Ex Parte Wu et al

Patent Trial and Appeal Board

Feb 7, 2018

13392035 (P.T.A.B. Feb. 7, 2018)

United States Patent and Trademark Office
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O.Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
13/392,035 02/23/2012 Xiaohui Wu 2009EM274 4425
34477 7590 02/09/2018
ExxonMobil Upstream Research Company
22777 Springwoods Village Parkway
(EMHC-E2-4A-296)
Spring, TX 77389
EXAMINER
BROCK, ROBERT S
ART UNIT PAPER NUMBER
2128
NOTIFICATION DATE DELIVERY MODE
02/09/2018 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
urc-mail-formalities @ exxonmobil. com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte XIAOHUI WU, DACHANG LI, ROSSEN PARASHKEVOV,
ADAM K. USADI, and YAHAN YANG
Appeal 2017-008964
Application 13/392,035
Technology Center 2100
Before ROBERT E. NAPPI, DAVID M. KOHUT, and
LYNNE E. PETTIGREW, Administrative Patent Judges.
PETTIGREW, Administrative Patent Judge.
DECISION ON APPEAL
This is an appeal under 35 U.S.C. § 134(a) from a final rejection of
claims 1, 3—17, 19, and 20. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm.
Appeal 2017-008964
Application 13/392,035
STATEMENT OF THE CASE1
Introduction
Appellants’ invention is directed to a method and system to decrease
the number of simulations that may be run to generate a model for predicting
the properties of a hydrocarbon reservoir. Spec. 12.
Claim 1 is illustrative of the invention and is reproduced below.
1. A method for improving hydrocarbon production by
lowering computational costs of multiple simulations,
comprising:
using a computer program product, comprising a
computer usable medium having a computer
readable program code embodied therein, said
computer readable program code adapted to be
executed to implement a method comprising:
identifying a lowest level training set (TS1) in a
lowest level physics based surrogate (PBM1),
wherein identifying TS1 comprises:
sampling a parameter space to obtain a first
set of sample points;
evaluating a response for PBM1 at each of
the sample points;
training a data fit surrogate (DFS°) over the
parameter space using the response at each
of the sample points; and
1 Our Decision makes reference to Appellants’ Reply Brief (“Reply Br.,”
filed June 6, 2017) and Appeal Brief (“App. Br.,” filed January 26, 2017),
and the Advisory Action (“Adv. Act.,” mailed October 26, 2016),
Examiner’s Answer (“Ans.,” mailed April 6, 2017), and Final Office Action
(“Final Act.,” mailed July 14, 2016).
2
Appeal 2017-008964
Application 13/392,035
identifying a second set of sample points at
TS1, wherein a response of the DFS°
exhibits a critical behavior at the sample
points;
generating at least one higher level training set
(TSk) for at least one higher level PBMk, wherein 1
< k < K, and wherein K represents a highest level;
and
training the at least one higher level PBMk at TSk
using a next higher level physics based surrogate
(PBMk+1);
generating a model for predicting the properties of
a hydrocarbon reservoir using at least in part a
result obtained by training the at least one higher
level PBMk at TSk, wherein one or more lower
level surrogates are used to evaluate an entire
parameter space for the model; and
using the model in performance planning of a
hydrocarbon reservoir, enhancement of a
hydrocarbon reservoir, or a combination thereof.
Rejections on Appeal
The Examiner rejected claims 1, 3—14, 17, 19, and 20 under 35 U.S.C.
§ 103(a) as unpatentable over Forrester (A. I. J. Forrester et al., Recent
advances in surrogate based optimization in 45 PROGRESS IN Aerospace
Sciences, 50—79 (2009)) in view of Wu (X. H. Wu et al., Reservoir
Modeling with Global Scaleup, Society of Petroleum Engineers 105237,
15th SPE Middle East Oil & Gas Show and Conf., 1-13 (2007)).
The Examiner rejected claim 15 under 35 U.S.C. § 103(a) as
unpatentable over Forrester, Wu, and Badru (O. Badru & C. S. Kabir, Well
3
Appeal 2017-008964
Application 13/392,035
Placement Optimization in Field Development, SPE Annual Technical
Conference and Exhibition, 1-9 (2003)).
