GB2142498A

GB2142498A - Earth horizon sensor

Info

Publication number: GB2142498A
Application number: GB08412392A
Authority: GB
Inventors: Dieter Sodeikat
Original assignee: Messerschmitt Bolkow Blohm AG
Current assignee: Airbus Defence and Space GmbH
Priority date: 1983-06-24
Filing date: 1984-05-15
Publication date: 1985-01-16
Also published as: IT8421544A0; FR2547917B1; GB2142498B; GB8412392D0; DE3322750A1; FR2547917A1; IT1176294B; DE3322750C2; JPS6015298A; IT8421544A1

Abstract

An earth horizon sensor 1a for a geostationary satellite has a chopper disc 3a (e.g. a disc with the apparent diameter of the earth) in the image plane of optical system 2a and which oscillates at a constant frequency so as to allow light from opposite earth horizons to fall on a single infra-red detector 9a. If the two earth horizons emit different amounts of infra-red radiation, an anomaly constituent is contained in the detector output signal. This can be determined by utilizing signals corresponding to different amplitudes of oscillation of chopper disc 3a. Since the anomaly-dependent component in the detector output signal depends on this chopper amplitude, the correction for the anomaly component can be determined. <IMAGE>

Description

SPECIFICATION A method of processing the output signals of an optical earth horizon sensor The invention relates to a method of processing the output signal of an optical earth horizon sensor of a geostationary satellite.

In order to regulate the position of geostationary satellites regulating signals are required which indicate the displacement of a fixed line of sight of the satellite with respect to a connecting line between the satellite and the central point of the earth. For positioning with respect to two axes optical earth horizon sensors are used. These are one type of zero-seeking sensor. One such earth horizon sensor operates in the infra-red range and is based on the mechanical chopper principle. The infra-red radiation of the earth is collected by an objective iens made of germanium and falls onto a circular chopper disc in the image plane of the lens.

This chopper disc has a diameter which corresponds approximately to the image of the earth and it is moved to and fro with a specific amplitude (the chopper amplitude) at a specific frequency (the chopper frequency).

Light travelling through the collecting optical system is interrupted by the chopper disc. Light from the two opposite earth horizons is alternately allowed through atthe chopper frequency and is conducted onto a detector, for example a pyro-electrical detector, by way of a secondary optical system consisting of a spherical mirror segment and a prism and by way of a spectrum filter for the infra-red range. The output signal of the detector is amplified and subsequently demodulated with the chopper frequency. If the light energy picked up by the detector from both earth horizons is the same, a ZERO signal is supplied through the demodulation.If this is not the case, for example because the line of sight of the earth horizon sensor does not coincide with the connecting line between the satellite and the central point of the earth, the output signal of the detector depends upon the difference between the picked-up amounts of light. This difference is thus a measure of the displacement angle of the line of sight of the earth horizon sensor with respect to the connecting line between the satellite and the central point of the earth.

The displacement angle in one plane can be indicated with such an earth horizon sensor. For three-dimensional stabilisation, two earth horizon sensors are necessary.

The chopper disc is reciprocated with a chopper frequency of about 40 Hz and an amplitude of about 1/17th of the diameter of the disc, i.e. approximately 1 mm.

As compared with other sensor systems, such earth horizon sensors which make use of the chopper principle have several advantages.

On the one hand, the mechanical chopper is a very simple spring/mass system which oscillates with its natural resonance frequency. Therefore, neither a drive motor nor a complex angle read-out are necessary.

On the other hand, the chopped infra-red radiation of each channel is conducted onto only one detector. In this way the problem of balancing or equalising several detectors as well as the ageing problem are abolished.

The measuring range of such an earth horizon sensor is about " 1". The aforementioned zero point error resulting from different radiation from opposite earth horizons is indeed siight, but it can in an extreme case amount to about - 16 percent of the indicated measuring range of one degree. Naturally, one strives to keep this error caused by the radiation anomaly, i.e. the anomaly constituent within the sensor displacement signal, as small as possible. This endeavour is understandable if one considers that geostationary satellites are intended to be used to establish radio and remote-directed television links with the earth.

