Title: Electromagnetic induction detection system
Abstract: A airborne electromagnetic induction (EMI) detection apparatus and system. In accordance with one embodiment, the EMI detection apparatus includes a transmitter element in the form of a transmitter coil that emits a primary, multi-frequency component magnetic field which induces a secondary magnetic field in an external body. A receiver element in the form of a receiver coil is mounted in a horizontal loop-loop orientation with respect to the transmitter coil and receives the secondary magnetic field. The detection apparatus further includes a magnetic shield disposed around the receiver coil to limit the lateral footprint diameter observed by the receiving element and to shield the receiver coil from the primary magnetic field and other external electromagnetic radiation to improve the gain and resolution of the detection apparatus.
Patent Number: 6,870,370 Issued on 03/22/2005 to Bryan
| Inventors:
|
Bryan; Melissa Whitten (Albany, GA)
|
| Assignee:
|
Agri Dynamics, Inc. (Albany, GA)
|
| Appl. No.:
|
616466 |
| Filed:
|
July 9, 2003 |
| Current U.S. Class: |
324/331; 324/326; 324/330 |
| Intern'l Class: |
G01V 003//11 |
| Field of Search: |
324/323,326,327-329,330-331,334
|
References Cited [Referenced By]
U.S. Patent Documents
Primary Examiner: LeDynh; Bot
Attorney, Agent or Firm: Saitta; Thomas C.
Parent Case Text
CROSS REFERENCE TO RELATED APPLICATIONS
This application claims the benefit of and priority from U.S. provisional
patent application Ser. No. 60/395,089 and filed on Jul. 10, 2002, the
content of which is incorporated herein in its entirety.
Claims
What is claimed is:
1. A electromagnetic induction detection apparatus comprising:
a transmitter element that emits a primary magnetic field which induces a
secondary magnetic field in an external body;
a receiver element that receives the secondary magnetic field; and
a magnetic shield disposed around said receiver element that limits the
lateral footprint diameter of the secondary magnetic field observed by
said receiver element.
2. The electromagnetic induction detection apparatus of claim 1, wherein
said magnetic shield is constructed of magnetic field absorbant or
magnetic field reflective material.
3. The electromagnetic induction detection apparatus of claim 1, wherein
said magnetic shield is cone-shaped, said receiver element concentrically
disposed at the narrow end of said cone-shaped magnetic shield.
4. The electromagnetic induction detection apparatus of claim 1, wherein
said magnetic shield comprises an outwardly angled shield wall.
5. The electromagnetic induction detection apparatus of claim 3, wherein
said outwardly angled shield wall is sloped to form an angle between the
shield wall and the footprint surface within an open end of the magnetic
shield from 28.degree. to 90.degree..
6. The electromagnetic induction detection apparatus of claim 1, wherein
said transmitter element is an inductive coil.
7. The electromagnetic induction detection apparatus of claim 1, wherein
said receiver element is an inductive coil.
8. The electromagnetic induction detection apparatus of claim 1, wherein
said transmitter element, said receiver element are disposed in a
horizontal loop-loop configuration on a substantially rigid,
non-conductive support platform.
9. The electromagnetic induction detection apparatus of claim 8, wherein
said receiver element is mounted in a coplanar, displaced manner with
respect to said transmitter element on said support platform such that
said receiver element is substantially shielded from the primary magnetic
field emitted from said transmitter element.
10. The electromagnetic induction detection apparatus of claim 8, wherein
said transmitter element, said receiver element and said non-conductive
support platform form a discrete electromagnetic induction detection
apparatus that may be flown in a suspended manner below an aircraft.
11. The electromagnetic induction detection apparatus of claim 10, further
comprising an aircraft that transportably positions said electromagnetic
induction detection apparatus.
Description
BACKGROUND OF THE INVENTION
1. Technical Field
The present invention relates generally to detecting electromagnetic
induction sensing and processing, and in particular, to a mobile and
compact electromagnetic induction sensing system that may be utilized in
overhead detection applications for detecting small discrete objects such
as landmines.
