WO1993006433A1

WO1993006433A1 - Position measurement by radio frequency transponders

Info

Publication number: WO1993006433A1
Application number: PCT/NL1992/000163
Authority: WO
Inventors: Paul Henri Frans Peteri; Wouter Anton Meijer
Original assignee: Industrial Contractors Holland B.V.
Priority date: 1991-09-24
Filing date: 1992-09-24
Publication date: 1993-04-01
Also published as: AU2668892A; NL9101615A

Abstract

The invention relates to a method and a system for determining a coordinate (x) of the distance between, on the one hand, a transponder (3) which is placed at a height (h) above a plane and, on the other hand, a body placed in said plane. The transponder (3) is irradiated by radiation having a primary frequency (fp) and, in response thereto, emits radiation having a secondary frequency (fs). The body interacts with a frequency source and transmitting coil (1) for the primary frequency (fp) and carries at least a first measuring coil (2a; 2b) attached to the body and a second measuring coil (2b; 2a), at a fixed distance therefrom, for the secondary frequency (fs). From the first and the second measuring coil, a first measuring signal (Ux) and a second measuring signal (Uh), respectively, are taken off which depend, according to a first and a second set of known relationships [f(x,h)] and [g(x,h)], respectively, on the distance coordinate (x) and the transponder height (h). From the second measuring coil (2b; 2a) a second measuring signal (Uh) is taken off which depends, according to a second set of known relationships [g(x,h)], on the distance coordinate (x) and the transponder height (h). By making use of calibration tables in which these relationships are defined and which are stored in one or more memories, the distance coordinate (x) in the plane is derived from the measured values obtained.

Description

Title: Position measurement by radio frequency transponders

The invention relates to a method and a system for determining a coordinate (x) , in a plane, of the distance between, on the one hand, a transponder which is placed at a certain height (h) above said plane, is irradiated by radiation having a primary frequency (fp) and, in response thereto, emits radiation having a secondary frequency (fs) , and, on the other hand, a body positioned in said plane, which body interacts with a frequency source and transmitting coil for the primary frequency (fp) and carries at least a first measuring coil attached to the body, and a second measuring coil, disposed at a fixed distance therefrom, for the secondary frequency (fs) .

Because of continuing automation, more and more systems appear in which a moving or mobile object interacts in one form or another with a device in a fixed position, without human intervention. In the case of an interaction of this type it is important that the mobile object and the fixed device are positioned effectively with regard to one another. This positioning can take place in two different ways: either the mobile object positions itself with respect to the fixed device in such a way that the interaction can take place reliably, or the fixed device "adapts" to the movable object by said device being provided with movable means to enable the interaction to proceed reliably. In both cases, the relative positions of the mobile object and the fixed device with regard to one another must be known.

In the first case, the mobile object (hereinafter called "vehicle") must know the position of the fixed device with regard to itself. This is the case, for example, if the vehicle always knows its position in the (2D) space with sufficient accuracy and, at the same time, knows the position in said space of the fixed device. This implies that the vehicle is provided with a "map" of the working space, and also with navigation means to keep track of its position in the working space. If the navigation system, and thus the position-measuring means, are sufficiently accurate, no separate position-measuring system is required on the vehicle to enable the vehicle to be positioned correctly. In the second case the vehicle is placed in the vicinity of the fixed device. Possible examples are an automatic vehicle which navigates, but which knows its position insufficiently accurately for the purpose of the interaction in question, or a manned vehicle whose positioning accuracy is limited. The fixed device will then have to determine the position of the vehicle with regard to itself and will, on this basis, control its means of motion in such a way that effective and reliable interaction is possible. The positioning system will have to be accurate to a greater or lesser degree depending on the type of interaction.

In both cases, a position-measuring system is required which can determine a relative position over a limited distance (for example 0.5 m or less) with the maximum possible accuracy (order of 1 mm) . Depending on the application, a measurement in two or three dimensions will be desirable.

The invention relates to those applications where measurement accuracies of the order of 1 cm or better are required. Such an application is, for example, an automated filling station for private cars. In this case, removal of the tank cap and filling the tank are performed by a robot arm. To this end it is necessary that the filling installation is able to determine accurately the position of the tank cap. Another example is the performance of an operation (for example mounting a component) by a fixed station on a workpiece attached to an unmanned vehicle. In these cases, the fixed station is equipped with a position- measuring system. It is also possible to mount the measuring system on the vehicle. This could, for example, be the case if workpiece and machine from the previous example are reversed: the workpiece is fixed and the machine sits on the vehicle. Another example is the lifting of a pallet from a stacked storage system. In this case the position of the pallet in the rack is not known exactly. If the vehicle can determine the relative position of the pallet, the forks can be controlled accurately. In this example the measuring system is placed in the mobile part instead of the static installation.