The object of the present invention is to provide a method of correcting the anomaly constituent within the sensor displacement signals which can be derived in a simple manner from the sensor displacement signals themselves.

Pursuant hereto, the present invention provides a method of processing the output signals of an optical earth horizon sensor of a geostationary satellite, in which respect the earth horizon sensor comprises an input optical system, a chopper disc which is situated in the focal plane of the input optical system, is reciprocated with a specific amplitude and frequency and has a diameter which corresponds approximately to the image of the earth, a secondary optical system and a detector, the output signal of which is amplified and is demodulated with the chopper frequency to form a sensor displacement signal which is a measure of the angle of displacement of the line of sight of the earth horizon sensor with respect to the connecting line between the satellite and the centre point of the earth, characterised in that, to reduce the error (anomaly) of the sensor displacement signals caused by different heat radiation from the earth horizons, the earth horizon sensor is operated with different chopper amplitudes, in that the displacement signals associated with the chopper amplitudes are compared with the values of the standard characteristic curves for chopper amplitudes without anomaly but with a common displacement angle, and, in the event of non-agreement between the corresponding values for a common displacement angle, the measured values are corrected until coincidence of the corresponding values occurs, the measured values being reduced by the correction amount (anomaly constituent) thus determined.

The invention starts from the realisation that the anomaly constituent of the sensor displacement signal is dependent only upon the anomaly V and the chopper amplitude A, so that, upon a change in the chopper amplitude A to a value Al,the anomaly constituent UDA also changes in proportion to the amplitude from UDA to UDA. Thus, upon a change in the chopper amplitude and the change, following therefrom, in the anomaly constituent, the earth anomaly and also the anomaly constituent of the sensor displacement signal can be derived directly.

This will be explained hereinunder: With a uniformly warm earth, both earth horizons at the earth edges radiate the same energy, in other words: EL = ER = E (1) Where E is the average radiation energy, EL is the radiation energy of the left-hand earth edge and ER is the radiation energy of the right-hand earth edge. If the radiation energies from the left-hand and right-hand earth edges are not the same, i.e. EL = ER, there is a so-called "earth anomaly".

For further calculations it is pre-supposed that the average energy is constant, i.e. that EL + ER = 2E = a constant (2) The different radiation energy of each earth edge is represented as EL=E-A ER = E + A (3) In this respect, A is the energy difference between the actual radiation energy and the mean value. What is designated as earth anomaly is the ratio of these two radiation energies: EL E(1 - #) 1 - # 1 - V V = - = --- = -- and # = -- ER E(1 + #) 1 + # 1 + V (4) In practice the anomaly V fluctuates in a range between 1.5/1 to 1/1.5.

If no earth anomaly exists, the characteristic curve of the earth horizon sensor can be represented approximately by the following formula (see Figure 1) 2 KA UDα = -.K.A.sin (α.KA.90) (5) # A The formula is valid forthe range -1 'a.AA A - Where: UD a = the sensor displacement signal which is dependent upon the displacement angle a and is measured as electrical voltage; K = a constant of proportionality which is determined by the sensor geometry; A = chopper amplitude; a = displacement angle of the line of sight of the earth horizon sensor; KA = a constant mechanical transfer factor.

Figure 1 shows the characteristic curve UD a. The entire characteristic curve runs over an angular region of about # 18 , but the measuring range for the sensor displacement angle is only about t 1 , as as indicated. The characteristic curve in this measuring range is shown in Figure 2. It is evident from this figure that for small displacement angles the characteristic curve can be considered linear by applying the formula: UD αx = K . KA . α (6) This 'linear' characteristic curve U0 αx is shown in broken lines in Figure 2.

