搜尋 圖片 地圖 Play YouTube 新聞 Gmail 雲端硬碟 更多 »
進階專利搜尋 | 網頁圖片 | 網頁紀錄 | 登入

專利

  
[merged small][merged small][graphic]
[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][table][merged small][merged small][merged small][merged small][merged small][merged small][table][merged small][merged small][merged small][merged small][graphic][merged small]

1

HYBRID NAVIGATION SYSTEM
OTHER APPLICATIONS

The present application is a continuation-in-part of our copending application, Ser. No. 402,966, filed on Oct. 9, 1964, now abandoned.

The present invention relates to hybrid navigation systems and, more particularly, to a class of such systems in which a means for determining the position of a vehicle is combined with an inertial navigation system in order to compensate errors accumulated by the latter system.

In known hybrid navigation systems, attempts are made to derive advantages from the measurements of a position by non-inertial means for calibrating the inertial navigation system and for reducing its systematic errors. In these prior systems, there is only a partial combination of the two kinds of systems since the determination of the position by non-inertial means does not derive all the available advantages of the presence of an inertial navigation system.

According to one distinguishing feature of the present invention, means aire provided not only for enabling the measurement of position errors of the inertial navigation system and to correct the same, but also and mainly to make use of the knowledge of the movement of the vehicle to be located as supplied by the inertial navigation system to improve considerably the sensitivity of the measurement of the position without sacrificing its precision.

The proposed system must, on the other hand, be able to withstand particularly strong jamming without introducing errors into the indications of the inertial navigation system even when the radio measurement is effected. For this purpose, use is made on the one hand of a signal extractor having a very narrow final pass band and a negligible delay which is considerably below the period of time during which the signal is "integrated" and, on the other hand, no direct action is effected upon the inertial navigation system but a computation of its systematic error is made and account is taken of said error in a suitable way.

In a system according to features of the present invention, the signal extractor is rendered insensitive to the displacements of the vehicle to be located. The information about its position given by the inertial navigation system duly corrected enables this result.

According to another feature of the present invention it is proposed to realize a precise locating measurement even when the ratio of the power of the signals to the power of existing noise is very much smaller than one. This is possible, according to known general theory, only if the energy of the received signals during all the duration of the measurements is of the order of the spectral density of noise (said noise being assumed to be "white noise"). In order that this "energy ratio" be sufficiently large, while permitting the "power ratio" to be small, it is necessary and it is sufficient that "the duration of the measurement", which is also called "duration of integration", be long.

One of the aspects of the present invention consists of integrating the position measurement during a period of time sufficiently long with respect to the period of time during which the vehicle has moved by an amount corresponding to the tolerated location error.

2

Such a long duration of integration is normally considered as prohibitive, but the present hybrid navigation method renders it practicable. If the duration of integration is long, there result two 5 consequences. First, the vehicle will be subjected to a substantial displacement during the time of integration. It is therefore necessary to use a method of signal extraction which must be insensitive to displacements of the vehicle. Further, the recursion frequency of the in10 formation of radio position becomes small. There is not available, therefore, at every instant of time, an information of radio position which could eventually be used for elaborating an error signal in the servo-loops of the inertial navigation system. 15 According to a further feature of the present invention, there is provided a means for representing the systematic errors of the inertial navigation system. This representation is in effect a mathematical formula giving the systematic errors, or drift of the inertial naviga20 tion system, as a function of time depending upon a number of parameters. The values of these parameters are obtained during the movement of the vehicle from the deviations between successive radio positions measured and the positions given at the same instants of 25 time by the inertial navigation system.

Any accurate localization system can, in principle, be combined with an inertial system according to features of the present invention. However, in effect, when one is interested in long range position systems and in very 30 fast aircraft, it is necessary to use long waves or waves utilizing the reflection from an artificial stable ionosphere or position methods by means of satellites or by radio or optical astronomy. There will be described hereinafter one embodiment 35 of the present invention, based on the use of long waves and intended for aircraft such as airplanes.