A position-measuring system is of interest not only for interaction between a fixed and a mobile object: a system of this type may also be of use in the case of unmanned vehicles. This is exemplified by an intelligent unmanned vehicle system which navigates over a grid of marking elements in a floor. Accurate measurement of the relative position of the marking elements with regard to the vehicle enables the vehicle to execute a pre-planned route accurately and reliably.

This summary is not exhaustive, many other examples can be given where it is advantageous to measure a relative position. A contact-less local position-measuring system which is based on the detection and localisation of a transponder, corresponding to the system according to the present invention, is known in practice. In this case, a transponder collects magnetic energy which has been emitted by the primary antenna of the measuring system. The technology is based on identification systems such as those produced by various manufacturers: a (passive) transponder is actuated by an electromagnetic field (radio frequency, of the order of 100 kHz) ; as a result of the transponder modulating the field strength in a coded manner, the transponder code can be read out. In this way it is possible, for example, to identify goods which are marked by a transponder. An important advantage of a transponder compared to, for example, a bar code is the fact that the read-out is not affected by dirt or wear. Furthermore, a reasonable spacing is possible between transponder and reader unit (for example 1 m) . The addition of measuring coils to the system has achieved a further development of the RF transponder technology, from identification to localisation. For this purpose, a modified transponder is used which is provided with a second coil and a frequency divider, as a result of which a signal is "returned" at a lower frequency which differs sufficiently from the primary frequency supplied to the transmitter antenna. As a primary frequency, for example, 120 kHz is used, the transponder returning a signal of 30 kHz. In this way it is possible to collect selectively, by means of filtering, the 30 kHz signal in the measuring coils without the signal becoming unusable due to the (much stronger) 120 kHz field of the primary antenna. For the purpose of measurements in one dimension (x) , the voltages induced by the transponder over two measuring coils coupled in opposition are used as a measure for the relative position of the transponder in the x-direction (see Figure 1) . By using a second set of coils it is also possible to obtain a y-measurement.

The biggest drawback of this known system is the height dependence of the x- and y-measurements. With increasing height h between transponder and measuring and transmitting coils, the voltage induced in the coils decreases rapidly. The differential voltage between the measuring coils, which is a measure for the relative positions, therefore also decreases. In addition to the measured signal generated in the measurement coils, a measured signal is also generated by the transponder in a transmitting coil. This also decreases with increasing height. By using this "height signal" as compensation for the purpose of amplifying the x- and y-signals ("automatic gain control") , the height dependence can be somewhat reduced. Because, however, the h-signal and the x- and y- signals do not depend on h in the same way, the accuracy of x and y will remain limited if h varies more widely (see Figure 2) . This is a fundamental limitation of the existing localisation system.

Moreover, the height signal is not constant at fixed h, but depends on x and y. As Figure 2 shows, the inaccuracy of the measured value increases as the measured value itself increases, which limits the useful range. All these properties of the existing localisation system have the effect that correction of the x- and y-measurement by means of the so-called "automatic gain control" by means of an h-signal taken off the transmitting coil is not really suitable as a measuring system, but only, at best, as a global localisation system having an accuracy of a few centimetres.

The object of the present invention is to overcome these limitations by explicitly taking into account the abovementioned non-linearities and relationships and, for this purpose, is characterised in that, from the first measuring coil, a first measured signal (Ux) is taken off which depends, according to a first set of known relationships [f(x,h)], on the distance coordinate (x) and the transponder height (h) , in that, from the second measuring coil, a second measured signal (Uh) is taken off which depends, according to a second set of relationships [g(x,h)], on the distance coordinate (x) and the transponder height (h) , and in that the distance coordinate (x) in the plane is derived, from the measured values obtained, by making use of calibration tables in which these sets of relationships are defined and which are stored in one or more memories.

Several advantageous embodiments of the present invention will appear from the subsidiary claims.