If the two earth edges radiate different energy, the entire characteristic curve UD is composed of the above undisturbed characteristic curve U0 α together with an anomaly constituent UDA: UD = UDA + UD a (7) The anomaly constituent UDA is then 1-V UDA = K .A . -- (8) 1 + V By reason of the formulae 5,7 and 8 there then emerges for the characteristic curve U0 of the earth horizon sensor 1-V 2 KA UD=K.A.-- + - .K.A.sin (α - . 90) 1 + V # A for the range KA -1#α - . 0 # + 1 (9) A For small values a; 1 - V UD = K.A. -- + K.KA.α (10) 1 + V From the formulae 8 to 10 it follows that, as indicated above, the anomaly constituent UDA of the sensor displacement signal depends only upon the anomaly V and the chopper amplitude A.

If, now, the sensor displacement signal in the case of two different chopper amplitudes A and Al is measured, then the anomaly constituent UDA becomes, at least for small angles, dependent upon the anomaly V. For small angles, the anomaly constituent A AA1 (U0-U01) (11) and for the simplified, corrected dharacteristic curve U0 ax which is valid for small displacement angles A UDαx = UD - -- (UD - UD1) (12) A - A1 For the earth anomaly V the following then applies: (A - A1) . K - (UD - UD1) V = --- (A - A1) . K+(UD - UD1) (13) Thus, the earth anomaly V is dependent only upon the chopper amplitude and the sensor displacements signals in the case of the different chopper amplitudes.For large displacement angles, of course, the complete formula for the characteristic curve must be used.

Accordingly one can, from only the values of the sensor displacement signals for different chopper amplitudes, draw conclusions for a corrected sensor characteristic curve which takes into account the earth anomaly.

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

In Figure 3 two characteristic curves U0 a and UD1 a without anomaly are shown. The characteristic curve U0 a applies for the amplitude A, and the other characteristic curve applies for the amplitude A1. Similarly, two characteristic curves U0 and UD1 are shown, which apply for an anomaly of V = 1.5/1 and are associated with the amplitudes A and A1 respectively.

Firstly the two voltage values U and U1 for -the chopper amplitudes A and A1 respectively are measured. As shown in Figure 4, the displacement angles, indicated by crosses, on the two standard characteristic curves without anomaly would correspond to these voltage values.

Secondly, the voltage values U' or U1' respectively for the amplitudes A and A1 respectively are measured.

These values too are shown as crosses on the two standard characteristic curves UD and UD1 a respectively.

The first values on the standard characteristic curve in accordance with the voltage values U and U1 still lie in the range in which the two standard characteristic curves overlap, whereas the values for U' and U1' no longer lie in the overlapping retion. Since the respective measured values cannot be associated with the same displacement angle on the standard characteristic curves, anomaly exists. In the linear region of the characteristic curve, the anomaly voltage UDA can be calculated directly from equation (11). The anomaly signal in accordance with the earth anomaly V is then calculated from equation (13). For the values U' and U1', however, the complete curve forms of the individual characteristic curves have to be used.In all cases the computing operations can be manipulated so that the corresponding characteristic curves without and with anomaly are all shifted relative to a common zero point and overlap in pairs. From the shifting of the individual characteristic curves the individual anomaly constituents UDA and UDA1 respectively may be calculated. Since the optical system of the earth horizon sensor causes further non-linear distortions of the characteristic curve or of the transfer function of the horizon sensor, the correction for large displacement angles is not derived from the mathematically given characteristic curve, but from the actually measured curve. However, nothing changes about the principle ofthe correction.

Various possibilities are envisaged for accomplishing the above-indicated method of correcting the sensor displacement signals. For example, the chopper disc can be moved intermittently with various amplitudes.

Of course, when the chopper amplitude switches over, the measuring procedure is interrupted so continuous measurement is not possible.

Also, displacement signals of two separate earth horizon sensors can be used for correction of the anomaly. The optical system can then either be doubled or be common to both chopper discs by way of a beam divider.