2. Description of the Related Art
Electromagnetic induction (EMI) detection techniques are applied in a
variety of contexts including non-destructive testing of manufactured
objects, mineral exploration, treasure hunting, security checkpoints, and
detection of mines and unexploded ordinance (UXO). The hand-held metal
detector used for treasure hunting as well as landmine sweeping is a
familiar example of an EMI sensor. The operation of EMI sensors is based
on principles of electromagnetic induction in which one or more inductor
coils are utilized to interact with buried or otherwise hidden metallic
elements or objects. Specifically, an EMI sensor includes a transmitter
coil that emits a primary magnetic field into the surrounding environment
(a ground surface or container, for example). The primary magnetic field
induces eddy currents within nearby electromagnetic reactive elements,
resulting in emission of a secondary magnetic field which is measured by
the EMI sensor as an electric potential or electromotive force across a
receiver coil. For ease of reference, the terms "EMI sensor," "EMI
detector," and "metal detector" are utilized herein synonymously
throughout.
The metal detectors used for mine and UXO detection are remarkably
sensitive, capable of detecting buried objects containing less than a gram
of metal. Therefore, even with the advent and increasing utilization of
low metallic content mines and other ordinance, EMI sensors remain a
staple in the field of mine and UXO detection. The basic objectives of
detection using EMI sensors include obtaining the highest probability of
detection (P.sub.d) and the lowest false alarm rate. Furthermore, for wide
area assessment there is an increasing interest in the speed at which an
area can be covered.
EMI mine and UXO detection may be deployed using hand-held, ground based
vehicle-mounted or airborne detectors. Vehicular-mounted EMI detection
provides faster ground coverage but is limited to vehicle-accessible
terrain. Although effective for reliable detection over terrain
inaccessible by vehicle, hand-held EMI sensors pose the highest risk of
human injury of any of the methods and are ineffective for providing rapid
assessments of vast areas such as is often encountered during or after
wartime conditions in which minefields may span hundreds or thousands of
square miles. Furthermore, some terrain conditions such as mine or UXO
contaminated underwater or wetland environments may preclude use of either
vehicle-mounted or hand-held EMI detection.
Airborne electromagnetic induction (A-EMI) in which the detector is mounted
to the underside of an aircraft, such as a helicopter or fixed-wing
airplane, is effective for determining apparent conductivity in
near-surface geophysical studies. A-EMI induction systems are currently
used in many areas of environmental and geophysical exploration including
detection of mineral deposits, saltwater intrusion studies, and petroleum
exploration. Conventional A-EMI detectors provide much faster ground
coverage but have a lower P.sub.d when used for individual landmine or UXO
detection than the other two methods. A problem with conventional A-EMI
sensing methods when applied to mine and UXO detection is that the
altitude distance between the detector and the ground surface results in a
lateral "footprint" of the receiver coil being on the order of tens of
meters. Because the sensor response is averaged over the receiver
footprint, adequate resolution is not provided for relatively small
conductive or dielectric objects such as landmines. Furthermore, A-EMI
systems are physically unable to gain close proximity to the ground in
order to reduce the size of the footprint and therefore increase
resolution.
It can therefore be appreciated that a need exists for an improved A-EMI
sensing apparatus and system that enables rapid, scalable detection and
imaging to provide accurate and rapid detection of mines, UXO and the
like. The present invention addresses such a need.
SUMMARY OF THE INVENTION
A airborne electromagnetic induction (EMI) detection apparatus and system
are disclosed herein. In accordance with one embodiment, the EMI detection
apparatus includes a transmitter element in the form of a transmitter coil
that emits a primary, multi-frequency component magnetic field which
induces a secondary magnetic field in an external body. A receiver element
in the form of a receiver coil is mounted in a horizontal loop-loop
orientation with respect to the transmitter coil and receives the
secondary magnetic field. The detection apparatus further includes a
magnetic shield disposed around the receiver coil to limit the lateral
footprint diameter observed by the receiving element and to shield the
receiver coil from the primary magnetic field and other external
electromagnetic radiation to improve the gain and resolution of the
detection apparatus.
All objects, features, and advantages of the present invention will become
apparent in the following detailed written description.
