 |
Description  |
|
|
BACKGROUND OF THE INVENTION
The present invention relates to fiber-optic sensors and particularly to a
serial interferometric fiber-optic sensor array for sensing changes in
environmental conditions or physical phenomena.
Fiber-optic sensors have been developed for use in many sensing
applications in a wide range of fields. This is due to the high
sensitivity of optical fibers to various environmental conditions or
physical phenomena which affect such optical fibers. For example, factors
such as temperature, pressure, electromagnetic fields, and acoustical
waves directly affect the light transmitting characteristics of optical
fiber. These changes in the optical fiber produce a change in the phase of
light signals traveling through that optical fiber. As a result, a
measurement of the change in phase of light signals propagated through the
optical fiber is representative of changes in those physical phenomena
which have affected that fiber.
In recent developments, fiber-optic sensors have been implemented into sets
or arrays so that a number of sensors can utilize light from a single
source and provide environmental information on physical phenomena from
various locations to a common processing location for subsequent
processing. Such arrays of sensors can be used, for example, in various
geophysical exploration and antisubmarine warfare applications.
A fiber-optic sensor array can be implemented in a variety of different
configurations, some of which being very complex. Typically, a fiber-optic
sensor can include a fiber input bus for carrying light to an array of
sensors, with each sensor imparting information about the local
environment to this light. This information can be collected by an output
fiber bus and propagated to the common processing location, where
information obtained from any selected one of the sensors can be readily
identified and analyzed.
In the development of these fiber-optic sensor arrays a number of different
approaches have been proposed by which information from each sensor in an
array is demultiplexed, or separated, for individual identification from
among all of the information arriving at the common processing location
from the output fiber bus. Some of these approaches are based on
conventional formats of frequency and time division multiplexing, while
other approaches involve more optically complex and specialized schemes
such as coherence multiplexing. As a result, present fiber-optic sensor
arrays essentially include many optical fibers and optical components,
which make them expensive and difficult to operate.
OBJECTS OF THE INVENTION
One object of the invention is to provide an improved fiber-optic sensor
array.
Another object of the invention is to provide a novel serial
interferometric fiber-optic sensor array.
Another object of the invention is to provide a simple, inexpensive,
compact array of interferometric fiber-optic sensors to be efficiently
addressed and demodulated using a minimal number of optical components and
connecting fibers.
A further object of the invention is to provide a serial array of
interferometric fiber-optic sensor elements which are tapped before the
first sensor element, between adjacent sensor elements, and after the last
sensor element by a single output fiber bus.
SUMMARY OF THE INVENTION
These and other objects of the invention are achieved by providing a serial
fiber-optic sensor array which comprises: an input fiber formed into a
series of N sensor elements at separated locations along the input fiber,
each of the sensor elements being of optical path length L and being
responsive to any change in an associated predetermined physical parameter
for changing its optical path length; a light source for selectively
transmitting a light pulse into the input fiber; an output fiber coupled
to the series of N sensor elements for coupling a predetermined portion of
the light pulse at each location before the first sensor element, between
adjacent ones of the sensor elements and after the Nth sensor element in
order to produce at the output of the Nth sensor element a series of N+1
pulses separated in the time domain; and output means of optical path
length L being responsive to the N+1 pulses for coherently mixing pulses
obtained from each pair of consecutive locations to obtain a series of N
interferometric signals respectively indicative of any changes in the
physical parameters to which the sensor elements are respectively
responsive.
BRIEF DESCRIPTION OF THE DRAWINGS
These and other objects, features and advantages of the invention, as well
as the invention itself, will become better understood by reference to the
following detailed description when considered in connection with the
accompanying drawings wherein like reference numerals designate identical
or corresponding parts throughout the several views, and wherein:
FIG. 1 is a schematic block diagram of a preferred embodiment of the serial
array configuration of the invention;
FIG. 2 illustrates waveforms useful in explaining the operation of the
invention; and
FIG. 3 illustrates a system in which the preferred embodiment of FIG. 1 can
be utilized.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT
Referring now to FIG. 1 a schematic block diagram of a preferred embodiment
of the invention is shown. An input pulse 11 of light from a laser source
(FIG. 3) is launched into an input optical fiber 13 formed into an
exemplary serial array 14 of N sensors or sensor elements at separated
locations along the input optical fiber 13. For purposes of this
description, let N=4. Thus, as shown in FIG. 1, the array 14 is comprised
of sensor elements 15-18. Each of the sensor elements 15-18 has a
preselected optical path length L. These sensors 15-18 monitor
environmental conditions or physical phenomena (such as changes in
temperature, pressure, electromagnetic fields and acoustical waves) which
selectively produce changes in the phase of the input pulse 11 traveling
through the sensors 15-18.
