FR2547917A1 - METHOD FOR PROCESSING THE OUTPUT SIGNAL OF AN OPTICAL SENSOR OF TERRESTRIAL HORIZONS OF A GEOSTATIONARY SATELLITE - Google Patents

Abstract

PROCESS FOR DETERMINING THE ANOMALY-COMPONENT IN THE OUTPUT SIGNAL OF A TERRESTRIAL HORIZON SENSOR FOR A GEOSTATIONARY SATELLITE. THE HORIZON SENSOR 1A INCLUDES A CONSTANT FREQUENCY DRIVEN CHOPPER DISK 3A WHICH REPRODUCES THE IMAGE OF OPPOSING EARTH HORIZONS ON A COMMON INFRARED DETECTOR 9A. IN THE EVENT OF A DIFFERENCE IN THE THERMAL RADIATION OF THE TWO HORIZONS, THE ANOMALY COMPONENT CONTAINED IN THE OUTPUT SIGNAL OF THE DETECTOR CAN BE DETERMINED BY MEASURING THE DRIFT SIGNALS FOR DIFFERENT AMPLITUDES OF THE CHOPPER DISK. AS THE ANOMALY DEPENDENT COMPONENT IN THE DETECTOR OUTPUT SIGNAL DEPENDS ON THIS AMPLITUDE OF THE CHOPPER, ABSOLUTE CORRECTION VALUES FOR THE MEASURED VALUES CAN BE DETERMINED ACCORDING TO THE ANOMALY COMPONENT.

Description

Method for processing the output signal of an optical sensor

  BACKGROUND OF THE INVENTION The invention relates to a method for processing the output signals of a terrestrial optical sensor of a geostationary satellite, said sensor comprising an optical input, a chopper disc arranged in the focal plane of the input optics, moved in a reciprocating motion with a determined amplitude (chopper amplitude) and a determined frequency (chopper frequency) and having a diameter substantially corresponding to image of the earth, a secondary optics and a detector whose output signal is amplified and demodulated at the chopper frequency into a sensor drift signal which is a measure of the deflection angle of the line of sight of the terrestrial horizon sensor with respect to the line of

  satellite connection / center of the earth.

  For the stabilization of orientation of geostationary satellites, control signals are used which indicate the deviation of a line of sight attached to the satellite with respect to the satellite connection line / center of the earth For this purpose, for two axes on uses, among other things, terrestrial optical sensors which can be classified in the class of zero-tracking sensors. Such a terrestrial horizon sensor operates in the infrared zone and is based on the principle of the mechanical chopper. earth is concentrated by a germanium objective lens and falls on a chopper disk in the image plane of the lens This chopper disk has a diameter corresponding substantially to the image of the earth and is moved in a movement of -and-comes with a determined amplitude, amplitude-chopper, and at a specific frequency, frequency-chopper Light, passing through convergent optics and interrupted by the chopper disk, two opposite terrestrial horizons and alternately released at the chopper frequency, is directed via a secondary optics constituted by a spherical mirror segment and a prism, and via a spectral filter for the infrared zone, on a detector, for example a

pyroelectric detector.

  The output signal of the detector is amplified and then demodulated at the chopper frequency. When the light energy, absorbed by the detector, of the two terrestrial horizons, is equal, the demodulation gives a zero signal. If this is not the case, for example because the line of sight of the terrestrial horizon sensor does not coincide with the satellite connection line / center of the earth, then the output signal of the detector depends on the difference of the amounts of light absorbed from the two horizons. difference is a measure of the angle of deviation of the line of sight of the earth horizon sensor from the satellite / center line of

Earth.

  With such a sensor of terrestrial horizons, the angle

  This deflection can only be indicated on one axis. For triaxial stabilization, two earth horizon sensors of this type are required.

  The chopper disk is driven back and forth at a chopper frequency of about 40 Hz and

  amplitude of about 1/17 of the disc diameter of about 1 mm.