BRIEF DESCRIPTION OF THE DRAWINGS
The novel features believed characteristic of the invention are set forth
in the appended claims. The invention itself however, as well as a
preferred mode of use, further objects and advantages thereof, will best
be understood by reference to the following detailed description of an
illustrative embodiment when read in conjunction with the accompanying
drawings, wherein:
FIG. 1 depicts an airborne electromagnetic induction (A-EMI) detection
system in accordance with one embodiment of the present invention;
FIG. 2 is an underneath view of an electromagnetic induction detection
platform incorporated in the A-EMI detection system of FIG. 1 and showing
the relative disposition of transmit and receive coils;
FIG. 3 is a block diagram illustrating the interfacing of signal and data
processing elements included within an A-EMI detection system in
accordance with one embodiment of the present invention; and
FIG. 4 is a high-level flow diagram depicting process steps utilized to
obtain multi-layer field conductivity profiles from received field
response data in accordance with one embodiment of the present invention.
DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT
The present invention is described in a preferred embodiment in the
following description with reference to the figures. While this invention
is described in terms of the best mode for achieving this invention's
objectives, it will be appreciated by those skilled in the art that
variations may be accomplished in view of these teachings without
deviating from the spirit or scope of the present invention. For example,
although the figures depict the system of the present invention as
deployed using aircraft transport and positioning, it should be noted that
the more general inventive principles taught herein are more broadly
applicable to any above-surface transport device.
The present invention is directed in general to an improved electromagnetic
induction (EMI) system that enables near placement of the detector to a
volume being examined and the detection, recording, and processing of
collected secondary magnetic field data. Specifically, the present
invention is directed to an apparatus, system and method for improving
response resolution when performing overhead scan EMI detection for mines,
unexploded ordinance (UXO), or other relatively small submerged or buried
objects. Previous problems with airborne systems used to locate landmines
and unexploded ordinance has been that the airborne system has a footprint
on the order of meters. Because the response is an average over the entire
footprint, a target measuring on the order of inches cannot be
discriminated from other signals in the footprint. The present invention
overcomes this problem by suspendedly flying an EMI detector from a
specified height, typically within a meter, from the ground.
The invention dramatically reduces the lateral detection footprint of an
A-EMI detection system, thus, making the system practical for use in
locating landmines and UXO. This system furthermore employs a method of
analyzing data from the search area that enables researchers to model the
region on a multiple-layer level in a timely fashion such that more
information is obtained from each survey.
In one embodiment, referred to herein as an airborne EMI detection system,
a remote-controlled drone helicopter is utilized to fly the detector as
EMI measurements are taken. The relatively small and maneuverable drone
tows an EMI detector platform that includes a transmitter and a receiver
coil. As explained below with reference to the figures, the receiver coil
is surrounded by a magnetic shield, which reduces the lateral footprint
through which a secondary magnetic field is received by the receiver coil,
enabling the detector system to detect relatively small objects as it
travels in a given direction at a given speed. Furthermore, the magnetic
shield reduces the effect of the primary magnetic field emitted from the
transmitter coil on the system response in addition to shielding the
receiver coil from other sources of electromagnetic interference. In a
preferred embodiment, the electromagnetic data obtained from the secondary
magnetic field response are processed using a modified Marquardt-Levenberg
type nonlinear least squares inversion algorithm to calculate the
resulting conductivities at corresponding depths for each data collection
point.
With reference now to the figures, wherein like reference numerals refer to
like and corresponding parts throughout, and in particular with reference
to FIG. 1, there is illustrated an airborne EMI (A-EMI) detection system
10 in accordance with one embodiment of the present invention. A-EMI
detection system 10 is designed to detect sub-surface objects, such as
landmines or UXO objects, in a discrete, individualized manner while
traversing at a given speed in a linear path over the object field surface
7. To this end, A-EMI detection system 10 includes a flying transport
craft which may be a fixed wing airplane or a helicopter. In a preferred
embodiment, the transport craft is a remote-controlled helicopter drone 2
having a compact design providing optimal maneuverability over rugged
terrain and under power lines and tree limbs, for example. Drone
helicopter 2 preferably has an approximate rotor diameter of 1.34' and is
capable of carrying a usable payload of about 22 lbs. Drone helicopter 2
is further preferably designed to maintain flying speeds between 15 mph
and 20 mph (ideally 17 mph) while collecting EMI data.