Fiber directional couplers 21-25 are utilized to couple the input optical
fiber 13 to an output optical fiber bus 27 at preselected tap points or
locations 31-35 along the input optical fiber 13, where the input and
output optical fibers 13 and 27 abut each other within the respective
couplers 21-25. These locations 31-35 are respectively selected to be
before the sensor element 15, between each of sensor element pairs 15 and
16, 16 and 17 and 17 and 18, and after sensor element 18.
Each of the directional couplers 21-25 is constructed, by conventional
means well known in the art, to have a power splitting ratio (or tap
ratio) of, for example, 1%. Furthermore, each of the couplers 21-25 has
input ports 1 and 2 and output ports 3 and 4, with the input optical fiber
13 passing between ports 1 and 3 and the output optical fiber 27 passing
between ports 2 and 4 of each of the couplers 21-25.
Because of the power splitting ratio of 1%, 99% of the optical power that
enters input port 1 of each of the couplers 21-25 propagates to the output
port 3 of each coupler and remains in the input optical fiber 13. On the
other hand, only 1% of the optical power that enters input port 1 of each
of the couplers 21-25 is cross-coupled to the output port 4 of each of the
couplers 21-25 and into the output optical fiber 27. In a similar manner,
99% of the optical power that enters input port 2 of one of the couplers
22-25 is passed to output port 4 of that coupler and remains in the output
optical fiber 27. No light is injected into or cross-coupled into input
port 2 of coupler 21.
Thus, only a small fraction K.sub.j (1% in this description) of the optical
power in fiber 13 is tapped-off to the output fiber bus 27 at each of the
tap points or locations 31-35. The output of the sensor array 14 is
developed at output port 4 of the coupler 25. If the optical propagation
delay in each of the sensor elements 15-18 is greater than the width of
the input pulse 11, the output of the array 14 consists of a pulse train
or series of N+1 pulses 37 which are separated in the time domain. Since
N+1=5 in this description, there are five pulses in the pulse train 37 at
the output port 4 of coupler 25. These five pulses in the pulse train 37
were respectively obtained from the consecutive tap points 21-25 in the
array 14. Apart from crosstalk effects, which will be discussed later,
these pulses in the pulse train 37 carry no direct interferometric
information.
The pulse train 37 is applied to a compensating interferometer 39, which is
environmentally isolated. The compensating interferometer 39 includes a
piezoelectric transducer 41, an optical fiber segment 43 formed into a
delay coil 45 which is wound around the transducer 41, and two 3dB fiber
directional couplers 47 and 49. Couplers 47 and 49 couple the fiber
segment 43 to the output optical fiber 27 at locations 51 and 53 along the
fiber 27, where the fiber segment 43 abuts against the fiber 27 within the
respective couplers 47 and 49. Each of the couplers 47 and 49 has input
ports 1 and 2 and output ports 3 and 4, with opposite ends of the fiber
segment 43 passing between ports 1 and 3 of the respective couplers 47 and
49, and the output optical fiber 27 passing between ports 2 and 4 of each
of the couplers 47 and 49. The compensating interferometer 39 has an
optical path length inbalance of L between the path through the fiber 27
and the path through the delay coil 45. This imbalance is due to the fact
that the delay coil 45, like each of the sensor elements 15-18, has an
optical path length L.
The waveforms of FIG. 2 will also be referred to at this time to better
explain the operation of the compensating interferometer 39.