The proposed localization system utilizes the measurement of the distances of the vehicle to three guiding radio transmitters for determining the location of 40 the aircraft. The signals used are analogous to those employed in the system well known under the name of LOR AN C. and consist of radio frequency pulses with steep leading edges (for example 30 for the rise time of the signals) arranged to eliminate at reception the 45 effect of the natural ionosphere.

The determination of the distance from the vehicle or craft to be located relative to a radio transmitter is not made, as in the LOR AN C. or in other known hyperbolic systems, by comparing distances between each other, but by comparing the phase of the received signal to that of a high stability oscillator carried on the craft. This measurement is made in practice by determining the difference, or deviation, existing between 55 the phase of the signal effectively received and the phase of the signal which would have been received if the position of the vehicle had been that indicated by the inertial navigation system duly corrected. This process has the result of rendering the signal extractor in60 sensitive to the displacements of the vehicle during the duration of the integration. Also, the proposed process permits avoiding automatically the ambiguity due to the exclusive use of the phase of the carrier of the transmitted signals. 65 The distances to the guiding radio-transmitters are obtained afterwards by correcting, by the amount of the measured differences, the distances given by the compensated inertial navigation system. These dis

3 4

tances being taken or measured at the moment at ematical-model used in program 11'were exact); said

which the integration ends, the time assigned to each computed propagation delays are obtained through an

radio measurement is that of the end of the integration, operation 13 of computation from the corrected posi

whatever the duration of this integration may have tion values 12.

been, provided however that the mathematical repre- 5 The signal extraction operation 1 consists of estab

sentation used for the correction of the errors of the in- lishing, after a measurement cycle, the differences of

ertial navigation system can describe same in a suffi- the true propagation delays (corresponding to radio

ciently precise manner during this interval of time. signal 2) with respect to the assumed delays (computed

Therefore, the integration does not bring any delay into propagation delays 6).

the measurement. 10 From these differences and from non-corrected posi

The position of the vehicle is deduced in a known tion values 9 and from assumed delays 6, a computa

manner from the three distances thus measured at the tion according to a program of computation 7 is made

guiding radio-transmitters. A comparison is afterwards of new or "updated" values of the parameters 8 which

made with the position given by the inertial navigation are comprised in the aforesaid mathematical model,

system (without error compensation) for computing 15 which mathematical model is computed according to

for the next period (by means of an electronic com- program 11' as stated above. The next cycle of mea

puter) a new set of parameters for the mathematical surements then starts.

model, or representation, which model represents the Referring now to FIG. 2, it may be seen that the radio

systematic errors of said inertial navigation system. signal 2, from a guiding transmitter (not shown) is re

A mixed or hybrid navigation system incorporating 20 ceived and amplified in receiver-amplifier 14, which

features of the present invention will next be described receiver is brought into operation by an unblocking or

in detail with reference to the accompanying drawings. control signal (C4 in FIG. 3) received via line 23 from

DRAWING a timC baSC 8enerator 18'

The amplified signal is then limited in amplitude in

FIG. 1 is an operation diagram illustrating various 25 Hmiter 15 and sampled by sampler 16 controlled by sig

features of the invention; nals CS received via line 17 from said time base genera

FIG. 2 is a block diagram of receiving and sampling tor 18. On the other hand, from the computed propaga

means coupled with extractor and time base means em- tion delays 6 as well as from the indications of an

ployed in accordance with the invention; atomic clock, 31, there is produced in time base gener

FIGS. 3 and 3(a) are charts of reference signal and 30 ator 18 reference signals such as C7 and C8 (FIG. 3)

blocking signals wave forms; which are transmitted on lines such as 19. From the

FIGS. 4 and 4(a) are block diagrams of a data han- limited-in-amplitude and sampled signal received via

dling system used in the practice of the invention; and line 20 and from the reference signals on lines 19, there

FIG. 5 is a block diagram showing the interconnec- are computed in a circuit 21, which will be discussed

tion of the devices of FIGS. 2 and 4. ^5 hereafter with reference to FIG. 4, numbers representing the deviations between the computed propagation delays 6 and the true transmission delays. Signals repre

FIG. 1 is a block diagram illustrating schematically a senting these numbers are presented at terminal T 51.

flow of operations incorporating features of the present These delays are given in the form of phases and their

invention. The block 10 represents an operation of in- 40 differences in the form of phase differences, the com

ertially generating position values 9, this operation putation of which is started at the end of each period

being carried out by a conventional inertial navigation of measurement, by synchronization signals on line 22

system. given by the time generator 18.

