CC BY
107
АВТОМАТИЗИРОВАННЫЙ ГЕОДЕЗИЧЕСКИЙ МОНИТОРИНГ С ПРИМЕНЕНИЕМ НЕДОРОГОЙ GNSS
Фолькер Швигер
Технический университет Штутгарта, Институт инженерной геодезии, Geschwister-Scholl-Str. 24D, 70174, Штутгарт, Германия, директор института, доктор-инженер, профессор, тел. ++49/711-68584041, e-mail: volker.schwieger@ingeo.uni-stuttgart.de, web-site: http://www.uni-stuttgart.de/ingeo/
Ли Чжан
Технический университет Штутгарта, Институт инженерной геодезии, Geschwister-Scholl-Str. 24D, 70174, Штутгарт, Германия, научный сотрудник, тел. ++49/711-68584049, e-mail:
li.zhang@ingeo.uni-stuttgart.de
В статье описывается использование приемников ГНСС и произвольных сетей для задач мониторинга. В эти задачи входит геодезический мониторинг зданий и инженерных сооружений и природных объектов, например оползни. Последнее время стали активно использоваться недорогие приемники. Наряду с обзором существующих систем мониторинга с использованием ГНСС приводятся и первые результаты, полученные новой системой, разработанной Институтом инженерной геодезии в Техническом университете Штутгарта. Точность определения положения составляет 2 см для измерений с интервалом 10 минут.
Ключевые слова: ГНСС, мониторинг, беспроводная сеть передачи данных, произвольная сеть, топология сетки.
AUTOMATED GEODETIC MONITORING BY LOW-COST GNSS
Volker Schwieger
University of Stuttgart, Institute of Engineering Geodesy, Geschwister-Scholl-Str. 24D, 70174 Stuttgart, Germany, Director of Institute, Professor, Dr.-Ing, tel. ++49/711-68584041, e-mail: volker.schwieger@ingeo.uni-stuttgart.de, web-site: http://www.uni-stuttgart.de/ingeo/
Li Zhang
University of Stuttgart, Institute of Engineering Geodesy, Geschwister-Scholl-Str. 24D, 70174 Stuttgart, Germany, Research Associate, tel. ++49/711-68584049, e-mail: li.zhang@ingeo.uni-stuttgart.de
The contribution deals with the use of GNSS (Global Navigation Satellite System) receivers and ad-hoc networks for monitoring tasks. These tasks enfold the monitoring of buildings and structures as well as natural objects, e.g. landslides. The current trend is to operate plenty of low-priced receivers. Beside an overview of available GNSS monitoring systems, first results of a new system developed at the Institute of Engineering Geodesy of University Stuttgart (IIGS) are presented. The accuracy is characterized by a maximum positional deviation of 2 cm for a measurement interval of 10 minutes.
Key words: GNSS, monitoring, WLAN, ad-hoc network, mesh topology.
INTRODUCTION
In general, a monitoring concept has to be realized for any civil structure (Welsch et al. 2000). The monitoring itself is one of the conceptual pillars to guarantee the reliability and the stability of structures. The other pillars are a measure and emergency concept for occurring irregularities and possible constructive measures. Monitoring can be subdivided into visual inspection, function tests on a regular basis, and the monitoring surveys. The measurements control the state and, if a high data rate is chosen, the reaction of the monitored objects. From a geodetic point of view, monitoring deals with geometric quantities and therefore rigid body movements and deformations of the object. Since this contribution focuses on GNSS, the monitoring from outside the civil structure is of interest only. The internal view into the object is not possible using GNSS.
The dominating trends for monitoring are automation and continuity. In the past years monitoring surveys were realized in defined time epochs (e.g. once a month or once a year, so-called epoch-wise measurements). Nowadays, information concerning the state of the monitored object should be available at any time (and at any place). In emergency cases, this allows a reaction in near realtime. To ensure continuous measurements the measuring instruments have to stay at the monitoring site and high investment costs are required compared to epoch-wise measurements. Total stations and GNSS receivers are the only possibility to continuously and automatically acquire three-dimensional positions. Up to now, the price of the GNSS receivers forbids their object-wide (e.g. bridges, dam walls, landslides) installation. Due to the fact that low-cost receivers are available, which deliver measurements of high quality, new possibilities arise. Currently, GNNS receivers may be purchased for definitely less than 100 €. If correct evaluation strategies and methods are used, the accuracy level is comparable to the one of expensive geodetic level receivers (Weston & Schwieger 2010).