The present invention will now be explained in more detail for the case of position measurement in one direction, by reference to the attached drawing. In this:

Figure 1 shows in diagrammatic form the arrangement known in practice of a transmitting coil and measuring coils coupled in opposition, for the purpose of measuring the distance in one direction;

Figure 2 shows diagrammatically the relationship known in practice between the horizontal displacement x and the measured signal Ux observed at the output of the measuring coils according to the configuration of Figure 1;

Figure 3 shows diagrammatically the relationship, for various transponder heights, between the horizontal displacement x and the measured signal [f(x,h)] observed a the output of the measuring coils and the measured signal [g(x,h)j observed at the output of the transmitting coil;

Figure 4 shows the first quadrant from Figure 3, also illustrating diagrammatically the method according to the invention;

Figure 5 shows a block diagram of a system for performing the method according to the present invention; and

Figure 6 shows a cross section of a transponder provided with a metal housing.

The following explanation starts from the known configuration which is shown diagrammatically in Figure 1. Distinction must be made herein between a transmitting coil 1, measuring coils 2a and 2b and a transponder 3. The position-measuring system according to the invention makes use - as does the known localisation system - of two signals for a 1-dimensional determination of location (x-measurement) ; the amplitude of the measuring signal intercepted by the transmitting coil 1 and the amplitude of the measuring signal at the measuring coils 2 for the x-direction. The amplitude Uh of the measuring signal at the transmitting coil 1 is, to a first approximation, a measure for the distance (h) between this coil and transponder 3, while similarly the value of the signal Ux at the measuring coils 2 is, to a first ap¬ proximation, a measure for x. In reality, Ux and Uh do not depend equally on h. Furthermore, Uh also depends on x (thus it is not a constant for a given h for all relevant values of x) . Both Ux and Uh therefore depend on x as well as h: Ux = f (x,h) Uh = g (x, h)

Because of the dependence of Uh on x, h cannot be derived directly from Uh. In that case the problem could be reduced to substituting h in f(x,h), whereupon x could be directly determined from Ux. However, assuming that f(x,h) and g(x,h) are known, it is still a matter of two equations with two unknowns, and it should therefore be possible to establish x (and h) with reasonable accuracy. Figure 3 shows that f(x,h) consists of a fan of curves in the (x,U)-plane - after all, for each h there is a corresponding curve f(x,hl), f(x,h2) etc. All these curves pass - at least in theory - through the origin because the measuring antennae must give the value Ux=0 for x=0, irrespective of the value of h. The smaller h is, the steeper is the curve f(x). Similarly, g(x,h) consists of a set of curves in the (x,U)-plane. The larger h is, the lower the curve g(x) comes to lie. Because an algebraic solution of the system of equations is not possible, a numerical method must be used. A method is described in the following text by reference to Figure 4. To start with, the procedure followed is that, for example, four curves of f(x) are measured, namely f(x,hl), ..., f(x,h4), and also four curves g(x) , i.e. g(x,hl) , .., g(x,h4) . These curves are stored in tabular form in the memory of the microprocessor. Here hl=hmin, and h4=hmax, hmin and hmax being the limits within which the height can vary. The determination of x (and h) based on two given measured values Uh (=g(x,h)) and Ux (=f(x,h)), can now be carried out as follows. The measured value Uh affords, with the help of the tables for g(x,hmin) and g(x,hmax) , a maximum and a minimum value for x, xmax and xmin respectively, where xmax therefore corresponds to hmin and xmin to hmax. With the help of the tables f(x,hmin) and f(x,max), the function values associated with the values xmax and xmin, respectively, can now be determined. These are shown in the (x,U)-plane of Figure 4 by reference number 4 and 5, respectively. The points form the extreme values of the set of solutions in the (x,U)-plane, formed by the combinations (xmax, hmin) and (xmin, hmax) . Intermediate x-values (with associated intermediate h-values) may be determined analogously by making use of the tables g(x,h2) and g(x,h3) and by subsequently combining the x-values found with f(x,h2) and f(x,h3) , which are indicated by 6 and 7, respectively. In this way, four points of the "solution curve" k(x,Uh) , associated with the measured Uh, have now been determined. The curve can now, for example, be approximated by an interpolation of the order of (n-1) , if n points of the curves have been found. The actual x-value must lie on the curve 8 thus found, but must also satisfy the measured value Ux. The intersection of the horizontal line U=Ux with the curve found earlier thus affords the actual x sought, from the set of possible x-values. Once x has been determined, h can be determined subsequently by means of one or other interpolation method.

This method can be extended to the three dimen- sional case, the arrangement being extended by measuring coils for the y-direction. This results in a system of three equations with three unknowns x, y and h.