The input pulse 11, as discussed above, causes the optical pulse train 37
to be developed at the output port 4 of coupler 25. When the optical pulse
train 37 is applied to the compensating interferometer 39, one-half of the
optical power in the pulse train 37 goes through the output optical fiber
27 to input port 2 of coupler 49 as a non-delayed pulse train 55. The
remaining one-half of the optical power in the pulse train 37 is delayed
one optical path length L by the delay coil 45 before arriving at input
port 1 of coupler 49 as delayed pulse train 57.
Time-coincident pulses in the non-delayed pulse train 55 and in the delayed
pulse train 57 are coherently mixed in the coupler 49 of the compensating
interferometer 39 to develop at the output port 4 of coupler 49 an output
pulse train 59 consisting of a series of N+2 pulses, which in this
description is a series of six pulses. Note that the first and last pulses
in the pulse train 59 contain no interferometric signals, while the
central four pulses in the pulse train 59 carry interferometric
information generated by the sensor elements 15-18. The interferometric
information in the central four pulses of the pulse train 59 indicates the
respective amounts of stress, and hence phase shifts, experienced by the
sensor elements 15-18 in the array 14.
A sinusoidal oscillator signal at a preselected high frequency, such as 20
KHz, is applied to the piezoelectric transducer 41 to produce in the delay
coil 45 a phase carrier signal of 2.pi. radians peak-to-peak at the
exemplary 20 KHz modulates the `interferometric signal` carried by each of
the central four pulses in the output pulse train 59. This phase carrier
signal is subsequently used to demodulate the four interferometric signals
produced by the compensating interferometer 39. In an alternative
embodiment of FIG. 1, the piezoelectric transducer 41 and its associated
oscillator signal may be eliminated.
In general, the coupling ratios of the directional couplers required to
equalize the returned power from each of the tap points in the array 14
can be derived by simply equating the power from the first tap in the
array 14 to that from the n.sup.th tap in the array 14, i.e., (1 n N),
where N=the number of sensor elements in the array 14 and n=an integer
between 1 and N. This produces the relationship:
K.sub.1.(1-K.sub.2).(1-K.sub.3). . . .(1-K.sub.N+1) =(1-K.sub.1). . .
.(1-K.sub.n-1).K.sub.n.(1-K.sub.n+1). . . .(1-K.sub.N+1). (1)
where K is a power tapping ratio. This sets a requirement of:
K.sub.1 /(1-K.sub.1)=K.sub.n /(1-K.sub.n), (2)
i.e., equal power splitting (K) at each tap point. The power in each pulse
at the output of the array 14 (at port 4 of coupler 25) is thus:
P.sub.n =n.P.sub.o =K.(1-K).sup.N.P.sub.o, (3)
where P.sub.o is the peak power in the input pulse 11. This neglects
multiple cross-coupling of the pulses in the array 14, and excess loss in
the system, which would be expected to modify this result only slightly
for losses <0.2 dB/coupler. The intrinsic crosstalk between sensors can be
shown to be directly related to the power tapping ratio K, and can be
assessed by considering the number of interfering pulses generated in the
output of the array 14. Taking into account first order crosstalk effects
only (i.e. pulses which cross-couple back from the output fiber 27 to the
sensor array 14 and back again), it can be shown that the number of
crosstalk pulses received in the n.sup.th time slot (i.e. t=nT, where T is
the optical delay through each sensor of length L) at the array output is
given by
M=Nn-n.sup.2 (4)
Each of these pulses is a factor K.sup.2 weaker than the primary tapped
pulses, but mix interferometrically at the output of the compensating
interferometer 39 with primary pulses derived from adjacent time slots
(i.e. (n-1)T and (n+1)T) to produce crosstalk. This leads to a worse case
time averaged crosstalk (sensor to sensor) for a centrally located sensor
of
##EQU1##
This is the result which would be expected intuitively, and again neglects
excess losses and polarization effects.
It should be noted at this time that in FIG. 1, each pair of adjacent
couplers (e.g. 21 and 22) and the intervening portions of the output fiber
27 and the input fiber 13, including the associated one of the sensor
coils 15-18 (e.g. 15), between that pair of adjacent couplers form a Mach
Zehnder interferometer. Therefore, in the embodiment shown in FIG. 1,
there are four Mach Zehnder interferometers serially disposed between the
couplers 21 and 25.
