Autonomous Permanent Automatic Monitoring System with Robot-Tacheometers

Transcription

1 Autonoous Peranent Autoatic Monitoring Syste with Robot-Tacheoeters Karl FOPPE, Gerany, Wolf BARTH, Gerany and Sebastian PREIS, Gerany Key words: Engineering Survey, Peranent Monitoring Syste, SUMMARY At the geodetic laboratory of the Technical University of Munich the peranent autoatic onitoring of historical architecture has a long tradition. In the year 2003 an autoatic syste for peranent geodetic observations called MoSTUM (Monitoring Syste TUM) was developed especially for the onitoring of the behaviour of buildings. MoSTUM bases on the application of robot-tacheoeters. Furtherore additional sensors for detection of exterior influences like eteorological values are supported. An optional obile data transfer solution by GSM allows the autonoous integration of MoSTUM at arbitrary sites. It offers the possibility to check the state of the object fro the office or even wireless fro any other place in the world. In critical cases MoSTUM can act as an alar syste by sending short essages to obile phones of the responsible engineers. The standard syste consists of a robot-tacheoeter and retro-priss for signalization of the object points. For the installation the choice of the position of the object points is free. Moreover MoSTUM allows changing the configuration of the object points during the period of observations easily. If there is no direct sightline between robot-tacheoeter and the retro priss at iportant object points, an indirect line of sight can be realized by the use of plane irrors. The collected data of the object geoetry and of the exterior influences like acting forces enable an interdisciplinary interpretation in post processing. In this paper the technical design of MoSTUM and the experiences of the first three years of the practical use in Bavarian churches are presented. 1/14

2 Autonoous Peranent Autoatic Monitoring Syste with Robot-Tacheoeters Karl FOPPE, Gerany, Wolf BARTH, Gerany and Sebastian PREIS, Gerany 1. INTRODUCTION INTO THE PROBLEM 1.1 Motivation for the Installation of Peranent Monitoring Systes Historical buildings like churches are objects of great cultural value and convey a certain religious eaning. Very often historical churches show daages like tearing obviously caused by oveents of the building especially subsidence, which are related to non sufficient foundation (Figure 1). A renewing of the foundation is associated with iense expenses. So the goal of investigation is to check the necessity for taking action to assure the stability of the church. First step is the investigation of the actual behaviour of oving of the church by geodetic easureents considering exterior influences on the building. This leads to the developent of a static odel. Under the control of further peranent precise onitoring the static odel becoes verified in a second phase. During this phase of control the expert becoes enabled to define the stability of the building and the necessity of structural alterations. In the third phase the building ay be controlled during the structural works. inster of Schäftlarn lateral cut nave Horizontal Displaceents Subsidence Figure 1: Unequal subsidence leads to horizontal displaceents 2/14

3 1.2 Deands on a Peranent Monitoring Syste A syste for peranent onitoring of historical architecture ust satisfy following requireents: Peranent onitoring of the behaviour of the building for the detection of regular periods (Detection of daily periods, annual periods etc. Repetitions are depending on Nyquist Frequency ) Peranent registration of eteorological values teperature (inside and outside the building), pressure (inside), and huidity (inside) as ean influencing values The concept of the syste should be very flexible to allow changes in the syste configuration or even the expansion of the syste by additional points on deand by inial costs. Soeties it ay becoe necessary to configure the onitoring syste as an alar syste during structural works or because of too big oveents of the building. This option ust be considered during planning of the syste. The required accuracies depend on the expected oveents. For three diensional observation systes in historical buildings usually the standard deviations for each coordinate coponent σ x = σ y = σ z 1 are sufficient. The syste ust be econoically convenient but also deal with the cultural eaning of the building. 1.3 Monitoring Methods for Peranent Observation of Historical Architecture At the beginning of onitoring of an historic building, whose oveents are obviously caused by subsidence, it is absolutely necessary to carry out an engineering levelling of highest precision inside and outside the building. Which kind of observation syste should be installed depends on the rates of expected oveents, directions of expected oveents, and the local conditions in the onitored building. The ost used ethods for peranent onitoring of historical architecture are: 1. Convergence easureents with invar wires Analogue systes (non autoatic) Digital systes 2. Convergence easureents with EDM (reflectorless laser distance easureent by fix installed Distos ) 3. Inclination easureents 4. Vibration easureents 5. Application of robot-tacheoeters Signalisation by retro-reflectors Reflectorless distance easureent 3/14