Suspended below helicopter drone 2 is a horizontal coplanar coil EMI sensor
apparatus that utilizes electromagnetic induction principles to detect and
map the location of buried landmines and the like. In accordance with
well-known EMI detection techniques, inductive coils are utilized to
induce and detect electromagnetic fields. Specifically, a current-carrying
transmitter coil 4 acts as a magnetic dipole that generates a primary
magnetic field, H.sub.p, proportional to a magnetic dipole moment,
.mu.=NIA, where A is the area of the loops of transmitter coil 4, I is the
current through the loops, and N is the number of loops in the coil. The
primary magnetic field induces eddy currents in magnetic field-reactive
bodies or targets (not depicted) within a volume located a specified
distance from the transmitter coil. The induced currents in the volume in
turn generate a secondary magnetic field, H.sub.s, which is detected as a
changing electric potential, or electromotive force, in a receiver coil 5
above the volume. The ratio of H.sub.s to H.sub.p is called the mutual
coupling ratio, Q, and is the quantity utilized by an EMI system utilized
to detect the location of buried magnetic objects.
For the horizontal coplanar coil configuration depicted in FIGS. 1 and 2,
also known as a vertical dipole configuration, Q is given by:
Q=H.sub.s /H.sub.p =-r.sup.3.intg..lambda..sup.2 R(.lambda.)J.sub.0
(.lambda.r)e.sup.-2.lambda.h d.lambda.,
wherein R represents the total complex distance between the primary dipole
and the secondary dipole in the conductor, h is the detector height above
the ground, J.sub.0 is the 0.sup.th order Bessel function, and the
integration limits are taken over the coil separation, r, from r=0 to
r-.infin.. The mutual coupling ratio is measured for a particular
frequency, and inversion of the equation for Q gives R(.lambda.), from
which the embedded parameters of apparent conductivity, .sigma..sub.a, and
skin depth, .delta..sub.s, are obtained. The apparent conductivity and
skin depth parameters, .sigma..sub.a and .delta..sub.s, are dependent on
coil separation, height of the detector and the transmitter frequency.
The EMI sensor apparatus depicted in FIG. 1 comprises a loop-loop
horizontal coplanar set of transmitter and receiver coils mounted on a
non-conductive EMI detection platform 8 that is towed below the
remote-controlled drone helicopter 2 using a cable 6 or other suitable
suspension support member. Specifically, and referring to the underneath
profile depiction of EMI detection platform 8 shown in FIG. 2 in
conjunction with FIG. 1, the EMI sensor apparatus includes a
small-diameter transmitter coil 4 and a small-diameter receiver coil 5
both mounted to the underside surface of EMI detection platform 8, which
is preferably constructed of a light, non-conductive material such as
light plastic polymers.
Transmitter coil 4 and receiver coil 5, having respective, specified radii
and numbers of loops are mounted on EMI detection platform 8 using epoxy
glue or other suitable adhesive or mechanical fastening means. The
windings of transmitter and receiver coils 4 and 5 are formed such that
the respective coil radii are held in exact circular uniformity and the
concentricity of the coil loops is maintained.
As further depicted in FIGS. 1 and 2, receiver coil 5 is displaced by a
specified distance, D, from transmitter coil 4 as measured between the
respective coil centers. The dimensions and relative positioning of
transmitter and receiver coils 4 and 5 is significant to A-EMI detection
because receiver coil 5 measures the in-phase (real) and quadrature
(imaginary) components of the secondary (induced) field in terms of a
percentage of the primary (source) field. Transmitter and receiver coils 4
and 5 are preferably mutually positioned on EMI detection platform 8 in a
manner that maximizes this ratio. Preferably, the coil separation D is
small compared to the skin depth of the system, such that the primary
field is reduced in the detected signal and the value of Q is greater. The
horizontal coplanar mode has a maximum sensitivity to conductive layers at
a depth of about 0.4 times the coil separation. Furthermore, the
displacement D between the centers of transmitter coil 4 and receiver coil
5 is preferably maximized to the extent practicable on EMI detection
platform 8 to reduce the source field's detection by receiver coil 5 and
further maximize Q. Additionally, the spacing distance D between the coils
must be less than the height of receiver coil 5 from the ground surface 7.
Typically, the detector is flown at a height of about 10-20" from the
ground, with the distance D between transmitter and receiver coils 4 and 5
preferably set between 15"-20" to give a maximum sensitivity at a depth of
penetration of about 6" to 8".