The Examiner rejected claim 16 under 35 U.S.C. § 103(a) as
unpatentable over Forrester, Wu, and Altamira (US 3,554,282 issued Jan.
12, 1971).
Issues on Appeal
The issues in this appeal are:
Did the Examiner err in finding that it would have been obvious to
combine Forrester and Wu and further to combine Forrester and Wu with
either Badru or Altamira? App. Br. 10—13; Reply Br. 5.
Did the Examiner err in rejecting independent claim 1, and similarly
independent claims 14 and 17, under 35 U.S.C. § 103(a) because the
combination of Forrester and Wu fails to teach or suggest “identifying a
lowest level training set (TS1) in a lowest level physics based surrogate
(PBM1), wherein identifying TS1 comprises: sampling a parameter space to
obtain a first set of sample points; evaluating a response for PBM1 at each of
the sample points; training a data fit surrogate (DFS°) over the parameter
space using the response at each of the sample points; and identifying a
second set of sample points at TS1, wherein a response of the DFS° exhibits
a critical behavior at the sample points; [and] generating a model for
predicting the properties of a hydrocarbon reservoir using at least in part a
result obtained by training the at least one higher level PBMk at TSk,
wherein one or more lower level surrogates are used to evaluate an entire
parameter space for the model”? App. Br. 7—10; Reply Br. 2-4.
4
Appeal 2017-008964
Application 13/392,035
ANALYSIS
For all of the proposed combinations of references, Appellants argue
that the Examiner impermissibly relied on hindsight, because the references
are directed to different technical problems, and, therefore, the record
remains unclear as to how and why the references would be combined. App.
Br. 10, 11, 13.
We are not persuaded by Appellants’ argument, which too narrowly
defines the particular problem addressed in the references. “Whether a
reference in the prior art is ‘analogous’ is a fact question.” In re Clay, 966
F.2d 656, 658 (Fed. Cir. 1992) (citing Panduit Corp. v. Dennison Mfg. Co.,
810 F.2d 1561, 1568 n.9 (Fed. Cir. 1987)). Two criteria have evolved for
answering the question: “(1) whether the art is from the same field of
endeavor, regardless of the problem addressed, and (2) if the reference is not
within the field of the inventor’s endeavor, whether the reference still is
reasonably pertinent to the particular problem with which the inventor is
involved.” Id. at 658—59 (citing In re Deminski, 796 F.2d 436, 442 (Fed.
Cir. 1986); In re Wood, 599 F.2d 1032, 1036 (CCPA 1979)). Appellants’
invention involves techniques to decrease the number of full physics
simulations that may be run to generate a model for predicting the properties
of a hydrocarbon reservoir. See Spec. 12. Therefore, the references that
deal with surrogate/modelling based analysis are reasonably pertinent to the
particular problem with which the inventor is involved because the
references logically would have commended themselves to the attention of
one considering the invention as a whole. See Final Act. 6, 15, 16. Thus,
Appellants’ hindsight argument is unpersuasive because it is based on the
incorrect conclusion that the references are non-analogous rather than
5
Appeal 2017-008964
Application 13/392,035
explaining why the Examiner’s articulated reasoning does not support the
conclusion of obviousness.
Regarding independent claims 1,14, and 17, the Examiner finds
Forrester teaches sampling a parameter space to obtain a first set of sample
points (initial samples (Final Act. 3; Adv. Act. 2; Ans. 5)); evaluating a
response for the lowest level physics based surrogate at each of the sample
points (true function evaluation (Final Act. 4; Ans. 3)); training a data fit
surrogate (Kriging model) over the parameter space using the response at
each of the sample points (refitting the surrogate to the data (Final Act. 4;
Adv. Act. 2; Ans. 5)); and identifying a second set of sample points as a
lowest level training set (initial samples and infill points (Final Act. 4; Ans.