4 The principle of the onitoring of buildings bases on the experiences with the ore siple ethods 1 to 4. The application of robot tacheoeters is the result of new technical developents. A detailed description of the ethods 1 to 4 is given in [3]. 2. MONITORING SYSTEM TECHNICAL UNIVERSITY MUNICH (MOSTUM) At the geodetic laboratory of the Technical University of Munich (TUM) an autoatic syste for peranent geodetic observations called MoSTUM (Monitoring Syste TUM) was developed. MoSTUM bases on the application Robot-Tacheoeters with Autoatic Target Recognition (ATR). Furtherore additional sensors for detection of exterior influences like eteorological values are supported. 2.1 Hardware Equipent Installed Inside the Building Robot-Tacheoeters Figure 2: Leica-TCA2003 Up to now MoSTUM bases on the application of LEICA Robot-Tacheoeters with Autoatic Target Recognition (ATR), especially the TCA2003 because of its excellent accuracies for easureent of horizontal or vertical directions σ Hz,V 0.15 gon and for electronic distance easureent σ s 1 +1 pp. The robot tacheoeter is ounted on a assive console, which should be installed inside the building at a possibly stable position with direct sight lines to all points of interest (object points). To avoid steel sight lines to the object points the tacheoeter should be installed close to the level of the object points. In our experience the ost stable position is given in the back of the church at a assive wall of the tower. To assure the stability of the console, soe fixed points should be installed as reference. The ost fixed area in an instable church is given close to the ground level. The stability of the reference area is to be controlled fro tie to tie by engineering levelling of highest precision Retro-Reflector Priss All object and reference points are fitted with retro-reflector priss. There are two sorts of priss offered by Leica given with different accuracies: Standard pris GPR1 (<<1) and ining pris GPR112 (<1). In several investigations in the Geodetic Laboratory no significant difference in quality of both priss was detected. As an additional feature the ining pris is equipped with a ebrane to avoid steaing up under wet conditions, which is a big advantage for a peranent installed syste. 4/14

5 Very often the visitors view on the ecclesiastical architecture should not be disturbed by the target priss. So the priss have to be varnished before installation (Figure 3). Figure 3: Leica-Pris GPR112 (<1) and GPR1 (<<1), varnished and installed Meteorological Sensors Meteorological values like teperature and huidity cause expansions or contractions of different parts of the building. So it is necessary to detect these values to estiate the acting forces and the resultant oveents. For detection of exterior influences like eteorological values the integration of additional sensors is supported. A very often used eteorological sensor is the Reinhardt DMT 1MV, which is a cobined sensor for teperature (σ T 0.3 C), huidity (σ H 2%), and pressure (σ P 0.8 hpa). Additional sensors for teperature easureent can be connected to the DMT 1MV because of the integrated data logger (Figure 4). So the cliatic circustances can be recorded at several points of interest in and outside the building. Figure 4: Sensors: Reinhardt DMT 1MV (left) and Leica-TCA2003 (right) 5/14

6 2.2 Diverted Sight Lines Soeties there is no chance to find a position for the tacheoeter inside the building fro which there are direct sight lines to all points of interest. Fro econoic point of view it is seldo possible to install a second tacheoeter. So there was the idea to divert the sight line fro tacheoeter to pris by the use of a plane irror. In the year 2003 the first investigations under practical conditions were realized with plane surface irrors in the church of onastery Schäftlarn. Meanwhile a second syste of irrors and priss is installed in the Jesuits Church in Landshut Basic Principles Denotation: T M P r d 1 t z r 0 d 2 r 0 t z n α position of tacheoeter (intersection of axes) position of plane irror (point of reflection) position of retro pris spatial vector fro T to P distance fro T to M bearing in T to M (with respect to the local coordinate syste) zenith distance in T to M vector of direction TM distance fro M to P vector of direction MP bearing of the outer noral to the irror zenith distance of the outer noral to the irror direction vector of the outer noral to the irror angle of incidence resp. angle of reflexion in M The shifts of the indirectly observed priss are coputed by vector analysis. P M d 2 d 1 M α α n r 0 T r o Figure 5: Basic Geoetrical Principles for Diverted Sight Lines 6/14