In addition to this software-based determination of the measured values, which provides much greater accuracy than is possible with the existing hardware-based system, the acquisition of the measurements (the data acquisition) is improved fundamentally, while the electronics have been simplified. That is to say, it is no longer necessary to rectify the signal, and the phase-detection circuit which was necessary to determine the sign (of the x- or the y- value) is omitted. These advantages result from making use of the properties of discrete Fourier sums for periodically sampled signals for the purpose of determining signal amplitudes U: N-l

U(kf) = Σ u(n) . exp(-j2πnk/N) n=0 The number of samples u(n) and the sample time are adjusted to the known measuring frequency in such a way that k=l in the above formula. Because the frequency is known, the measuring method, in essence, comes down to determining the amplitude U(f) from a weighted average of a number of samples. In addition to the amplitude, the phase with regard to the start time of the measurement is determined, from the samples, by this method. Depending on the configuration used of the measuring coils, the phase may or may not be of interest, but it is available in any case as a result of the amplitude determination by means of Fourier sums. An additional advantage of this measuring method is the fact that the (weighted) summation of a number of measured values automatically reduces the measured noise, in comparison with one single measurement.

The present invention, apart from the greater accuracy, also has other advantages over existing locali¬ sation systems. An advantageous hardware configuration, shown in Figure 5, of a system according to the invention is simplified compared with the known localisation systems, as the use of a fast A/D converter 9, of a microprocessor 10 and of software has made it possible to sample the secondary signal directly without rectification taking place. In addition, the phase-detection circuit which determines the sign of the measured value in the existing system is omitted. In the existing system the task of the microcontroller is limited to intercepting and buffering the final measured value and transmitting it to the control of the vehicle or machine. The basis of the system of the invention was the active use of the possibilities created by software, which has produced a system which can deal "intelligently" with the measured values. The most important advantages thereof have already been discussed above. Other advantages are - simpler adjustment by using programmable amplifiers 11, which make it possible to adjust the pre¬ amplifiers automatically (software-controlled) - "self-setting" filters 12 to counteract changes which are a consequence of fluctuations in ambient temperature: the filter is adjusted by means of software in such a way that the response is at a maximum

- flexibility: all the methods described above of filtering, averaging, interpolation, may be further adapted or refined if this is required or if new insights should suggest this. If another transponder with other properties is used, the memory of the microprocessor should be provided with new calibration curves associated with this trans¬ ponder. This should also be done if a "known" transponder is provided with a metal, for example steel, housing (Figure 6) , in order to minimise the influence of metal, in the near vicinity of the transponder, on the measured results.

In addition to the antenna configuration described herein (Figure 1) , in which one set of measuring coils is used per direction, other configurations are also possible. Instead of two coils coupled in opposition, two loose measuring coils per dimension may also be used. Particularly for the purpose of enlarging the range in one direction, use may be made of an array of loose measuring coils which are polled one by one via the analog multiplexer 13 (Figure 5) . The characteristics of the measured values Ux as a function of the distance x have the same shape, in the case of an array of loose measuring coils, as those of g(x,h) from Figure 3. An important advantage of loose coils is that two loose coils afford two measured values Uxl(x,h) and Ux2(x,h) which contain more information than the combined differential signal of the coils coupled in opposition. This offers the possibility of determining the transponder height and (x) from these two x-signals without Uh(x,h) . The phase of the signals in this case is no longer significant and therefore needs no longer be determined. Variations of antenna configurations with measuring coils, which can be achieved within the system of the invention and which thereby make it possible to measure more unknowns (x,y,h, transponder orientations, temperature, transponder response) because more equations are available, can easily be derived based on the above.

Claims

CLAIMS 1. Method for determining a coordinate (x) , in a plane, of the distance between, on the one hand, a transponder (3) which is placed at a certain height (h) above said plane, is irradiated by radiation having a primary frequency (fp) and, in response thereto, emits radiation having a secondary frequency (fs) , and, on the other hand, a body positioned in said plane, which body interacts with a frequency source and transmitting coil (1) for the primary frequency (fp) and carries at least a first measuring coil (2a; 2b) attached to the body, and a second measuring coil (2b; 2a) , disposed at a fixed distance therefrom, for the secondary frequency (fs) , characterised in that from the first measuring coil (2a; 2b) a first measuring signal (Ux) is taken off which depends, according to a first set of known relationships [f(x,h)], on the distance coordinate (x) and the transponder height (h) , in that, from the second measuring coil (2b; 2a) , a second measuring signal (Uh) is taken off which depends, according to a second set of relationships [g(x,h)], on the distance coordinate (x) and transponder height (h) , and in that the distance coordinate (x) in the plane is derived, from the measured values obtained, by making use of calibration tables in which these sets of relationships are defined and which are stored in one or more memories.