The position values 9 are then corrected according to During the intervals separating the reception of two a program 11 of correction of these position values fol- 45 consecutive pulses of radio waves from guiding translowing a computation 11' of this correction, according mitters, there is applied via lines 23 and 17 to the reto a mathematical model or representation of the sys- ceiver 14 and to the sampler 16 blocking signals genertematic error of the inertial navigation system, i.e., of ated by the time base generator 18. These signals are the deviation between the position values 9 and the ^ intended to render the system insensitive both to ionotrue position values. spheric echoes and to jamming between two consecu

The corrected position values 12 thus obtained tive pulses,

through execution of program 11 are available for use In a system embodying features of the present inven

in controlling the governing of the vehicle, and for a se- tion, the reference signals used for determining the

quence of operations which aims to update the parame- J5 phase of the received signals are constituted by square

ters of the mathematical model and consists of comput- waves C7, C8 as shown in FIG. 3. It is possible to visual

ing, according to a program 13, propagation delays 6 ize said signal wave shapes as a sine and a cosine wave

from values 12 and of establishing by a signal extrac- limited in amplitude, the phase of which is that com

tion-operation 1, at each end of periodic radio- puted from the indications of the inertial navigation

measurements, the differences between computed ^ system corrected of the deviation. Due to this correc

propagation delays and true propagation delays and tion, one knows the phase of the carrier wave within an

then of computing, according to a program 7, new or approximation of ± dO, d"0 having a maximum equal to

updated values of the parameters 8 of the mathematical w/4 (this corresponds to knowing the position with an

model. approximation of 300 meters for a wave length of 3,000

The above mentioned "computed propagation de- 6J meters),

lays", are those which should be obtained if the cor- The reference signals are produced by flip-flop cir

rected position values 12 were true (obviously : if the cuits, the instants of flip-flopping of which are obtained

correction performed according the deviation-math- by counting, with a frequency sufficiently high such as,

DETAILED DESCRIPTION

5 6

for example, 10 megacycles per second, the pulses where it must be noted that all connecting lines are

which are provided by the atomic clock. This fre- represented schematically according to a "one line rep

quency (of, for instance, 10 magahertz) will, from now resentation" in which there is represented one line for

on, be referred to as the "high frequency". It must be each category of signal. While, in FIG. 4, the plurality

understood that, since the reference signals have a 5 T 19 of output terminals of computer G 13 is generally

much lower frequency, they are formed of a large hum- represented by one terminal, it is resolved, in FIG. 5,

ber of H.F. pulses (equal to the ratio of respective fre- into two lines, both belonging to the plurality T 19 of

quencies). The reference signals might be, for illustra- outputs of computer G13 and each concerning a differ

tive purposes, considered as resulting from a "square ent use.

modulation" of said H.F. pulses. 10 The pulses of the sampled signals are applied to ter

FIG. 3 shows the shape of the reference signals C7 minals Tl or T2 according to their polarity (positive at

and C8 and the shape of the blocking signals during the Tl, negative at T2). The reference signal 90° out of

reception of one pulse with one of the two possible op- phase is applied at terminal T3 or T4, according to its

posite phases for the carrier wave. Curve CI shows the polarity. Likewise, the reference signal in phase is ap

phase of the carrier such as computed from the cor- 15 plied at terminals T5 and T6.

rected indications of the inertial navigation system. The During the reception of one radio-frequency pulse, curves C2 (in dotted lines) and C3 (in dash and dot or during one radio-frequency period in the case of the lines) correspond to the maximum phase differences to above described alternative, the gates G7 and G8 conbe considered. Curve C4 is the blocking signal applied trolled by pulses from terminals T3 and T4 and the on line 23 to the receiver and curve C5 the blocking 20 "OR" circuit G9 transmit the impulses from terminals signal controlling the output of the sampler. Tl and T2 to the counting device G 10. At the end of