  Compared to other sensor systems, terrestrial horizon sensors of this type operating according to the chopper principle have several advantages: firstly, the chopper or mechanical vibrator is a very simple spring-mass system which oscillates with its natural resonant frequency It therefore requires neither drive motor nor complicated angular reading. On the other hand, channel-minced infrared radiation is directed only at a detector. This eliminates the problem of

  the balancing of several detectors as well as the problem

me of aging.

  The measurement range of such a terrestrial horizon sensor is about + 1. The zero deviation mentioned as a result of the difference in radiation of the opposite terrestrial horizons is certainly not very great, but can, however, in the extreme case, rise to approximately + 16% of the measuring range of a given degree. An attempt is made to keep this signal as small as possible.

  erroneously caused by the radiation anomaly, i.e., the anomaly component in the sensor drift signal.

  This effort is understandable if one thinks that geostationary satellites should be used to establish

  television links by radio-controlled herzian beams with the earth.

  The object of the invention is to develop a method for the correction of the anomaly component in the sensor drift signals, which correction can be carried out in a simple manner from the drift signals themselves.

of the sensor.

  This result is achieved according to the invention in that to reduce the errors that are caused by the differences in heat of the edges of the earth (anomaly) and which are reflected in the sensor drift signals, it operates it to different amplif 25 deshacher, that for these chopper amplitudes the drift signals obtained are compared with the values of the normal characteristic curves for chopper amplitudes without anomaly for a common deflection angle, in the case of nonconformity of the corresponding values for an angle the values measured until a concordance of the corresponding values appears and the measured values of these values of

correction (component-anomaly).

  The invention is based on the fact that the anomaly component of the drift signal of the sensor only depends on the anomaly V and the amplitude-chopper A, so that if

254791?

  If this amplitude-chopper A is modified to give it the value A 1, the anomalous component UDA varies proportionally with the amplitude and goes from UDA to UD Al. Thus: in the event of a modification of the amplitude and 5 consecutive variation of the component-anomaly, we can find directly the anomaly of the earth and also the component anomaly of the drift signal of the sensor, this is explained in the following: In case of uniformly hot earth, both hori10 zons or edges of the earth radiate the same energy, that is to say that we have:

EL E = ER = E (1)

  In this formula, E is the average energy of the radiation, EL the energy of the radiation of the left edge of the earth and ER the energy of the radiation of the right edge of the earth If the energies of radiation of the left and

  are not equal, that is to say, if EL ER, this state is referred to as terrestrial anomaly.

  For the other considerations, it is assumed that the average energy is constant, that is to say that EL + ER = 2 E = constant (2) The radiation energy different from the two edges of the earth can be written :

EL = E

ER = E + A (3)

  In these formulas, Ar Erésente the difference of the radiation energies relative to the average value As the anomaly of the earth, V denotes the ratio of these two radiation energies: -5

EL E (1-A) 1 A 1-V

ER = E (1 + A) = 1 + A = -V (4)

E 1 + V

R

  The anomaly V oscillates in practice in a position between 1, 5/1 to 1 / 1.5.

  If there is no anomaly, the characteristic line of the earth horizon sensor can be roughly represented by the following formula (see Figure 1)

2 KA

  U Da = K A sin (a 90) (5) This formula is valid for the KA range

-1 A + 1

  In this formula: U Da denotes the drift signal of the sensor which depends on the deflection angle α and which is measured as electrical voltage; K represents a constant proportionality factor which is determined by the geometry of the sensor; A represents the chopper amplitude; a represents the deflection angle of the target line of the terrestrial horizon sensor; KA represents a mechanical transmission factor,

also constant.

  FIG. 1 represents the curve of the characteristic line UD. The whole of the characteristic line C130 has an angular zone of approximately +18.

  for the deflection angle of the sensor is about + 1.