Transmitter and receiver coils 4 and 5 preferably have a radius of about
0.5". Unshielded, the receiver footprint diameter grows in proportion to
height of the detector apparatus above the ground (approximately 19" with
the detector less than 20" from the ground). As explained in further
detail below, a magnetic shield 11 is deployed around receiver coil 5 to
absorb or deflect unwanted secondary field values to limit the footprint
size. The reduced lateral footprint area resulting from magnetic shield 11
enables the EMI detection apparatus to maintain adequate resolution when
the detection platform 8 is flown at a higher altitude over rugged or
obstacle-filled terrains.
As further depicted in FIGS. 1 and 2, and in a preferred embodiment of the
present invention, receiver coil 5 is mounted within the inner volume of
magnetic shield 11, which protrudes downwardly from the underneath surface
of EMI detection platform 8. Its constituent materials and contour enable
magnetic shield 11 to limit the lateral footprint, i.e. surface area of
incoming magnetic field exposure, of receiver coil 5 to a specified width
suitable for detecting EMI reactive objects having dimensions on the order
of inches. Magnetic shield 11 may be constructed entirely of a magnetic
shielding material such as aluminum, or in the alternative may be
constructed of a lightweight non-conductive material layered with magnetic
shielding material. Secure attachment to platform 8 may be achieved by
inserting an outer sleeve of magnetic shield 11 into a circular hole
within EMI detection platform 8 using epoxy glue or other suitable
adhesive or mechanical fastener means such that receiver coil 5 is
maintained level, i.e. coplanar, with respect to transmitter coil 4.
The effective resolution of the EMI detection apparatus is significantly
increased by the disposition of magnetic shield 11 around receiver coil 5
in a two-fold manner. Specifically, in addition to limiting the lateral
footprint observed by receiver coil 5, the disposition of magnetic shield
11 around receiver coil 5 serves to greatly reduce or eliminate the
primary magnetic field and environmental sources of electromagnetic noise
received by receiver coil 5. As depicted in FIGS. 1 and 2, magnetic shield
11 is preferably cone or frustum contoured. However, magnetic shield 11
may be implemented using other shapes and contours without departing from
the spirit or scope of the present invention.
In a preferred embodiment, the slope angle of the side of magnetic shield
11 is such that the angle, .theta..sub.fp, between the shield side and the
footprint surface (as represented by the open mouth surface area bounded
by the rim of magnetic shield 11) is between 28.degree. and 90.degree.. At
28.degree., none of the secondary magnetic field lateral footprint is
shielded due to the field characteristics of the secondary field which
leaves an angle between the footprint and the ground of approximately
28.degree.. At 90.degree., wherein the shield is effectively a cylinder
around receiver coil 5, nearly all of the secondary magnetic field
response is blocked.
In combination, the relative positioning of transmitter coil 4 and receiver
coil 5, in which the coils are mutually separated as much as is
practicable on EMI detection platform 8, and shielding function provided
by magnetic shield 11, minimizes the effect of the primary magnetic field
and noise on the response signal received by receiver coil 5, thereby
improving resolution by maximizing the detected ratio of the secondary
magnetic field to the primary magnetic field.
As further illustrated in FIG. 1, a dual-frequency global positioning
system (GPS) receiver 16 is mounted on EMI detection platform 8 directly
above receiver coil 5. In accordance with well-known GPS technology, GPS
receiver 16 receives longitude, latitude and altitude information from a
set of GPS satellites (not depicted). As explained in further detail with
reference to FIG. 3, the spatial position information collected by GPS
receiver 16 is advantageously utilized for mapping the collected EMI data
points.
Also mounted on EMI detection platform 8 are a detonator device 14 and an
altimeter device 12. Detonator device 14 may be advantageously utilized in
conjunction with the spatial position data provided by GPS receiver 16 and
target detection data to detonate or otherwise neutralize landmines or UXO
detected by the EMI detection apparatus. Detonator device 14 may be a
small-caliber gun, such as a .22 caliber gun, which is lightweight and
uses lightweight ammunition. In the alternative, detonator device 14 may
be a high-power laser or a device that controllably emits a high-energy
sonic boom as a detonation mechanism. Regardless of the detonation
mechanism employed, and as depicted in FIG. 3, detonator device 14
receives targeting instructions in accordance with current position data
from GPS receiver 16 as well as target position information from a host
data processing system 52 within helicopter drone 2 enabling detonator
device 14 to accurately aim and strike detected targets. Altimeter device
12 is preferably a laser altimeter that detects and reports to data
processor 52, the current height of platform 8 above the ground surface 7
thereby enabling helicopter drone 2 to adjust its altitude as necessary
over, for example, uneven or steeply sloping terrain.