MONITORING BY GNSS
Since 1970, GNSS have been used for geodetic tasks. At the beginning, exclusively the American Global Positioning System (GPS) was used; meanwhile the Russian GLONASS is completely operable. The European Galileo and the Chinese Compass are currently in development. In the near future, the users may use 4 GNSS comprising more than 100 navigation satellites (e.g. Kleusberg 2010).
Absolute GNSS positioning is not useful for monitoring tasks, since the accuracy available is around some meters RMS. In contradiction, geodetic receivers can deliver positions in the sub-cm level, since they measure relatively using the precise carrier phase as observation. These geodetic receivers were used very early for geodetic monitoring issues. As written before, the development has gone from epoch-wise to continuous measurements. For the latter, GNSS receivers are ideal, since their data acquisition is always continuous and automatic. First experiences were gained by Baumker & Fitzen (1996). In the sequel a multitude of automated GPS monitoring systems were developed. In addition to measurement and evaluation, the system GOCA comprises communication and deformation analysis (Kalber et al.
2000). Besides, a GPS monitoring system was investigated at TU Graz (Hartinger
2001). In recent years the relevant research came into accelerated motion. Particularly the University of the Armed Force Munich has developed a system based on low-priced Novatel receivers. It has been used for monitoring of landslides in nearrealtime (Glabsch et al. 2009). In the following, the authors describe the development and the operation of mass-market GPS receivers within a WLAN ad-hoc monitoring network at IIGS. First results will be presented too.
WLAN AD-HOC MONITORING GNSS NETWORK
Normally, monitoring systems are controlled centrally and a router takes over all tasks as access point. Consequently, data acquisition is controlled by this central station and all clients transmit their data to this central station. This is the typical network structure that is applied within WLAN networks, too. Principally, the nodes within wireless ad-hoc networks are organized in a different way. Hereby the dynamic mesh topology is the central element (Johnson et al. 2007). Each node comprises a router and an antenna and communicates with the nearest neighbor node. This means that any node may serve as access point. Crucial to the monitoring system is the fact that the data can search for their own way through the net, independent from the path. Therefore it is without importance which path the data have taken. Besides, this way is not detectable without an intervention into the network system. As a rule, the communication in ad-hoc networks as well as in centrally controlled networks is based on standards of the Institute of Electronic and Electronical Engineers IEEE802.11X and uses the WLAN frequency 2.4 GHz. The major advantages of the mesh topology are the improved robustness and the increased band width. Both is possible due to the parallel data communication on alternative paths within the network. Consequently, ad-hoc routing algorithms are required. These are the essential base for self-organisation and self-networking of the individual nodes.
In general, the band width is no drawback with respect to monitoring problems, since individual measurement values like inclinations and temperatures or classical geodetic quantities like distances and angles are transmitted. In the case of GNSS the data volume is definitely higher: for each receiver 4.7 kbit/s are transmitted for a sampling rate of 1 Hz. These requirements are fulfilled by the CabLynx Wireless Router that is integrated into an ad-hoc network for recent investigations at the IIGS. The router is used as WLAN router, but has the possibility to realize data transport via UMTS and GPRS, too. Since the network is self-organized, the receiving and the transmitting directions for the data are not known a-priori. Additionally, these directions may change, if e.g. a link is disturbed by external influences. Hence, an omni-directional antenna is required (for the described system the respective antenna of the company VIMCOM is used). The autonomous system is completed for each node by a solar panel, a charge controller, and a battery. The heart of the router are the U-Blox GPS antenna ANN-MS and the newest U-Blox GPS receiver LEA-6T. The LEA-6T delivers GPS raw data like code and phase measurements through a clearly defined interface into a proprietary data format. This is the foundation for the evaluation possibility of the GPS data in the cm-accuracy-level. Fig. 1 shows an overview of the system components, Fig. 2 presents the equipment of one node. The
low-cost GPS antenna was upgraded with a metal shielding (ground plate) to reduce multipath effects.