  The curve of the characteristic line in this indicated measuring range is shown in FIG. 2. In this figure, it can be seen that for small angles of deflection the characteristic line can be linearized by the following formula:

254791?

x

UD = K KA (6)

  Dcx 6 This linearized characteristic line U Da x is presented as a dashed line in Figure 2. If the two edges of the earth radiate different energies, the whole of the UD characteristic line consists of the undisturbed upper characteristic line UDC and a component-anomaly UDA: 10

UD = UDA + UD (7)

  The anomaly component EDA is then 1-V UDA = K A 1 + v (8) On the basis of formulas 5, 7 and 8, the characteristic line UD of the terrestrial horizon sensor is thus obtained for

1-V 2 K KA O + I (9)

  UD = K + K A sin (90) -1 _ and for small values a UD = K A 1 -V + K KA a (10)

DI + V

  From formulas 8 to 10 it follows that, as indicated above, the anomaly component of the sensor drift signal depends only on the V anomaly and the A-beam amplitude. If the sensor drift signal is now measured for two different chopping amplitudes A and A 1, the anomalous component UDA at least for small angles becomes independent of the anomaly V For small angles, the anomalous component has 35 AU (A

DA (UD UD) (11)

  A-A 1 and for the characteristic line V Da, simplified, corrected, and valid for small angles of deflection a x A U Du = UD (UDUD (12) D 5 to D A-A 1 For the terrestrial anomaly V,

(A-A 1) K- (UD-UD 1) 13)

V = (13)

V =

(A-A 1) K + (U D-U D 1)

  from which it appears that the terrestrial anomaly only depends on the amplitude and drift signals of the sensor for the different amplitudes-chopper For large

  angles of deviation, one must of course use the complete formula for the characteristic line.

  From this, it is only from the values of the sensor drift signals for different chopper amplitudes that we can draw conclusions for a corrected characteristic line of the sensor in which

  takes into account the anomaly of the earth.

  The correction of the characteristic line is explained graphically in Figures 3 and 4.

  FIG. 3 shows two characteristic lines U Da and UD 1a with no anomaly, the characteristic line

  U Da applies to the amplitude A, and the other line is characteristic of the amplitude A 1. Two UD and UD 1 characteristic lines are also drawn which apply to an anomaly of V = 1.5 / 1. and are associated with the amplitudes A and A 1.

  In a first case, the two voltage values U and U 1 are measured for the chopper amplitudes A and A 1. As can be seen from FIG. 4, at these voltage values correspond the deflection angles marked with a cross on

  the two normal characteristic lines without anomaly.

  In a second case, the voltage values are measured

  U 'and U 1 for the amplitudes A and A 1; these values are 35 of a cross on the two lines also marked with a cross on the two lines

254791?

  UD and UD normal ticks 1 The first values on the normal characteristic line corresponding to the voltage values U and U 1 are still in the area where the two normal characteristic lines overlap, while the values for U 'and U' 1 are no longer in the overlap area Since the values measured each time-can not be associated with the same deflection angle on the normal characteristic lines, there is an anomaly In the linear zone of the characteristic line, 10 UDA anomaly can be calculated directly from equation (11) h The anomaly signal corresponding to terrestrial anomaly V then results from equation (13) For U 'and UV values, the values must be used. forms or forms of complete curves of the different characteristic lines In all cases, the calculation operations may be interpreted in such a way that the corresponding characteristic lines without and without The anomalies are shifted together at a zero point and overlap in pairs. By moving the different characteristic lines, the different UDA and UDA 1 anomalous components are obtained. As the optics of the terrestrial horizon sensor causes other nonlinear distortions of the characteristic line or the transfer function of the horizon sensor, the correction for large deviation angles is not determined by the characteristic line given mathematically, but by the measured curve actually. nothing changes

in the principle of correction.

  Different possibilities can be envisaged for implementing the method indicated above for the correction of the drift signals of the sensor. Thus, for example, the chopper disk can be displaced intermittently with different amplitudes. By modifying the amplitude of the chopper, one interrupts of course the measurement process so that no continuous measurement is possible.

  For the correction of the anomaly, it is also possible to

  -9 to read the drift signals of two horizon sensors

  The optics can then either be performed in duplicate or be used in common for the two chopper disks via a ray splitter.