Differing conductive properties of elements fluxed by the primary magnetic
field affect the response detected by receiver coil 5. Even dielectric
materials cause anomalies in the detected secondary magnetic field when
exposed to a high frequency source. Thus, a broad range of frequencies
must be considered in a multiple frequency electromagnetic induction
system in order to exploit the dielectric properties of less conductive
materials in the volume below a receiver coil. These changes in the
responses from a dielectric material are detectable and are evident in the
processed data.
The present invention exploits the frequency response characteristics of
different materials and furthermore uses multi-frequency induction and
data processing techniques to accurately detect objects at different
ground depths. With reference to FIG. 3, there is depicted a block diagram
illustrating the interfacing of signal and data processing elements
included within A-EMI detection system 10 in accordance with one
embodiment of the present invention. Specifically, a data processing
system 52 is shown which may be deployed either on drone 2 or EMI
detection platform 8. As shown in FIG. 3, data processing system 52
includes a pulse width modulated (PWM) waveform generator 54 that converts
a digital signal input from an alternating current (AC) 53 source into an
arbitrary multi-frequency component waveform. In a preferred embodiment,
waveform generator 54 generates a pulse width modulated output signal
having multiple frequency components which are input as a multi-frequency
current signal into the windings of transmitter coil 4. Using a source
signal comprised of multiple frequency components results in more
comprehensive location information because responses are detected and
recorded for each corresponding input frequency. The different frequencies
induce responses at different layers in the volume being examined below
the transmitter/receiver coil system. In general, lower frequencies induce
responses in targets that are deeper within the volume, while higher
frequencies induce responses in targets closer to the surface.
In this manner, so-called frequency depth sounding occurs wherein multiple
frequencies are used to see further into the region below a sensor. Each
frequency in a frequency-domain system models a different layer of the
conductive earth below the transmitter/receiver pair. Low frequency
signals look deeper into the earth, while high frequency signals can only
travel short distances and sees only shallow structures. Higher frequency
signals are more practical for detecting mines and UXO for two reasons.
First, higher frequency signals induce target responses at shallower
volume depths where mines and UXO typically reside. Additionally, higher
frequency signals exploit the dielectric properties of less conductive
materials, commonly found in explosives, which are consequently
detectable.
The multi-frequency signal generated by waveform generator 54 induces a
primary magnetic field having corresponding multi-frequency
characteristics that is transmitted from transmitter coil 4. Among the
multiple frequency components, high-frequency signals exploit the
dielectric properties of less conductive materials, enabling the system to
detect dielectric materials such as explosives. Additionally, a broad
range of frequencies allows for a broad range of depths to be examined.
The primary magnetic field induces eddy currents in targets contained
within the ground volume below the detector that in turn generate a
secondary, or response magnetic field which is detected by measuring the
electromotive force, or electric potential in receiver coil 5 as receiver
coil 7 follows transmitter coil 4 over the sampled location. Receiver coil
5 detects the secondary magnetic field as a percentage of the primary
field. The resulting induced secondary magnetic field is preferably
recorded at a sampling rate of 30 times per second for a 60 Hz power
supply on a separate tape or other data recording media according to
fiducial numbers.
Data processing system 52 further includes processing means in the form of
a microprocessor 55 and a digital signal processor (DSP) 56. In accordance
with the depicted embodiment, the induced electromotive force generated
from the secondary magnetic field is converted to a digital signal by an
analog-to-digital (A/D) converter 58 before being received and processed
by DSP 56 and microprocessor 55 which then compare the phase and amplitude
of the current from the transmitter coil with the phase and amplitude of
the induced electromotive force from the receiver coil to deduce
corresponding detected secondary to primary magnetic field ratio from
which parameters such as conductivity and depth can be extracted.