Amplitude modulation of the chopper amplitude and evaluation of the sensor signal associated with the respective chopper amplitude by means of a synchronous demodulator is particularly advantageous since the anomaly signal and the correction of the angle displacement are then generated directly. If the modulation frequency is made higher than the band width of the sensor displacement signal, the amplitude modulation of the sensor displacement signal can be suppressed. With this arrangement continuous measurement of the signal for the angle displacement and of the anomaly is possible. Also, by using amplitude modulation a completely autonomous sensor-internal correction of the earth horizon sensor is possible.The output signals which the earth horizon sensor emits to the computer for position regulation of the geostationary satellite are accordingly already corrected and no longer need to be corrected in the computer.

The invention will be described further, by way of example, with reference to the accompanying drawings, in which: Figure 1 represents the characteristic curve of an earth horizon sensor without anomaly over the entire detection range; Figure 2 represents the characteristic curve of an earth horizon sensor without anomaly for displacement angle between t 1"; Figure 3 represents two characteristic curves without amonaly for two different chopper amplitudes and the equivalent characteristic curves with anomaly; Figure 4 shows a portion of Figure 3 to an enlarged scale; Figure 5 represents different characteristic curves of the earth horizon sensor in the case of amplitude modulation of the chopper amplitude;; Figure 6 represents the characteristic curves for a mean value of the anomaly constituent of the sensor displacement signal in the case of amplitude modulation of the amplitude of the chopper disc; and Figures 7to 10 are, in each case, block diagrams for different embodiments of earth sensors for carrying out the method of the invention.

In the block diagrams in Figures 7 to 10, for the same or similarly-acting components the same reference numerals are used but the letters a, b, cord respectively are added.

With reference to Figure 7, an optical earth horizon sensor la has an input optical system 2a which is sensitive to the infra-red range. Arranged in the focal plane of the optical system 2a is a chopper disc 3a which is driven by a chopper drive 4a, for example a magnet/spring system oscillating in resonance. The chopper disc 3a is driven in regulated manner with a constant frequency of, for example, 40 Hz with a specific chopper amplitude. The chopper amplitude is switched-over in a specific rhythm between two different amplitudes A and A,. This timed switch-over is controlled by a timing generator 5a, which also controls the timing of the entire evaluation. The chopper amplitude is measured with a sensor 6a which is not described here in detail, and the actual value is passed to a summing location 7a, which receives the desired value at its other input.The resulting displacement signal and a timing signal are fed to a regulator 8a which monitors the chopper drive 4a accordingly. The chopper frequency is fed as synchronisation signal to the timing generator 5a, which also serves to balance the other participating circuit arrangements and in particular maintains a balance with a computer Ca for the position regulation of the satellite.

The infra-red radiation allowed through periodically by the chopper falls onto an infra-red detector 9a which has an amplified output signal. The amplified detector signal is fed in timed manner to a demodulator incorporating a low-pass filter and an amplifier 1 Oa. The output of the demodulator 10a provides a timed series of voltage values U0 and UD1 respectively, which correspond to the sensor displacement signals with the two amplitudes A and A1 respectively. These displacement signals ascertained in sequence are fed on the one hand to a summing member 1 lea and on the other hand to a correction circuit 1 2a in which anomaly errors are corrected.This correction, which depends on the displacement angle, is effected, as explained above, with reference to the given equations or by a more or less complicated mathematical or graphical comparison of individual characteristic curves. In the correction circuit 1 2a, the anomaly constituents UDA and UDA1 as well as the anomaly signal V are calculated. The anomaly constituents are fed in proper phase relation to the summing member 1 lea and combined therewith the output signal UD and U01 respectively of the demodulator. The output of the summing member 11 a is then the anomaly-corrected displacement signal of the lines of sight of the earth horizon sensor.