  An amplitude modulation of the chopper amplitude and the interpretation of the sensor drift signal corresponding to each chopper amplitude by means of a synchronous demodulator are particularly advantageous. In this case, the anomaly signal and the correction of Angular deviation is generated directly If the modulation frequency is higher than the bandwidth of the sensor drift signal, the amplitude modulation of the drift signal can be suppressed. In this way, a continuous measurement of the signal for the deviation angular and anomalous is possible Through the use of amplitude modulation, a perfectly self-contained correction and at the internal level 'of the terrestrial horizon sensor is possible The output signals that the terrestrial horizon sensor delivers to the calculator for the stabilization of orientation of the geostationary satellite are therefore already corrected and no longer need to be corrected

in the calculator.

  According to an advantageous characteristic of the invention, the drift signals UD, UD 1 are measured for two different chopper amplitudes A, A 1, and the UD value, measured for the largest amplitude, of the anomaly component is decreased.

U A (UU 1

DA A-A (UD UD 1)

  and this corrected value is associated with the corresponding deflection angle on the normal characteristic line.

  According to another feature, the chopper amplitude is modulated with a frequency above the bandwidth of the sensor drift signal and the measured drift signal is synchronously demodulated, the demodulated component is compared with the values of the sensor. normal characteristic line and one decreases the measured drift signal

  an error signal corresponding to the comparison.

  The invention will be better understood with the aid of the description of an embodiment taken as an example, but not limited to, and illustrated by the appended drawing, in which: FIG. 1 represents the curve of the characteristic line of FIG. a sensor of terrestrial horizons without ano10 malia over the entire detection zone; FIG. 2 represents the curve of the characteristic line of a terrestrial horizon sensor without anomaly for angles of deviation of between + 1; FIG. 3 represents the curve of two characteristic lines without anomaly for two different amplitudeshacheur and the curve of these characteristic lines with anomaly; Figure 4 shows a detail of Figure 3 on an enlarged scale; FIG. 5 represents the curve of different characteristic lines of the terrestrial horizon sensor with an amplitude modulation of the chopper amplitude; FIG. 6 shows the curve of the characteristic lines for an average value of the analog component of the sensor drift signal with an amplitude modulation of the amplitude of the chopper disk; Figures 7 to 10 show respectively block diagrams for different embodiments of

terrestrial probes.

  In the block diagrams of FIGS. 7 to 10, the same references are used for the identical or identical-action components, to which, however, we have added

the letters a, b, c and d.

  A terrestrial horizons optical sensor comprises an input optic 2a which is sensitive in the infrared region. In the focal plane of optics 2a is installed a chopper disk 3a which is driven by an optical system. 4a, for example a resonant magnet and spring system The chopper disk is driven at a constant frequency of, for example, 40 Hz with a given chopper amplitude. The chopper amplitude is switched at a given rate between two different amplitudes A and A 1 This rhythmic switching as well as all the other interpretation systems are controlled by a clock 5a. For this purpose, the chopper amplitude is measured with a sensor 6a which is not described in detail, the value effective is sent to a summing stage 7a to the other input of which is applied the set value The drift signal as well as the clock signal are sent to a regulator 8a which monitors accordingly the chopper drive system The chopper frequency is applied as a clock signal to the clock 5a, which also balances the other circuit elements concerned, in particular the balancing with a calculator

  Ca for orientation stabilization.

  The infrared radiation transmitted periodically by the chopper falls on an infrared detector; the output signal is amplified. This is shown schematically by the block 9a. The amplified signal of the detector is delivered in time to a demodulator with low-pass filter and ampli fi erator 10a. At the output of the demodulator 10a a series of clocked values appears. UD and UD voltage signals 1 which correspond to the drift signals of the sensor having the two amplitudes A and A 1. These successively determined drift signals are, on the one hand, applied to a somatractor member 11a and, on the other hand, to a correction circuit

  12a in which the errors due to the anomaly are corrected This correction is effected, as explained above, according to the angle of deviation using predetermined equations or by a mathematical or graphical comparison more or less complicated with different characteristic lines.