The responses are recorded within processor memory or a local storage
device 51 as a ratio of the secondary field to the primary field. A
wireless network interface 59, as an RF transmission interface, may be
utilized to transmit field-collected data from onboard data processing
system 52 to remote networks or computers where the data may be processed
in accordance with the post-collection processing techniques disclosed
herein. In a preferred embodiment, the collected field data is processed
using a modified Marquardt-Levenberg type nonlinear least squares
inversion algorithm applied to a multiple-layer model, which will convert
the detected response to conductivity values at corresponding depths for
each latitude/longitude collection point as determined by GPS receiver 16.
Once the data has been successfully inverted to conductivity and depth
values at each collection point, contour graphs of conductivity vs.
latitude/longitude and corresponding depth vs. latitude/longitude will be
used to precisely map the mine locations. Using a laptop PC, this least
squares algorithm can process a line of 477 data points in approximately
60 seconds. Visualization can be performed using any commercial plotting
software program.
The dual-frequency GPS system (airborne version), comprising GPS receiver
16, is used to record the data collection point lateral position to an
accuracy of 1 cm and the corresponding altitude to an accuracy of 2 cm.
The GPS data collected at each acquisition site will be recorded on a tape
according to fiducial numbers. Maps of the area can be generated to show
conductivity and depths at each latitude/longitude.
Combination of GPS data with secondary field data will be performed offsite
by matching fiducial numbers. Further analysis will be performed using a
modified Levenber-Marquardt algorithm applied to a multiple-layer model,
which will convert the detected response to conductivity values at
corresponding depths for each latitude/longitude collection point. Once
the data has been successfully inverted to conductivity and depth values
at each collection point, contour graphs of conductivity vs.
latitude/longitude and depths vs. latitude/longitude will be used to
precisely map the mine locations.
In accordance with a preferred embodiment, a modified Marquardt-Levenberg
type nonlinear least squares inversion algorithm is utilized to estimate
values for the model parameters, using a multiple-layer forward algorithm
to compute the frequency-domain responses to a signal transmitted by a
horizontal coplanar loop-loop orientation, and compares these responses to
the actual measured data using a trust region approach. The inverse
algorithm then uses the discrepancies between predicted and actual to
improve its guess for the parameters. This procedure is iterated to
improve the estimates. The forward algorithm computes the Hankel functions
using a method of weighting the zeroes of the Bessel function. The model
can be modified for two to ten layers with the horizontal coplanar
loop-loop system. The processing time for this algorithm applied to three
layers is approximately 60 seconds for 480 data points for a six-frequency
detection signal.
The collected data are processed using a modified Marquardt-Levenberg type
nonlinear least squares algorithm, which estimates values for the model
parameters, uses a forward model to predict the response to a signal
transmitted by the source, and compares this response to the actual
measured data. The algorithm then uses the discrepancies between predicted
and actual to improve its guess for the parameters. This procedure is
iterated to improve the estimates. The forward algorithm computes the
Frischknecht integral by computing the Hankel functions using a method of
weighting the zeroes of the Bessel function.
The multiple-layer forward algorithm computes the frequency-domain
responses for a horizontal coplanar loop-loop orientation by computing the
Hankel functions using a method of weighting the zeroes of the Bessel
function. The subroutine can be modified for two to ten layers with a
horizontal coplanar loop-loop system.
If the observed values for inphase are Iph(i) and for quadphase are Qph(i)
at each frequency, i, then let a forward algorithm take model x and
compute the corresponding inphase value for the model x at each frequency
i, Ifr[i,x], and quadphase value for the model x at each frequency i,
Qfr[i,x]. The least squares residual function is
r(x)=1/2.SIGMA.((Iph(i)-Ifr[i,x]).sup.2 +(Qph(i)-Qfr[i,x]).sup.2)
where the sum is over the frequencies used
The Marquardt-Levenberg type nonlinear least squares inversion algorithm is
an IMSL subroutine that is used to produce a series of models x.sub.k that
converges to an approximate minimizer of the residuals r(x). This
algorithm updates a model x.sub.k to x.sub.k+1 =x.sub.k +p.sub.k by a
trust region approach that seeks p.sub.k as the solution to
min.sub.p.parallel.J.sub.kp +r(x.sub.k).parallel..sup.2
subject to
.parallel.p.sub.k.parallel..ltoreq..DELTA.,
where .DELTA. is the radius of the trust region and J.sub.k is a finite
difference approximation to the Jacobian of r(x) at x.sub.k. The minimum
will be a quasi-Newton step
J.sub.k.sup.T J.sub.k p.sub.k =-J.sub.k.sup.T r(x.sub.k)
if p.sub.k is within the trust region so that .parallel.p.sub.k
<.DELTA.. If .parallel.p.sub.k.parallel.=.DELTA., a line search method
finds a scalar .lambda.>0 such that
(J.sub.k.sup.T J.sub.k +.lambda.I)p.sub.k =-J.sub.k.sup.T r(x.sub.k).