Referring now to FIG. 3, a system is shown for utilizing the embodiment of
FIG. 1. Light from a suitable laser 61 is passed through an isolator (not
shown) and into an optical gate or acousto-optic modulator (AOM) 63. The
AOM 63 is typically a Bragg cell.
The light from laser 61 normally passes through the AOM 63 without being
transmitted into the input optical fiber 13 (FIG. 1). However, each time
that a short RF pulse is applied from a pulse generator 65 to the AOM 63,
the laser light is deflected through the AOM 63 and is launched as input
pulse 11 into the input fiber 13 and serial array 14 of FIG. 1, as
discussed before. Typically, 70-80% of the light from the laser 61 is
deflected into the optical fiber 13 of FIG. 1 during the RF pulse.
The oscillator signal that is applied to the piezoelectric transducer 41 in
FIG. 1 is generated by a conventional demodulation electronics circuit 67.
It will be recalled that this oscillator signal to the transducer 41
causes a non-reciprocal phase carrier signal of 2.pi. radians peak-to-peak
at 20 KHz to be produced in the sensor coil 45 to modulate the
`interferometric signal` carried by each of the central four pulses in the
output pulse train 59.
The output pulse train 59 that is produced by the compensating
interferometer 39 of FIG. 1 is applied to a conventional time-division
demultiplexer 69. For proper timing, the same RF pulse that is applied to
the AOM 63 to initiate the operation to develop the output pulse train 59
is also applied as a synchronizing signal to the demultiplexer 69.
The time division demultiplexer 69 comprises a number of gates, switches
and channels (not shown). Basically, the demultiplexer 69 performs a
sample and hold operation on each of the interferometric signals contained
in the output pulse train 59 and separates or demultiplexes them into
respective output channels. Since only four exemplary sensing elements
15-18 are used in the embodiment of FIG. 1, only four channels are needed
at the output of the demultiplexer 69. Thus, in response to the
synchronizing RF pulse from pulse generator 65, the time-division
demultiplexer 69 separates or demultiplexes the central four pulses in the
output pulse train 59 into four respective output channels.
The four time-division demultiplied outputs from demultiplexer 69 now have
to be demodulated by the demodulation electronics circuit 67. The
demodulation process performed by the circuit 67 linearizes each
interferometric signal in the four output channels of demultiplexer 69.
The interferometric signal generated by each overlapping pair of pulses,
or time-coincident pulses, in the non-delayed pulse train 55 and in the
delayed pulse train 57 is a (1+cos .phi.) function of the phase difference
.phi. between those pulses. This is not a linear function.
Demodulation of the time-division demultiplexed outputs of demultiplexer 69
can be achieved using either `phase generated carrier` homodyne or
synthetic-heterodyne techniques applied to the compensating interferometer
39 of FIG. 1. In this description the outputs of demultiplexer 69 are
demodulated by using the synthetic-heterodyne technique. This technique,
as described before, involves the application of the oscillator signal to
the piezoelectric transducer 41 to stretch the fiber in the sensor coil 45
and thereby produce a (false) phase carrier signal for interferometric
signals developed by the compensating interferometer 39 (FIG. 1). Such a
phase carrier signal is phase modulated by the interference signals in the
output pulse train 59. This resultant phase modulated carrier on each of
the four time-demultiplexed outputs of demultiplexer 69 can be demodulated
by the demodulation electronics circuit 67 by using standard electronic
circuitry, such as phase-locked loops and FM discriminators. The
demodulated outputs of the circuit 67 are four linearized sensor outputs.
Therefore, what has been described is a new serial interferometric
fiber-optic sensor array configuration which can be multiplexed using
time-division addressing. The configuration is based on a serial network
or array of sensor elements which is tapped between adjacent elements and
before the first element and after the last element by a single output
fiber bus.
It should therefore readily be understood that many modifications and
variations of the present invention are possible within the purview of the
claimed invention. It is therefore to be understood that within the scope
of the appended claims, the invention may be practiced otherwise than as
specifically described.
* * * * *
|
|
 |
|
 |
Description |
|