  The correction circuit 12 calculates the anomic components

  UDA and UDA 1 as well as the anomaly signal V The anomaly component is applied in phase to the summing element 11 ao it is combined with the output signal UD and UD 1 of the demodulator At the output of the summing element 11 a ap5 then appears the drift signal corrected for the anomaly,

  lines of sight of the earth-horizon sensor.

  The calculator Ca for the satellite orientation stabilization receives the corrected de-drift signal, the

  anomaly signal and, the clock, the balance and rhythm signal.

  In the embodiment which has just been described, the control of the amplitude and the correction of the anomaly can also be carried out in the computer Ca for the orientation stabilization; the calculator must then

  only select the rhythm of the measurement and the balancing according to the rhythm of the amplitude.

  FIG. 8 shows a terrestrial horizon sensor 1b in which the interpretation of the drift signal corresponding to the chopper amplitude is effected by means of a synchronous modulator 13b. With the aid of a clock 5b for amplitude modulation, the chopper amplitude is modulated with a frequency which is higher than the bandwidth of the drift signal. The output signal of the detector 9b is delivered to a demodulator 10b whose output signal can generally be represented as follows: (14) Am + 1 KXA i + m) -sin (LOA 9 sin m

O _ _ _

U Dm A U Dgri.

  The first term-U Dm A indicates the average value of the component abnormality, while the second term U Dma indicates the portion of the drift signal of the sensor which depends on the deflection angle. Anomaly V can be calculated from the following formulas: U Dm = K 1 + V Am + K- KA '(A Max amine) K (Um x-Umin) (1) biax minmax mun

(A = = A) -0

  At 2 m K (Umax-Umin) V = _ Am.2 r K + (Umax -Uin) In these formulas, Amax and Amin represent the maximum and minimum amplitudes of the chopper, Umax and Umin the drift voltage in case of maximum amplitude and in case of minimum amplitude and K the proportionality factor

  mentioned above of the horizon sensor.

  As can be seen from Equation 14, the signal of the demodulator consists of an anomaly-dependent part and a position-dependent part. In the above embodiment with two fixed amplitudes A and At 1, the error caused by the anomaly was expressed by a fixed absolute value. In the case of an amplitude modulation according to the embodiment of FIG. 8, the component-anomaly is also modulated. This component is demodulated in FIG. the demodulator 13b; the component depending on the anomaly and the anomaly itself are then determined in the correction and linearization circuit 12 b The output signal of the demodulator 10b is applied to the summing member 11b via a low-pass filter 14b The summing element 11b further receives from the correction circuit 12b the average value of the anomaly component The output of the summing member

  11 b is then the corrected drift signal of the sensor.

254791?

  This signal and the anomaly signal V are sent to the computer Cb for satellite orientation stabilization. The curve of the average value of the anomaly component U Dm A for the anomaly values of V = 1/1 and V = 1.5 / 1 is represented in FIG. 6. In addition, the U Dm characteristic lines have been drawn for the average drift voltage for the same

anomaly values 1 and 1.5.

  For the correction of the anomaly, it is also possible to use the signal trains of two sensors of independent terrestrial horizons for a measurement axis. In such a two-channel sensor, the two sensors can oscillate with different amplifiers. one of the sensors with constant amplitude and the second with variable amplitude or the two sensors can modify their amplitude. The interpretation of the drift signals of the two channels can here also be executed either externally in the computer, or also at inside the sensor

  itself or in combination of both possibilities.

  If the two sensors oscillate with different chopper amplitudes, the drift signal and the anomaly signal can be picked up equally quickly and, consequently, the deflection angle can also be corrected without delay. The disadvantage in this case is the drift. different from the two independent channels of the sensor If one of the sensors operates with constant amplitude and the other with variable amplitude, the channel for the second sensor can be adapted

  in its characteristic line to that of the first channel.

  The disadvantage here is as in the first example of

  performing the rhythmic correction of the anomaly.

