Data processing: measuring – calibrating – or testing – Measurement system in a specific environment – Earth science
Reexamination Certificate
1999-12-14
2001-06-26
McElheny, Jr., Donald E. (Department: 2862)
Data processing: measuring, calibrating, or testing
Measurement system in a specific environment
Earth science
Reexamination Certificate
active
06253156
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a synchronized seismic signal acquisition method allowing resynchronization of seismic signal acquisition on an exterior event, and to a device for implementing it.
The method according to the invention finds applications in many fields where various measured signals are to be sampled for example, by means of different acquisition chains, by imposing that the series of samples taken are substantially synchronous with an exterior event. This is notably the case in the field of seismic exploration where an initial time from which significant seismic signals are recorded is to be fixed for an acquisition system.
The initial reference time generally selected is the time of triggering of a source of seismic waves. The waves emitted are propagated in the subsoil and received by seismic pickups distributed at the ground surface for example. The signals delivered by these pickups are transmitted to a central control and recording station, generally by means of acquisition devices distributed in the field. Each one of them is suited to amplify, filter, digitize and store all the signals picked up after each source triggering. The stored data are transmitted to a central station from each acquisition device at fixed time intervals (after each emission-reception cycle for example) or “with the stream”, as soon as a transmission time interval is available. Seismic acquisition systems are for example described in patents FR-2,511,772 (U.S. Pat. No. -4,583,206) or FR-2,538,194 (U.S. Pat. No. -4,628,494).
In each acquisition device, the seismic signals are applied to an acquisition chain. A conventional acquisition chain structure comprises a steady-gain preamplifier, a high-pass filter, an anti-aliasing low-pass filter and an analog-to-digital (ADC) converter. The converters deliver for example 24-bit numerical words. They are for example (sigma-delta type) oversampling analog-to-digital converters associated with digital filters (FIR).
Oversampling converters produce numerical words of reduced format in relation to conventional converters, but with a much higher frequency. The normal dynamic range is restored by applying to the signals coming from the converter a digital filter referred to as decimation filter which, besides its anti-aliasing filtering functions, is suited to stack a determined number of samples with appropriate weightings as it is well-known to specialists.
An analog-to-digital converter digitizes series of analog samples taken from a signal at times fixed by an internal clock. This is no drawback when the converter works in isolation. It becomes a drawback in all the cases where it is desired to precisely fix an initial reference time in relation to which a sequence of events is located, and especially when signal acquisitions are to be carried out by a series of different converters.
In seismic prospecting operations notably, the seismic waves coming from the subsoil as a result of an emission by a source of seismic waves are picked up by a multiplicity of receivers and converted to digitized samples by an often considerable number of different acquisition chains provided each with an analog-to-digital converter. A reference time is selected, generally the time of triggering of the seismic source, and one tries to adjust in relation to this time the first significant sample taken by the various converters on each signal picked up by the receivers. If the sampling time of each converter only depends on an internal clock, there is no reason to be synchronized with the exterior event selected as the reference. A certain random delay or jitter follows therefrom, which is generally different from one acquisition chain to the next. The consequence thereof is a lack of synchronization that is very disturbing when signals received and acquired by different acquisition chains have to be combined, as it is generally the case in conventional seismic processing.
BACKGROUND OF THE INVENTION
Patent FR-2,666,946 (U.S. Pat. No. 5,245,647) filed by the applicant discloses a signal sampling device comprising in combination a sigma-delta type oversampling converter associated with a FIR type digital filter performing decimation of successive series of oversamples and a device for synchronizing the samples delivered with an exterior event such as the time of triggering of a seismic source for example. The solution used in this prior device essentially consists in a memory inserted between the sigma-delta converter and the decimation filter, wherein a series of oversamples is permanently stored. On reception of an exterior reference signal, the device is suited to find in the inserted memory the oversamples formed before reception of this signal and to command transfer thereof in the decimation filter so as to produce the first of the resynchronized samples.
Although this solution is perfectly operational, it has the drawback of requiring complex and expensive electronic components inserted between the delta-sigma modulator and the FIR anti-aliasing filter, i.e. a memory and relatively complex means for managing it.
There are also well-known fractional (less than one unit) delay processing techniques notably described by: Laakso T. I. et al : Splitting the Unit Delay; in IEEE Signal Processing Magazine; 1996, allowing to carry out, by means of calculations, time readjustment of the signal sampling. Certain principles thereof, useful for better understanding of the method, are reminded hereafter.
x[n] denotes a series of digitized samples S
k
, S
k+1
, S
k+2
. . . S
k+p
, etc, taken (
FIG. 1
) from a measuring signal from an initial time t
0
on, with a sampling interval &Dgr;t, by an analog-to-digital converter, and y[n] denotes a series of samples S′
1
, S′
2
, S′
3
. . . S′
p+1
, etc, taken with the same interval from the same measuring signal but readjusted in time from a reference time T
R
after t
0
. The readjustment time difference D is a positive real number.
This number can generally be written as follows: D=int(D)+d, where int(D) corresponds to a whole number of sampling periods and d is a fraction of a period.
We must have: y[n]=x[n−D].
In order to obtain a delay int(D), it is sufficient to delay the initial signal x[n] by a simple translation. The samples of y[n] are those of x[n] whose index is simply delayed (renumbered) by int(D). The sample bearing number k in the first series for example becomes the sample bearing number 1 in the second series, with k.&Dgr;t=int(D). For the fractional part of this time difference, the readjusted samples y[n] will be somewhere between the values of x[n] at two successive sampling positions by the local clock and they must best correspond to the effective amplitudes of the sampled signals at these intermediate positions. This delay with readjustment can be obtained by applying a digital filtering F (FIG.
2
).
With the notations specific to the z transform, this delay by digital filtering can be expressed as follows:
Y
(
z
)=
X
(
z
).
z
−D
.
The frequency response of the ideal filter H
ID
is:
H
id
=z
−D
=e
−j&ohgr;D
with
z=e
j&ohgr;
.
The amplitude and phase responses of the ideal filter for any &ohgr; are therefore:
|H
id
(
e
j&ohgr;
)|=1
and
arg[H
id
(
e
j&ohgr;
)]=&thgr;
id
(&ohgr;)
=−D&ohgr;.
The phase is often represented as a phase lag defined by:
τ
p
⁡
(
ω
)
=
-
θ
id
⁡
(
ω
)
ω
,
a lag that is here D.
The corresponding impulse response is obtained by inverse Fourier transform:
h
id
⁡
[
n
]
=
1
2
⁢
π
⁢
∫
-
π
π
⁢
H
id
⁡
(
ⅇ
j
⁢
⁢
ω
)
⁢
ⅇ
j
⁢
⁢
ω
⁢
⁢
n
⁢
⁢
Δ
⁢
⁢
t
⁢
ⅆ
ω
for any n, hence:
h
id
⁡
[
n
]
=
sin
⁡
[
π
⁡
(
n
⁢
⁢
&Delt
Bui-Tran Van
Nguyen Thi Thu
Thierry Gerard
Antonelli Terry Stout & Kraus LLP
Institut Francais du Pe'trole
McElheny Jr. Donald E.
LandOfFree
Synchronized seismic signal acquisition method and device does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Synchronized seismic signal acquisition method and device, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Synchronized seismic signal acquisition method and device will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2529556