Figure 1: Realised system architecture
Figure 2: Equipment of one autonomous node
Fig. 3 presents the advantages of the dynamic mesh topology using the example of the IIGS GPS monitoring system. Three GPS-equipped nodes were built up. It could be shown that WLAN communication still works for distances longer than 1 km, if the WLAN antennas are installed in a height of approximately 3 m. The acquired data should reach the central station that is responsible for evaluation and analysis. Between client 1 and the central station line-of-sight was not available. Nevertheless all data reached the central station. If now client 2 was shut down and therefore excluded from the network topology, the central station did not get any data from client 1. Hereby two facts could be shown empirically: At first that a disturbed line-of-sight does not permit communication; at second that the mesh topology autonomously finds a suitable alternative.
Figure 3: Test area in Stuttgart-Vaihingen
FIRST RESULTS
To evaluate raw data, the following steps of the procedure are realized in the central station (Roman 2011):
- Identification of the node for any data set and generation of node resp. station-related files in proprietary format (own software),
- Generation of files in a defined exchange format (RINEX-Format) per station using the freeware TEQC (TEQC, 2012),
- Determination of relative station co-ordinates with respect to the central using the GPS evaluation software WA1 (WA1, 2012).
Using the example of the measurement configuration shown in the last section (Fig. 3) the accuracy of positioning is determined for different observation time intervals. The co-ordinates of all stations are known at sub-mm-level for evaluation purpose. Two one-hour sessions were measured, meaning that 6 baselines (threedimensional relative co-ordinates) could be evaluated. First one hour was completely evaluated. Subsequently, the observation interval was reduced in 5 minute steps up to a value of 10 minutes. Fig. 4 presents the repeatability standard deviations for the time spans between 10 and 20 minutes for the baseline between client 1 and the central station. For horizontal co-ordinates standard deviations below 4 mm are reached, independent of the observation time span. The vertical component, the height, leads to a worst three-dimensional standard deviation below 1 cm.
Since the co-ordinates of the individual stations are given with sub-mm accuracy, a measure of accuracy can also be determined with respect to these given co-ordinates. Fig. 5 shows the deviation of the average to the given co-ordinates for all observation time spans for the same baseline. For the north component and the height a marginal dependence from the time span can be seen. The east component shows higher values and no dependence of this sort. Altogether, the deviations stay below 2 cm for the three-dimensional positional deviation. The systematic east deviation may have its cause in the use of non-calibrated antennas. The calibration should improve the results (see e.g. Schwieger & Wanninger 2006).
Figure 4: Standard deviations for baseline client 1 - central station in dependency on
the observation time span
20
Observation Time [minutes]
Figure 5: Deviations to given co-ordinates for baseline client 1 - central station in
dependency on the observation time span
The discussed results are at the same level for the other baselines not shown in this contribution. Altogether the dependency on observation time span is very small. Obviously it not needed to measure one hour to reach the above-mentioned accuracy. A ten minute time span seems to be sufficient at first glance. The investigations of
Roman (2011) show that one needs 20 minutes for a reliable and automated solution. For shorter time spans manual analysis is required.
Outlook
Up to now the evaluation is realized in post-processing. For the future the evaluation has to be automated to reach near-realtime evaluation. Additionally, a Web-interface should be created to have the possibility to reach the monitoring data from any place in the world. In Zhang et al. (2012) some steps forward are presented. But only when these improvements are completely implemented, the system could be used on a regular base for monitoring surveys.
Furthermore, the accuracy has to be improved. Options for this are antenna-individual calibration, development of an improved antenna shielding and a test of alternative low-cost antennas. The progress into the some-mm accuracy level would prepare the ground towards an area-wide monitoring of landslides or bridges.
The expected improvement of GNSS-signal reception of the receivers - GPS and GLONASS are upgraded by Galileo and Compass - will lead to improved availability and reliability. Nevertheless, one has to consider that GNSS may only be used in environments without or with few shadowing effects. For shadowed objects or object parts complementary sensors like total stations or inclination sensors have to be used.
Remark
The content of this contribution was before published to a large extent in German language in Schwieger & Zhang (2012).
References
1. Baumker, M., Fitzen, H.P. (1996): Permanente Uberwachungsmessungen mit GPS. In: Brandstatter, Brunner, Schelling: Ingenieurvermessung 96 - Beitrage zum XII. Internationalen Kurs fur Ingenieurvermessung, Dummler, Bonn, 1996.