The corrected displacement signal, the anomaly signal and the balancing and timing signal from the timing generator 5a are fed to the computer Ca for positional regulation of the satellite.

In the exemplified embodiment just described, the amplitude control and the anomaly correction can also be effected in the positional regulation computer Ca. The computer Ca merely has to select from the amplitude timing, the measuring and the balancing timing.

Represented in Figure 8 is an earth horizon sensor 1b in which evaluation of the displacement signal associated with the chopper amplitude is effected by means of a synchronous demodulator 13b. With the aid of a timing generator 5b for amplitude modulation, the chopper amplitude is modulated with a frequency which is greater than the band width of the displacement signal. The output signal of the detector 9b is fed to a demodulator lOb, the output signal of which can be generally represented as

The first term UDmA indicates the mean value of the anomaly constituent, whilst the second term UDm a indicates that part of the sensor displacement signal which is dependent upon the displacement angle. For small displacement angles, the anomaly signal V can be calculated from UDm x = K xK% (Amax - Amin) .K - (Umax-Umin) (15) V = --- (Amax - Amin) . K + (Umax - Umin) Am . 2m . K - (Umax - Umin) V = --- Am . 2m . K + (Umax - Umin) In this respect, Amax and Amin are the maximum and minimum chopper amplitudes respectively, Umax and Umin are the displacement voltages upon maximum amplitude and upon minimum amplitude respectively, and K is the constant of proportionality of the horizon sensor.

As is evident from equation 14, the demodulator signal is composed of an anomaly-dependent part and a position-dependent part. In the case of the above-described embodiment with two fixed amplitudes A and A1 respectively, the error occasioned by the anomaly was expressed by a fixed amount. In the case of amplitude modulation as in the presently described embodiment which is shown in Figure 8, the anomaly constituent is also modulated. This constituent is demodulated in the demodulator 13b. Subsequently, in the correction and linearisation circuit 1 2b the anomaly-dependent constituent and the anomaly itself is determined.The output signal of the demodulator 10b is fed by way of a low-pass filter 14b to the summing member 1 1b. The summing member 11b also receives the mean value ofthe anomaly constituent from the correction circuit 1 2b. The output of the summing member 11 b is then the corrected sensor displacement signal. This signal and the anomaly signal V are fed to the computer Cb for position regulation of the satellite. The graph of the mean value of the anomaly constituent UDmA for anomaly values of V=1/1 and V=1.5/1 is shown in Figure 6.

Also shown are the characteristic curves UDm for the mean displacement voltage for the same anomaly values 1 and 1.5 respectively.

The signals of two independent earth horizon sensors can be used to correct the anomaly for one measuring axis. In the case of such a two-channel sensor, one of the sensors can oscillate on a constant amplitude and the second can oscillate with a varying amplitude or else both sensors can change their amplitude. Evaluation of the displacement signals of the two channels can either be executed externally by the computer or internally in the sensor itself. A combination of both is also possible.

If both sensors oscillate with different chopper amplitudes, the displacement signal and the anomaly signal can be acquired equally rapidly and the displacement angle can be corrected without any delay.

However, the different drift behaviour of the two independent sensor channels may cause problems. If one of the sensors is operated with a constant amplitude and the other is operated with a variable amplitude, the channel for the second sensor can be assimilated to that of the first channel as regards in the characteristic curve. As with the first described embodiment, the timed correction of the anomaly is somewhat disadvantageous.

Figures 9 and 10 are two circuit arrangements for anomaly correction which each work in a two-channel manner. In both cases the two earth sensors 1 C and 1 c' and 1 d and 1 d' respectively change their amplitude.

The essential difference between the two arrangements is that in the embodiment of Figure 9, the evaluation is effected externally in the computer Cue for position regulation, but in the embodiment of Figure 10 it is effected internally of the sensor.