2. Method according to claim 1, characterised in that the transponder height (h) is also derived.

3. Method according to claim 1, characterised in ha based on the second measuring signal (Uh) , for each value, corresponding with a calibration table, of the transponder height (h) the value of the distance (x) associated with said value (h) is determined from the calibration table of [g(x,h) ] associated with the particular value (h) , in that subsequently the values of [f(x,h)] associated with these combinations of values for (x) and (h) are determined from the calibration tables in question, in that, based on the values found for [f(x,h)], a curve [k(x,Uh) ] is determined by means of interpolation, which curve is associated with the second measured signal (Uh) in question, and in that, from the curve thus obtained, the associated value of x is derived for the value (Ux) of [k(x,Uh)].

4. Method according to claim 3, characterised in that n values of [k(x,Uh)] are determined and, for the purpose of the interpolation between said values of [k(x,Uh)], in order to determine the curve for [k(x,Uh)], an interpolation of the order of (n-1) is performed.

5. Method according to claims 2 to 4, characterised in that, based on the distance (x) already determined and the measured value (Uh) , the transponder height (h) is derived by interpolation in the calibration tables for [g(x,h)J.

6. Method according to claim 5, characterised in that_/ based on the distance (x) found, m values of [g(x,h)] are derived from the calibration tables in question and in that, for the purpose of the interpolation between the values (h) , an interpolation of the order of (m-1) is performed.

7. Method according to one of the preceding claims, characterised in that the amplitudes of (Ux) and (Uh) are determined by means of Fourier sums.

8. Method according to one of the preceding claims, characterised in that a determination of the distance is also performed in a second direction (y) , based on a third measured signal (Uy) .

9. Method according to one of the preceding claims, characterised in that the first (2a; 2b) and second (2b; 2a) measuring coil are coupled in opposition, from which the first measuring signal (Ux) is taken off, and in that the transmitting coil (1) of the primary frequency (fp) is used as the measuring coil for the secondary frequency (fs) , from which the second measuring signal (Uh) is taken off.

10. System for implementing the method according to Claims 1 to 9, comprising a transponder (3) which, when irradiated by radiation having a primary frequency (fp) , in response thereto emits radiation having a secondary frequency (fs) , a body having a frequency source and transmitting coil (1) for emitting radiation having the primary frequency (fp) and at least one first measuring coil (2a; 2b) which is attached to the body, and a second measuring coil (2b; 2a) , disposed at a fixed distance therefrom, for the secondary frequency (fs) , the body further comprising band-pass filters (12) , amplifiers (11) , a multiplexer (13) , an analog/digital converter (9) and a central processing unit (10) with memory, the first measuring coil (2a; 2b) and the second measuring coil (2b; 2a) each being connected to a band-pass filter (12) , which filters, of the radiation received, pass a narrow frequency band situated around the secondary frequency (fs) and are connected to the inputs of amplifiers (11) whose outputs are coupled with the inputs of a multiplexer (13) , the output signal of the multiplexer (13) being used as the input signal of the analog/digital converter (9) whose digital output signals are supplied to the inputs of the central processing unit (10) .

11. System according to claim 10, characterised in that the amplifiers (11) are programmable and have inputs which are connected to the central processing unit (10) in order to make it possible to adjust the preamplifiers under the control of the central processing unit (10) .

12. System according to claims 10 or 11, characterised in that the filters (12) also have an input which is connected to the central processing unit (10) , in order to enable them to be set by the central processing unit (10) in such a way that the response of the filter (12) is as high as possible.

13. System according to one of claims 10 to 12, characterised in that the primary frequency (fp) and the secondary frequency (fs) have a frequency range of 100 to 140 kHz and 20 to 40 kHz, respectively.

14. System according to one of claims 10 to 13 for implementing the method according to Claims 1 to 9, characterised in that the body carries a third measuring coil for the measurement of (Uy) , which is disposed in such a way that it forms an angle with respect to the first measuring coil for the measurement of (Ux) .

15. System according to one of claims 10 to 14, characterised in that the first (2a; 2b) and second (2b; 2a) measuring coil are coupled in opposition and in that the transmitting coil (1) for the radiation having the primary frequency (fp) is used, at the same time, as a measuring coil for the secondary frequency (fs) .