therefore, a maximum sensing range cut-off exists at about R=50 m. Eddy current sensors of larger sensing face are able to produce longer, wider and stronger electromagnetic fields, which increase their sensing range. Resolution is usually controlled by the speed, sensitivity and accuracy of the measuring circuits or feedback loops; noise level and thermal drift impose significant influences also. Sensors adopting more advanced materials and manufacturing processes can achieve higher resolution; for example, high-quality resistive film potentiometers have a resolution of better than 1 m over a range of 1 m (i.e. 106 positions) whereas typical coil potentiometers achieve only 103 positions.2.1.2. Sensing frequency  sensing range chart (Fig. 2)When a displacement sensor is used to monitor an oscillating body, a consideration of sensing frequency becomes relevant. Fig. 2 displays the upper limit of sensing frequency and the sensor range for each class of displacement sensor. It is assumed that the smallest possible sensing range of a displacement sensor equals its resolution; therefore in Fig. 2, the left-hand side boundary of each sensor class corresponds to its finest resolution.4 However, sensors close to this boundary are only suitable as simple switches, or where few discrete positions are to be measured.Lines of slope 1 in Fig. 2 link classes of sensors with the same sensing speed, fR. For contact sensors such as the LVDT and linear potentiometer, the sensing speed is limited by the inertia of moving parts. In contrast, many non-contact sensors utilise mechanical or electromagnetic waves and operate by adopting the time-of-flight approach; therefore, their maximum sensing speed is limited by the associated wave speed. For example, the maximum sensing speed of magnetostrictive sensors is limited by the speed of a strain pulse travelling in the waveguide alloy, which is about 2.8×103 m s1.The sensing frequency of displacement sensors is commonly dependent on the noise levels exhibited by the measuring electronic circuits. Additionally, some physical and mechanical limits can also impose constraints. For example, the dynamic response of a strain gauge is limited by the wave speed in the substrate. For sensors with moving mass (for example, linear encoder, LVDT and linear potentiometer), the effects of inertial loading must be considered in cyclic operation. For optical linear encoders the sensing frequency increases with range on the left-hand side of the performance chart, according to the following argument. The resolution becomes finer (i.e. decreases in an approximately linear manner) with a reduced scan speed V of the recording head. Since the sensor frequency f is proportional to the scan speed V, we deduce that f increases linearly with , and therefore f is linear in the minimum range of the device.2.2. Linear velocity sensorsAlthough velocity and acceleration are the first and second derivatives of displacement with respect to time, velocity and acceleration measurements are not usually achieved by time differentiation of a displacement signal due to the presence of noise in the signal. The converse does not hold: some accelerometers, especially navigation-grade servo accelerometers, have sufficiently high stability and low drift that it is possible to integrate their signals to obtain accurate velocity and displacement information.The most common types of velocity sensor of contacting type are electromagnetic, piezoelectric and cable extension-based. Electromagnetic velocity sensors use the principle of magnetic induction, with a permanent magnet and a fixed geometry coil, such that the induced (output) voltage is directly proportional to the magnet's velocity relative to the coil. Piezo-velocity transducers (PVTs) are piezoelectric accelerometers with an internal integration circuit which produces a velocity signal. Cable extension-based transducers use a multi-turn potentiometer (or an incremental/absolute encoder) and a tachometer to measure the rotary position and rotating speed of a drum that has a cable wound onto it. Since the drum radius is known, the velocity and displacement of the cable head can be determined.5Optical and microwave velocity sensors are non-contacting, and utilise the optical-grating or Doppler frequency shift principle to calculate the velocity of the moving target. Typical specifications for each class of linear velocity sensor are listed in Table 3.Table 3. Specification of linear velocity sensorsSensor classMaximum sensing range (m/s)Resolution (number of positions)Maximum operating frequency (Hz)Magnetic induction253605×1045×1051001500PVT0.251.31×1055×1057000Cable-extension0.7151×1051×1061100Optical and microwave131651×105> 10,000目录1. 简介22. 传感器性能图表22.1位移传感器32.1.1分辨率 - 感应范围图（图1）42.1.2.检测频率 检测范围图（图2）52.2线性速度传感器6问题3-4，2001年第46卷，页461-504 传感器的选择J Shieh, J.E Huber, N.A FleckM.F Ashby剑桥大学工程系，英国剑桥CB2的1PZ，Trumpington街_摘要对于一个特定的应用系统来说要选择最为合适的传感器。大量种类的传感器存在，并且许多传感器是基于耦合的电气和机械现象，如压电，磁致伸缩和焦电效应。传感器的性能图表是从商用设备供应商提供的数据而来。选择适当的传感器是基于传感器的经营特色，以匹配应用程序要求。最终的选择是根据外加的其他因素，如成本，阻抗匹配。这些案例研究说明了选拔程序。关键词传感器选择感应范围检测分辨率检测频率_1. 简介“牛津英语大辞典”定义传感器“一个能够检测测量环境或一些变量，且能够记录，显示，或以其他方式收到信息的设备”。因此，传感器具有将刺激转换成可测量信号的功能。这些刺激可以是力学，热学，电磁学，声学，或起源于化学（等）的刺激，而测得的信号通一般是电信号，虽然气动，液压和光信号也可以采用。基于广泛而最基本物理原理的传感器是工程设备中必不可少的组成部分。考虑到市场上种类繁多的传感器，对于工程设计人员为一个新的应用程序选择合适的传感器是一项艰巨的任务：这篇文章的目的就是提供一套简单的挑选步骤。本研究是对胡贝尔等对执行机构选择问题的延伸和补充。传感器比执行机构具有更为广泛的应用：根本原因，驱动器需要有效的比配功率，而传感器是实现现性电子产品所要求的的高稳定性和增益性并能将其转换成强有效的电信号地刺激。传感器种类的研究将在附录中详细的阐述。2. 传感器性能图表 在本节中，传感器性能数据以性能为横轴的二维图中进行展示。这些数据是基于当前市场上一般可用的传感系统的。而不是具有工艺理论研究的理论知识。如成本，实用性（如阻抗匹配）和可靠性等问题也需要从备选传感器性能列表中进行对比，然后在做最后的选择。在阐释图表之前，我们需要介绍一些有关传感器特性的定义，在表1.1中给出的性能多是厂商会给出的。然而，传感器的可靠性和鲁棒性很