This last signal stops one-eighth of a period after the the reception, the figure recorded on this counting deinstant of time computed from the beginning of the sig- vice is equal, up to a multiplicative constant, to the nal and starts again one-eighth of a period before the value of the product of the sampled signals by the signal instant of time computed at which arrives the first iono- 25 900 out of phase. A first synchronization signal applied spheric echo. Thus, there are obtained two guarding to terminal Til empties the counting device and transintervals C6. The curves C7 and C8 represent refer- mits this figure to the memory G12 of the computer ence signals in phase and 90° out of phase respectively. G13. During that period of time, the gates G14 and

As an alternative, in some cases it is more advanta- G15, controlled by pulses from terminals T5 and T6, geous, particularly with respect to the calibration, to 30 and the "OR" circuit G16, transmit in their turn the reduce the duration of reception of one pulse to a sin- pulses from terminals Tl and T2 to the counting device gle period of the radio frequency wave. FIG. 3(a) indi- G10 where they arrive through the "OR" circuit G9 cates the shapes that are assumed in this case by the after having been delayed in G17 by a period of time reference signals in phase, curve C7', and signals 90° sufficient for enabling the counting device G10 to reout of phase, curve C8', as well as the blocking signals, 35 turn to zero (about 200 microseconds). The new figure for the receiver, curve C4', and for the sampler, curve recorded gives, with the same constant of proportionalC5'. ity, the product of the sampled signal by the reference

The carrier wave with a phase such as computed signal in phase. A second synchronization signal trans

from the corrected indications of the inertial navigation mitted to terminal Til empties this second figure into

system is represented by the sinusoid wave CI' (solid 4^ the memory G12. Other synchronization signals from

line). The curves C2' (dotted line) and C3' (dash and terminal 11 which are applied to the computer, intro

dot line) correspond to waves having the maximum duce the addresses of all the figures which are given out

phase differences to be considered. by the counting device G10. To these figures are allo

The interval of time during which the receiver is un- cated six sections of the memory G12, two sections

blocked and during which the reference signals are dif- 45 being allocated to each radio guiding transmitter, one

ferent from zero is chosen in order to extend from of these sections corresponding to the product of the

three-eighths of the value of a period before the com- sampled signal by the signal 90° out of phase and the

puted starting time of the signal, through nine-eighth of other to the product by the signal in phase. In each

a period after this same instant of time. The blocking case, there is accumulated the sums of the figures given

signal C5' of the sampler is chosen in order to stop out by the counting device G10.

three-eighth of a cycle after the unblocking instant of Since the reference signals contain a definite number

time of the receiver and to start again eleven-eighths of of high frequency pulses it is obvious that if they are

a cycle after the same. These values of three-eighth and presented, together with the sampled signals, to the

eleven-eighth of a cycle correspond to an integral num- 5g gates G15, G14, G8, G7, this will cause each one of

ber of high frequency cycles. said gates to present at its output, a number of pulses

There is thus provided an interval C6' of three- which depends upon the percentage of time during

eighths of a cycle before unblocking the sampler which which the waves of the reference signal and of the sam

permits the establishment of the permanent values of pled radio signal (applied to its inputs) are simultathe signal. Likewise, there is provided an interval C6' 6Q neously of like sign (plus on gates 7 and 14, minus on

of one-eighth of a cycle between the blocking of the gates 8 and 15).

sampler and that of the receiver. For instance, the presentation of the reference signal In FIG. 4, there is shown a data handling system in- C7 (in FIG. 3), in phase with the "computed phase", corporating features of the present invention. This sys- to the inputs T5 and T6 of gates T14 and T15 respectem is adapted to extract the phase deviation of the car- 65 tively and the presentation of the sampled signal to terrier wave with respect to the reference signals. minal T2 and Tl, inputs of said gates will produce, on The interconnection of this system with the blocks of their respective outputs, a number of pulses which is FIG. 2 may be seen on the block diagram of FIG. 5 equal to the maximum number of pulses which can be

« 上一頁繼續 »