7 r 0 is changed over by reflexion in M to in which r 0 = r 0-2 cos α n (1) cos α = cos (t t ) sin z sin z + cos z cos z (2) The vector r fro instruent T to target P of real interest is consisting of two sections r = d 1 r 0 + d 2 r 0 = (d 1 + d 2 ) r 0 2 d 2 cos α n (3) Observable quantities are the direction t and the zenith distance z, both together representing r 0, as well as the total distance (d 1 + d 2 ). The unknown noral of the plane irror is derived fro approxiate coordinates of the geoetrical configuration: n = [(d 1 + d 2 ) r 0 r] / (2 d 2 cos α) (4) α ay be coputed as half of the interior angle at M in the three diensional triangle TMP. Because of the condition that n has to be an unit vector of length one, equation (4) is adjusted to this desired value via odification of the denoinator. The coponents n i of n yield the angles of alignent for the irror: tan t = n 2 / n 1 ; cos z = n 3 (5) Fro geoetrical point of view M is located on an ellipse with T and P as focal points and the sei-ajor axis a = (d 1 + d 2 ) / Deforations To deterine deforations of P, the exact paraeters for the orientation of the irror are not needed, but required to be teporal constant. (To detect possible translations of the plane irrors, there ust be a directly easured pris installed in the surrounding of the irror.) If P is oving to P, in general the point of reflection M at the surface of the irror is displaced. (In a very trivial case P is part of the straight line fro M to P.) Due to the planarity the noral n should not vary at least in theory. A variation dα 1 of the angle α is causing a prolongation of d 1 about (d 1 tan α dα 1 ), whereas d 2 is shortened about the sae aount, concerning its starting point M. So the observed su (d 1 + d 2 ) reains unchanged fro this point view, but is indicating the behaviour of the end point P directly. Differentiation of forula (2) expresses dα = dα 1 + dα 2 in ters of the altered instruent readings and oreover of the warped irror in practice: 7/14

8 -sin α dα 1 = sin (t t ) sin z sin z dt + [cos (t t ) cos z sin z sin z cos z ] dz (6) -sin α dα 2 = sin (t t ) sin z sin z dt + [cos (t t ) sin z cos z cos z sin z ] dz (7) Forula (3) is odified to: r1 + dr1 cos (t + dt) sin (z + dz) 2 2 sin sin r + dr = d' (t + dt) (z + dz) 2 (d' -d1 - d r3 + dr3 cos (z + dz) 1 tan α dα ) 1 cos (α + dα + dα 1 2 cos (t ) sin (t + dt + dt cos (z ) sin (z ) sin (z + dz ) + dz + dz ) ) (8) By now d denotes the total distance fro T via M to P. Obviously the section fro M to P or rather P should be as short as possible with respect to the error propagation. M α + dα 1 + dα 2 α d 1 + d 1 tan α dα 1 Figure 6: Deforation of the Object Point P Observed by Diverted Sight Lines 8/14

9 2.3 MoSTUM Software The onitoring syste is controlled by the MoSTUM software, which was developed at the Geodetic Laboratory (Figure 7). Therefore a PC station with separated power supply has to be established inside the onitored building. It is possible to install ore than one robottacheoeter inside the building. The connection between the tacheoeter stations is realized by D-LAN or W-LAN. The data transfer and the aintenance are realized by a aster station via internet by a separate phone line to office. There is also an alar option. By the use of GSM obile phone odes it is possible to send short essages (SMS) concerning the syste status directly to the responsible user. In case of power blackout the full syste is protected by an independent power supply. The analysis of the tie series is done in post processing in office. easuring period deliberation about strucural interests and the owner SMS: syste status warnings etc. other gaging stations up to 5 easuring and calculation systes D-LAN / W-LAN: datatransfer control station easuring and calculation syste reote aintenance controll calls systecheck in situ tie series of point displaceents GSM evaluation syste tie series Eail ISDN analysis of trends treshholding Figure 7: MoSTUM in actual state MoSTUM is an easily utable syste. There is no proble to change the configuration of the object points or to iport additional points. So even in cases of changes in the behaviour of the building there is no proble to adapt MoSTUM by inial costs. 9/14

10 3. APPLICATION IN PRACTICE 3.1 Syste Setup and Validation in Geodetic Laboratory Upon the installation of the sensors and the object points a first zero epoch easureent has to be arranged to teach the syste. This zero epoch is the reference for all following differential views. The process of teaching the syste is a anual one. Therefore the engineer ais each pris to give the syste horizontal and vertical angels, which are exported by the tacheoeter to a first eory file. After the teaching phase is finished the zero epoch can be started autoatically in MoSTUM. The syste reads out the eory file and oves the tacheoeters according to both stored angels and targets the pris in view exactly by the ATR. Thus the zero epoch will be passed for all priss. Now the reference angels, distances and local coordinates are stored enuerated and dated in the final zero epoch eory file. Motions of priss can now be detected by coparing the coordinates of following easureent epochs with the zero epoch. Therefore a lot of tests were taken in the Geodetic Laboratory of the TUM by installing the syste on a trial and oving the targets in well-defined steps. Also different configurations with the plane irror were tested in the Geodetic Laboratory. All-in it contains oveents of priss and irrored priss, furtherore translations and rotations of the irror itself. For realization a irror was installed on a tacheoeter to perfor oveents in a welldefined way (Figure 8) to get the possibility of testifying the atheatical bases. Figure 8: Mirror test configuration 3.2 Exaple Jesuit s Church in Landshut Meanwhile MoSTUM is installed in three iportant Bavarian churches. Norally one tacheoeter is used. In one case of a very big church MoSTUM is running with two Leica TCA2003 siultaneously. In all installation the syste runs stable and reliable. Figure 9 shows the constellation of the onitoring syste in the Jesuit s Church in Landshut. The tacheoeter is built at a assive wall of the church tower. Two eteorological sensors are installed inside and outside the church. All object points are equipped with priss which are in the sae height as the tacheoeter. For two object points there are four plane irrors to divert the sight lines. So an independend control of the results for the not directly observed points is reached. The full syste fro the tacheoeter s point of view is shown in Figure /14