The .parallel.p.sub.k.parallel.=.DELTA. case occurs when x.sub.k is far
from a minimizer and J.sub.k.sup.T J.sub.k may have zero or near zero
eigenvalues. The .lambda.I term in (J.sub.k.sup.T J.sub.k +.lambda.I)
increases the eigenvalues by an amount .lambda. and assures a positive
definite coefficient matrix. Once a minimizer is approached, the
.parallel.p.sub.k.parallel.<.DELTA. case takes control and the
algorithm assumes the rapid convergence characteristics of traditional
Gauss-Newton methods. Further explanation of data processing techniques
applied to gathered AEM induction data is set forth in "Comparison Of MIM
and Least Squares Inversions For Barataria Bay AEM Data".COPYRGT.,
authored by Dr. Melissa Whitten Bryan. The content of "Comparison Of MIM
and Least Squares Inversions For Barataria Bay AEM Data".COPYRGT. is
incorporated in its entirety herein by reference.
Referring to FIG. 4, there is illustrated a high-level flow diagram
depicting process steps utilized to obtain multi-layer field conductivity
profiles from received electromagnetic induction field response data (i.e.
data retrieved from the EMI response signals received by receiver coil 5)
utilizing the modified Marquardt-Levenberg type nonlinear least squares
inversion algorithm in accordance with one embodiment of the present
invention. The process begins as shown at step 62 and proceeds to step 64
depicting one or more requests for initial parameter estimates, or
guesses, based on the number of frequencies utilized in the EMI
transmitter/receiver detection system. The initial parameter estimates
(i.e., layer depths and corresponding conductivities) are applied as the
initial forward model solution as illustrated at step 66. Next, as
depicted at step 68 a forward model subroutine that can be modified for
two to ten layers is utilized to compute the frequency-domain response of
a signal transmitted by the source of a horizontal loop-loop
configuration. The subroutine preferably computes the Frischknecht
integral using a method of weighting the zeroes of the Bessel function.
Proceeding to steps 72 and 74, the Jacobian of the residual function is
computed at a point using a finite difference approximation and an
inversion method is utilized to invert the responses to the model
parameters. If the current assessment is not complete, the resultant
predicted response is compare with the actual measured data using a trust
region approach that seeks to minimize the sum of the Jacobian and the
least squares residual function as shown at steps 76 and 78. Next, as
illustrated at steps 82 and 84, the discrepancies between the predicted
response and the actual response is utilized to adjust the next set of
parameter estimates which are again processed beginning at step 68 in an
iterative manner.
The foregoing description discloses a compact A-EMI detection system that
employs a magnetic shield to greatly reduce the size of the footprint and
increase the ability to discriminate small objects as well as dramatically
reduce the time required to detect and therefore clear mine fields.
Sampling at a rate of 30 times per second and flying a speed of 17 mph,
this system is capable of detecting mines or UXO at the rate of 10 acres
per day. Using multiple input frequencies give more information about the
volume being examined; therefore, a broader range of materials can be
detected including dielectric materials such as explosives and landmines
made using very little or no metal. Additionally, multiple depths can be
examined in a volume to see objects buried at various depths. After
processing using the modified Marquardt-Levenberg type nonlinear least
squares algorithm, maps can be produced of conductivity vs. position as
well as conductivity vs. depth.
While this invention has been described in terms of several embodiments, it
is contemplated that alterations, permutations, and equivalents thereof
will become apparent to one of ordinary skill in the art upon reading this
specification in view of the drawings supplied herewith. It is therefore
intended that the invention and any claims related thereto include all
such alterations, permutations, and equivalents that are encompassed by
the spirit and scope of this invention.
*