  Figures 9 and 10 show mountains for the correction of anomalies, which function respectively with two channels, the two terrestrial sensors lc and 1 c 'and ld and 1 of modifying in both cases their amplitude The essential difference lies in the fact that in the example embodiment according to FIG. 9, the interpretation

  outside of the Cc calculator for sta-

  15 in the embodiment of Figure 10 it is carried out inside the sensor. The two sensors 1c and 1c 'of FIG. 9 are of identical construction and, like the sensor of FIG. 7, each comprise an optic 2c, 2c', a chopper disk 3c, 3c ', a control system 4 c, 4 c 'drive for the chopper, a 6 c, 6 c' sensor for the chopper amplitude, and a 8 c, 8 c 'regulator for the chopper drive system. The output signal of the two IR detectors 9c and 9c is demodulated in a demodulator 10c or 10c 'with low-pass filter and amplifier. The drift signals of the two channels are applied to the calculator Cc for stabilization.

  guidance and are interpreted in accordance with the above examples.

  The magnitudes of the choppers in the two channels are varied. The amplitudes can be switched between two values A and A 1, the respective switching times intersecting with the corresponding amplitudes. In this way, the transient oscillating sensor having the largest amplitude-chopper always delivers the drift signal, while the other channel of the sensor having the smallest amplitude produces the signal necessary for the correction of the anomaly. Thanks to this intersection, the corrected drift signal is always available without interruption The output signals of the two channels are compared for the corresponding amplitudes. This comparison is used to balance the characteristic lines.

  that is, to compensate for drift.

  In Figure 10, the interpretation of the signals of

  sensor output is carried out inside the latter.

  For both channels 1d and 1d, there is provided a common clock 5d which controls switching of chopper amplitude and interpretation for both channels. The output signal of demodulator 1 Od in channel 1d is applied. has a

  summing member 11d via a pass filter

  This summing member receives secondly from the second channel 1 an abnormality correction signal determined in the correction element 11 d ', so that at the output of the summing element 11 d, the corrected waveform is obtained At the output of the correction and linearization element 11 the anomaly signal appears. The two signals are applied to the calculator Cd for the satellite orientation stabilization. The correction element 11 d ', as in the embodiment of Figure 7, receives the demodulator 1 Od' with low pass filter and amplifier of the second channel 1 d ', and the demodulator 10 d of the first channel ld the output signal uncorrected Grace

  at this cross connection of two channels, it is possible to continuously measure both the drift signal and the analog signal.

Claims (3)

  1. A method for processing the output signals of an optical sensor of terrestrial horizons of a geostationary satellite, this sensor comprising an input optics, a chopper disc arranged in the focal plane of the input optics, moved in a back-and-forth movement with a determined amplitude (chopper amplitude) and a determined frequency (chopper frequency) and having a diameter corresponding substantially to the image of the earth, a secondary optics and a detector of which the output signal is amplified and demodulated at the chopper frequency into a sensor drift signal which is a measure of the deflection angle of the line of sight of the earth-horizon sensor with respect to the satellite connection line. / center of the earth, characterized by the fact that to reduce the errors that are caused by the differences in heat of the edges of the earth (anomaly) and which are reflected in the drift signals of the sensor, o It is operated at different amplitudes, whereas for these chopper amplitudes the drift signals obtained are compared with the values of the normal characteristic curves for chopper amplitudes without anomaly for a common deflection angle. If the corresponding values are not compliant for a common deflection angle, the measured values are corrected until a concordance of the corresponding values appears and the measured values of these values are reduced. correction (component-anomaly).   2. Method according to claim 1, characterized in that the drift signals (UD, UD 1) are measured for two different chopper amplitudes (A, A 1), and the value (UD) is decreased. , measured for the largest amplitude, of the anomaly component A UDA = (UD-UD 1) A-A 1 254? 917   and that this corrected value is associated with the corresponding deflection angle on the normal characteristic line.   3. Method according to claim 1, characterized in that the chopper amplitude is modulated with a frequency above the bandwidth of the drift signal of the sensor and the synchronization signal is synchronously demodulated. measured drift, comparing the demodulated component to the values of the normal characteristic line   and that the measured drift signal of an error signal corresponding to the comparison -