The two channels lc and 1 c' in Figure 9 identically constructed and, like the sensor in Figure 7, each comprises an optical system 2c, 2c', a chopper disc 3c, 3c', a chopper drive 4c, 4c', a chopper amplitude sensor 6c, 6c', and a regulator 8c, 8c' for the chopper drive 4c, 4c'. The output signal of the two IR-detectors 9c and 9c' is demodulated in each case in a demodulator 10c and 10c' respectively with low-pass filter and amplifier. The displacement signals of the two channies are fed to the computer Cc for position regulation and evaluated there in accordance with the above statements.

In both channels the chopper amplitudes are varied. The amplitudes are switched between two amplitude values A and Al,the switching occurring at the same time but in the opposite direction for the respective channels. In this way, the sensor with the greatest chopper amplitude always supplies the displacement signal, whilst the other sensor channel with the smaller amplitude generates the signal which is necessary for correcting the anomaly. The corrected displacement signal is available, without interruption, as a result of this overlapping. The output signals of the two channels are compared for corresponding amplitudes; this comparison is used for balancing the characteristic curves, i.e. for drift compensation.

In Figure 10 the evaluation of the sensor output signals is effected internally of the sensor. For both channels 1 d and 1 d' a common timing generator 5d is provided, which controls the switching of the chopper amplitudes and the evaluation for both channels. The output signal of the demodulator 10din the channel 1d is fed by way of a low-pass filter 14dto a summing member 1 lid. This summing member 1 also receives the correction signal for the anomaly from the second channel 1 d', which signal has been ascertained in the correction member 11 d', so that the corrected displacement signal is present at the output of the summing member 11 d. The anomaly signal is, of course, present at the output of the correction and linearisation member 1 lid'. Both signals are fed to the computer Cdfor position regulation of the satellite. As in the embodiment shown in Figure 7, the correction circuit 11 d' receives the uncorrected displacement signal from the demodulator lOd' with low-pass filter and amplifier of the second channel 1 d' as well as from the demodulator 1 Od of the first channel 1 d. Due to this cross-connection of the channels, the displacement signal and the anomaly signal can be measured continuously.

Claims

1. A method of processing the output signals of an optical earth horizon sensor of a geostationary satellite, in which respect the earth horizon sensor comprises an input optical system, a chopper disc which is situated in the focal plane of the input optical system, is reciprocated with a specific amplitude and frequency and has a diameter which corresponds approximately to the image of the earth, a secondary optical system and a detector, the output signal of which is amplified and is demodulated with the chopper frequency to form a sensor displacement signal which is a measure of the angle of displacement of the line of sight of the earth horizon sensor with respect to the connecting line between the satellite and the centre point of the earth, characterised in that, to reduce the error (anomaly) of the sensor displacement signals caused by different heat radiation from the earth horizons, the earth horizon sensor is operated with different chopper amplitudes, in that the displacement signals associated with the chopper amplitudes are compared with the values of the standard characteristics curves for chopper amplitudes without anomaly but with a common displacement angle, and, in the event of non-agreement between the corresponding values for a common displacement angle, the measured values are corrected until coincidence of the corresponding values occurs, the measured values being reduced by the correction amount (anomaly constituent) thus determined.

2. A method as claimed in claim 1, characterised in that the displacement signals (U0, UD1) for two different chopper amplitudes (A, Al) are measured, and in that the value (UD) measured for the greater amplitude is reduced by the anomaly constituent A Urn = As (UD UD1), and in that the corresponding displacement angle on the standard characteristic curve is correlated with this corrected value.

3. A method as claimed in claim 1, characterised in that the chopper amplitude is modulated with a frequency above the band width of the sensor displacement signal and the measured displacement signal is demodulated synchronously, in that the demodulated constituent is compared with the values of the standard characteristic curve, and in that the measured displacement signal is reduced by an error signal corresponding to the comparison.

4. A method of processing the output signals of an optical earth horizon sensor of a geostationary satellite substantially as hereinbefore described with reference to the accompanying drawings.