[雙語翻譯]超聲波測距外文翻譯--低成本超聲波測距系統(tǒng)的設(shè)計考慮和性能（英文）

1、 Abstract—Numerous systems for reliable, low-cost, indoor, ultrasonic sensing have been demonstrated during the past few decades. Existing solutions span very low-cost transducers to high-grade equipment. This paper p

2、resents design details of a low-cost, single-tone, ranging system. A set of experimental results is included in order to evaluate consistency between ranging results obtained from receivers with various archite

3、ctures. The effects of channel equalization and signal pre- coding evaluated show how these systems can deal with realistic reflectors. Index Terms— ultrasonic transducers, Cramer-Rao bounds, time of arrival estimation

4、, indoor environments, CFAR, Tomlinson-Harashima pre-coder I. INTRODUCTION ANGING is an essential element in many robotics, industrial, and ambient-intelligence applications. Ultrasonic sensing is a potential solut

5、ion for reliable and low- cost indoor localization and ranging. In the experiments described within this paper, ultrasonic transducers, acting as both emitters and receivers, were used to locate reflectors based on th

6、e Time-of-Flight (ToF) of sound waves. The performance of ultrasonic distance measurement depends on employed signal bandwidth, which directly limits the amount of available information. The presented system uses low-

7、cost, narrow-band, piezoelectric transducers (PZT). A focus of the work was to consider ranging consistency and investigate the benefit of bandwidth extension by channel equalization and a constant-false alarm rate (

8、CFAR) [1] based detection of quality of ranging measurements. The system suggested presents core technology for more complex, sensor-array based localization or bearing estimation [2][3]. Such systems ordinarily use

9、Direct Sequence Code Division Multiple Access (DS-CDMA) to manage the multiple access of a number of transducers. This was stimulus to additionally explore the impact of Pseudo-random Noise (PN) signaling to the syste

10、m [4]. A. Organization The paper's organization is as follows: Section II briefly D. S. Zivkovic is with the Bitgear Wireless Design Services d.o.o., 11000 Belgrade, Serbia (e-mail: dragan.zivkovic@bitgear.rs).

11、B. R. Markovic is with the Bitgear Wireless Design Services d.o.o., 11000 Belgrade, Serbia (e-mail: bogdan.markovic@bitgear.rs). D. Rakic is with the Bitgear Wireless Design Services d.o.o., 11000 Belgrade, Serbia (e-

12、mail: dejan.rakic@bitgear.rs). S. Tadic is with the Bitgear Wireless Design Services d.o.o., 11000 Belgrade, Serbia (e-mail: srdjan.tadic@bitgear.rs). introduces the ultrasonic sensing problem and related prior work.

13、 Section III presents design details of the developed system. Experimental results are described in Section IV. Section V gives an overview of processing hardware and related practical considerations. The conclusions

14、are given in Section VI. II. ULTRASONIC SENSING AND RELATED WORK The ultrasonic sensor will measure distance by emitting a sound wave and then “l(fā)istening“ for a set period of time, allowing for the return of echoes b

15、ouncing off the target, before re-transmitting. Due to the phenomenon called ringing – a continued vibration of the transducer after its excitation pulse – the sensor cannot reliably measure within a so-called “dead

16、zone” in its proximity. The accuracy of measurement depends on the accuracy of the speed of sound used in calculation. The speed of sound (c) can be derived as Mγ = c RT (1) where

17、γ, R, T, and M are the specific heat ratio, universal gas constant, absolute temperature (in Kelvin), and molar mass. As temperature increases, the sound waves travel more quickly. Speed is also affected by the medium

18、's composition. This paper considers air as the propagation medium. Other reasons that can lead a sensor to incorrectly locate reflectors include air currents, in-band acoustic noise, or interference from other t

19、ransducers. Reflector surface, geometry, distance, size, and beam angle define the levels of absorption and scattering, which limit maximum detectable target distance attainable. Maximal distance also depends on air h

20、umidity as well as sound frequency, since they directly impact attenuation. Change in sound frequency or speed also affects wavelength. The consequence is change in resolution, accuracy, and minimum detectable reflec

21、tor size. There are also other difficulties, such as sensitivity to input voltage and frequency-temperature dependence. Listed effects exist within both most-common low-cost transducer technologies – Microelectromech

22、anical systems (MEMS) and PZT. Within the indoor scenario, surrounding walls and the floor behave as “background” and cause multipath propagation. Higher-level decision logic should be aware of this. A brief overview

23、of the theory of ultrasonic positioning is given in [5]. This chapter will briefly list two important parameters of an ultrasonic ranging system. First important parameter is a “range resolution” - the term originati

24、ng from radar technology, also denoted as “axial resolution” in medical systems. It is the minimal range difference needed to distinguish movement of a target along one bearing. In case of pulse compression ultrasonic

25、 system as RDesign Considerations and Performance of Low-Cost Ultrasonic Ranging System Dragan S. Zivkovic, Bogdan R. Markovic, Dejan Rakic, Srdjan Tadic 978-1-4673-6033-3/13/$31.00 ©2013 IEEEThe transducer frequen

26、cy response is shown in 4. The heterodyne baseband receiver is shown in Fig. 5. The complex baseband signal is filtered by a Root Raised Cosine (RRC) filter, and down-sampled to the PN chip rate of 4 kHz. Baseband i

27、mpulse response sampled at chip rate is shown in Fig. 6. The complex impulse response is 19 samples long, and its corresponding Z –transform ( ) z Hbis represented as a complex polynomial of 18th degree. The baseband po

28、lynomial can be represented with the following equation (4), using spectral factorization: ( ) ( ) ( ) z S z S = z H + b ?(4) Where polynomial ( ) z S+is formed by the roots of ( ) z Hbto be minimum phase and ( ) z

29、S ?is derived from the roots out of the unit circuit. The Tomlinson-Harashima (TH) pre-coder (Fig. 7.) is applied to the PN sequence in order to increase the spectral efficiency of the transmitter [9]. The TH pre-code

30、r is designed according to pulse response, while reflections are captured from a single small object in empty space. The background reflections are eliminated, and the transfer function is defined only by the transcei

31、ver and receiver. PN symbol ( ) m sis summarized by modulo 1 with its recursive part formed by a polynomial with minimum phase, guaranteeing stability. The “modulo-one” adder is also being used in order to prevent sign

32、al accumulation. The resulting spectrum, made of pre-coded and PN symbols, is shown in Fig. 8. This pre-coded spectrum is more equally distributed over the band of interest than the one created by using only a trans

33、ducer. On the receiver side, a predefined Zero Force (ZF) equalizer is applied in order to increase detection resolution. The ZF equalizer should be made to, as much as possible, satisfy (5). ( ) ( ) N z z H z S ? ?

34、≈(5) 36 38 40 42 44 46-100-80-60-40-200204060Frequency [KHz]Magnitude[dB]EqualizerChannel Equalized channelFig. 10. Frequency response of equalized channel ( ) s nT cos( ) s nT sin ?( ) s tr nT r( ) B tr nT a( ) B tr nT

35、bFig. 5. Heterodyne baseband receiver for 4 kHz signal ( ) z S + ? 1( ) m s ( ) n u ( ) t txFig. 7. Transmitter with TH pre-coder 36 38 40 42 44 46-60-40-200204060frequency [KHz] magnitude [dB] pre-coder used transducer