4)), wherein a response of the data fit surrogate exhibits a critical behavior at
the sample points (infill points used to isolate a local minimum (Final Act. 4;
Ans. 4—5)); and generating a model for predicting properties using at least in
part a result obtained by training the at least one higher level physics based
surrogate at a higher level training set, wherein one or more lower level
surrogates are used to evaluate an entire parameter space for the model
(exploration-based infill (Final Act. 5 (citing Forrester p. 68,12
“Exploration-based infill has its niche in design space visualization and
comprehension where the object is to build an accurate approximation of the
entire design landscape to help the designer visualize and understand the
design environment they are working in.”))).
Appellants contend that the Examiner erred in finding the
combination of Forrester and Wu teaches or suggests the limitations of
claim 1. App. Br. 6—9. Specifically, Appellants argue Forrester does not
start with simulations at the lowest level surrogate, but begins with
6
Appeal 2017-008964
Application 13/392,035
calculations at the highest level (expensive data) and evaluates the surrogate,
and then uses calculations at the lower levels (cheap data) to correct the
surrogate. App. Br. 7—9. Appellants further argue the Examiner
misconstrued Forrester’s surrogate as the physics based surrogate when it is
the data fit surrogate. App. Br. 9; Reply. Br. 3. Appellants argue Wu does
not cure the deficiencies of Forrester. App. Br. 9.
Appellants’ arguments do not persuade us of error with respect to the
Examiner’s findings concerning Forrester. The Examiner finds Forrester’s
surrogate optimization is the same as the claimed “identifying a lowest level
training set” (sampling, evaluating, training), where the optimization is done
using a physics based surrogate corrected with a data fit surrogate, as
claimed. Ans. 3^4. The Examiner finds the multi-fidelity analysis method
of co-Kriging takes advantage of using a physics based surrogate model and
the data fit surrogate model, using expensive data (calls to the true function)
and cheap data (the physics based surrogate) in the correction process. See
Final Act. 4; Adv. Act. 3; Ans. 4—5. We also note that Forrester teaches
space-mapping in addition to co-Kriging as a multi-fidelity analysis method,
another physics based surrogate method. Forrester § 3.7.2 (citing Bandler J,
Cheng Q, Dakroury S, Mohamed A, Bakr M, Madsen K, et al. Space
mapping: the state of the art, IEEE Transactions on Microwave Theory and
Techniques 2004: 52: 337—61). Thus, the weight of the evidence supports
the Examiner’s findings that the surrogate in Forrester is a lower level
physics based surrogate and with each interpolation of the surrogate model
(data fit surrogate), the surrogate will be refitted to the data interpolated at
the initial sample points until the surrogate is optimized to a higher level
surrogate for use in generating a model for predicting properties, as claimed.
7
Appeal 2017-008964
Application 13/392,035
Accordingly, Appellants have not persuaded us that the Examiner erred in
finding the combination of Forrester and Wu teaches or suggests the
limitations as recited in claim 1 and similarly required in claims 14 and 17.
Claims 3—13, 19, and 20 depend from one of claims 1,14, and 17 and are
not argued separately. We sustain the Examiner’s rejection of claims 1, 3—
14, 17, 19, and 20 under 35 U.S.C. § 103(a) as unpatentable over Forrester
and Wu.
Regarding the rejection of claims 15 and 16, Appellants present
arguments that Badru and Altamira do not remedy the deficiencies of
Forrester and Wu. App. Br. 10—13. As we find no deficiencies in the
Examiner’s rejection of independent claims 1, 14, and 17 as unpatentable
over Forrester and Wu for the reasons discussed above, we likewise sustain
the Examiner’s rejection of claim 15 under 35 U.S.C. §103(a) as
unpatentable over the combination of Forrester, Wu, and Badru and the
Examiner’s rejection of claim 16 under 35 U.S.C. § 103(a) as unpatentable
over the combination of Forrester, Wu, and Altamira.
CONCFUSION
On the record before us, Appellants have not persuaded us that the
Examiner erred in rejecting claims 1, 3—17, 19, and 20 under § 103(a) as
unpatentable over the combinations of cited references.
DECISION
The Examiner’s rejections of claims 1, 3—17, 19, and 20 are affirmed.
AFFIRMED
8