2. Glabsch, J., Heunecke, O., Schuhback, S. (2009): Monitoring of the Hornbergl landslide using recently developed low cost GNSS sensor network. Journal of Applied Geodesy, de Gruyter Verlag, Issue 4, 2009.
3. Hartinger, H. (2001): Development of a Continuous Deformation Monitoring System using GPS. Shaker Verlag, Aachen 2001.
4. Johnson, D., Matthee, K., Sokoya, D., Mboweni, L., Makan, A., Kotze, H. (2007): Building a rural wireless mesh network, Version 0.8. Wireless Africa Team of the Meraka Institute, Sudafrika, 30. Oktober 2007.
5. Kalber, S., Jager, R., Schwable, R. (2000): A GPS based online control and alarm system. GPS Solutions, Vol. 3, Issue 2, 2000.
6. Kleusberg, A. (2010): GNSS - Uberblick. In: GNSS 2010 - Vermessung und Navigation im 21. Jahrhundert 100. DVW-Seminar, 04./05. October 2010, Koln, Schriftenreihe des DVW, Vol. 63, WiBner Verlag, Augsburg.
7. Niemeier, W. (2007): Monitoring - was ist der Beitrag der Geodasie.
Ingenieurvermessung 07, Beitrage zum 15. Internationalen Ingenieurvermessungskurs Graz, 2007, Wichmann Verlag, Heidelberg.
8. Roman, M.A. (2011): Commissioning and Investigations regarding a Low-Cost GPS Monitoring System. Master Thesis, Faculty of Geodesy - Technical University of Civil Engineering Bucharest & Institute of Engineering Geodesy - University of Stuttgart (not published).
9. Schwieger, V., Wanninger, L. (2006): Potential von GPS Navigationsempfangern. In: GPS und Galileo. 66. DVW-Seminar, 21./ 22. February 2006, Darmstadt, Schriftenreihe des DVW, Vol. 49, WiBner Verlag, Augsburg, 2006.
10. Schwieger, V., Zhang, L. (2012): Automatisches geodatisches Monitoring mit Low-Cost GNSS. In: Messtechnik im Bauwesen, Wilhelm Ernst & Sohn Special, Berlin, in print.
11. TEQC (2012): http://facility.unavco.org/software/teqc/teqc.html. Last accessed: February
2012.
12. Ubeda, S. (2008): Ad Hoc Networks: Principles and Routing, in Wireless Ad Hoc and Sensor Networks (ed. H. Labiod), Wiley-ISTE, London, UK.
13. WA1 (2012): http://www.wasoft.de/wa1/index.html. Last accessed: February 2012.
14. Welsch, W., Heunecke, O., Kuhlmann, H. (2000): Auswertung geodatischer
Uberwachungsmessungen. Vol. 2, Reihe Handbuch Ingenieurgeodasie, Herbert Wichmann Verlag, Heidelberg.
15. Weston, N.D., Schwieger, V. (2010): Cost Effective GNSS Positioning Techniques. FIG Publication No 49, FIG Commission 5 Publication. The International Federation of Surveyors, Copenhagen, Denmark.
BIOGRAPHICAL NOTES
Prof. Dr.-Ing. habil. Volker Schwieger
1983 - 1989 Studies of Geodesy in Hannover
1989 Dipl.-Ing. Geodesy (University of Hannover)
1998 Dr.-Ing. Geodesy (University of Hannover)
2004 Habilitation (University of Stuttgart)
2010 Professor and Head of Institute of Engineering Geodesy,
University of Stuttgart Dipl.-Ing. Li Zhang
2002 - 2003 Studies of Geodesy in China (University of Wuhan)
2004 - 2009 Studies of Geodesy in Germany (University of Stuttgart)
2009 Research Associate at Institute of Engineering Geodesy,
University of Stuttgart
CONTACTS
Prof. Dr.-Ing. habil. Volker Schwieger / Dipl.-Ing. Li Zhang
University of Stuttgart
Institute of Engineering Geodesy
Geschwister-Scholl-Str. 24 D
D-70174 Stuttgart
GERMANY
Tel. + 49/711-685-84040 /-84049 Fax + 49/711-685-84044
Email: volker.schwieger@ingeo.uni-stuttgart.de / li.zhang@ingeo.uni-stuttgart.de Web site: http://www.uni-stuttgart.de/ingeo/
© V. Schwieger, Li Zhang, 2012