11 Figure 9: Constellation of the onitoring syste in the Jesuit s Church in Landshut Figure 10: View on the onitoring syste in the Jesuit s Church in Landshut 11/14

12 A view on the screen of the syste in situ is given in Figure 11. The analysis of the tie series is realized in postprocessing in the office. Figure 11: View on the onitoring syste in the Jesuit s Church in Landshut The experiences of the first three years of the practical use of MoSTUM in the Jesuit s Church in Landshut delivers accuracies for the object points of σx= σy= σz 1. REFERENCES [1] Bergann, N., Stepfhuber, W., Zinsberger, St.: Online Monitoring historischer Kirchen it eine Präzisionstacheoeter ittels reflektorloser, direkter oder indirekter Winkel- und Streckenessung, Ingenieurveressung 2004, Zürich 2004 [2] Foppe, K.: Ingenieurgeodätische Überwachungsessungen in und an historischen Bauwerken, Manuskript zu Vortrag an der Hochschule Neubrandenburg, [3] Foppe, K.: Peranent Autoatic Monitoring of Historical Ecclesiastical Architecture, 3 rd IAG Syposiu on Geodesy for Geotechnical and Structural Engineering / 12 th FIG Syposiu on Deforation Measureents, Baden (Austria), May 22 nd -24 th, 2006 [4] Maurer, W., Roßeier, F., Schnädelbach, K.: Deterination of Periodic Displaceents of Buildings and Machines with the Aid of a Laser-Interferoeter, 5th International FIG Syposiu on Deforation Measureents, Fredericton N.B., Canada, [5] Maurer, W., Schnädelbach, K.: Measureent of Displaceents in Buildings With Invar Wires, The Earth and the Universe, a Volue Dedicated to Porfessor Lyssiachos Mavridis on the Occasion of his Copleting Forty-Five Years of Acadeic Activities, Aristotle University of Thessaloniki, /14

13 [6] Schnädelbach, K.: Deforationsessungen it Invardrähten, Veressungswesen und Rauordnung 48, p , 1986 BIOGRAPHICAL NOTES Dr.-Ing. Karl Foppe Acadeic experience: Scientific Eployee at the Geodetic Institute Hanover Scientific Assistant at the Geodetic Institute Hanover Current position: Head of the Geodetic Laboratory at the Chair of Geodesy at the TU Munich Practical experience: engineering surveying, onitoring surveys, dynaic odeling, hydrostatic leveling, inertial survey, calibration of instruents Meberships: Meber of Working Group 5 (AK 5) Veressungsinstruente und Methoden of Deutscher Verein für Veressungswesen (DVW), Meber of Working Group III (AK III) Messethoden und Systee of Deutscher Verein für Veressungswesen (DVW), Head of subcoission Kineatische Meßethoden und Systee Meber of Working Group Kalibrierung Geodätischer Messittel of Deutsche Geodätische Koission (DGK), Dr.-Ing. Wolf Barth Acadeic experience: Current position: Practical experience: Scientific Assistant at the Institute of Astronoical and Physical Geodesy, TU Munich, Work Package Manager in ERS- 1 Satellite Project, Geran Geodetic Research Institute, Dep. 1, Custodian at the Chair of Geodesy of the TU Munich gravity field odeling, processing of satellite orbits, engineering surveying Dipl.-Ing. Sebastian Preis Acadeic experience: Dipl.-Ing. at the TU Munich in 2005 Current position: Scientific Eployee at the Chair of Geodesy at the TU Munich Practical experience: engineering surveying, onitoring survey, deforation easureents, Assisted and Indoor GPS solutions 13/14

14 CONTACTS Dr.-Ing. Karl Foppe Dr.-Ing. Wolf Barth Dipl.-Ing. Sebastian Preis Chair of Geodesy at the TU Munich Arcisstraße 21 D München GERMANY Tel Fax Eails: karl.foppe@bv.tu.de wolf.barth@bv.tu.de preis@bv.tu.de Website: 14/14