US6477469B2
Expired / Lapsed
US
⏱️ 0.0 years remaining
Coarse-to-fine self-organizing map for automatic electrofacies ordering
Assignee: Elf Exploration Production SAS
•
Filing: 2001-01-08
•
Granted: 2002-11-05
•
Expires: 2021-01-08
Abstract
Abstract A method for ordering electrofacies to assist in identification of mineral deposits is disclosed. Automated ordering of electrofacies allows geologists to draw inferences about the geological settings in which sediment deposit occurred without directly examining core samples or outcrops. The electrofacies order is determined by (a) training a one-dimensional linear self-organizing map to form an initial neural network that includes a plurality of neurons. The number of neurons is small in comparison to the number of electrofacies kernels (i.e., not greater than one-third the number of electrofacies kernels). (b1) A neuron is selected from the initial neural network. In the next step (b2), the processor determines if more than one electrofacies kernel is attached to the neuron. (b3) If more than one electrofacies kernel is attached to the neuron, then the neuron is split into the number of electrofacies kernels attached to the neuron. (c) Steps (b1)-(b3) are repeated until all neurons in the initial neural network have been processed. In the next step, (d) a self-organizing map is trained to form a final neural network using the split neurons in the initial neural network as initial state. (e) Steps (b1)-(d) are repeated if more than one electrofacies kernel is attached to a neuron with the initial neural network equal to the final neural network. In the last step (f), each electrofacies kernel corresponding to a neuron in the final neural network is correlated to an order number.
Inventors
- 👤 Shin-Ju Ye
- 👤 Philippe J. Y. M. Rabiller
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-10-26 19:47
CPC Classifications
G; G01; G01V; G01V3/00; G01V3/38
Claims Analysis
Total Claims: 22
Claim 1
Independent Claim
115 words
A method of ordering a set of kernels in a multi-dimensional data space, wherein the method includes: placing an ordered set of neurons in initial positions in the multi-dimensional data space; training the ordered set of neurons on the set of kernels; determining a number of kernels attached to each neuron in the ordered set; augmenting the ordered set of neurons by replacing each neuron with an ordered replacement set of neurons, wherein the replacement set includes a number of neurons that equals a number of kernels attached to the neuron being replaced; repeating said training, determining, and augmenting with the augmented set of neurons until each neuron has no more than one kernel attached.
Claim 2
Independent Claim
115 words
A method of ordering a set of kernels in a multi-dimensional data space, wherein the method includes: placing an ordered set of neurons in initial positions in the multi-dimensional data space; training the ordered set of neurons on the set of kernels; determining a number of kernels attached to each neuron in the ordered set; augmenting the ordered set of neurons by replacing each neuron with an ordered replacement set of neurons, wherein the replacement set includes a number of neurons that equals a number of kernels attached to the neuron being replaced; repeating said training, determining, and augmenting with the augmented set of neurons until each neuron has no more than one kernel attached.
Claim 3
Dependent Claim
13 words
The method of claim 1 , wherein the initial positions are randomly determined.
Claim 4
Dependent Claim
40 words
The method of claim 1 , wherein the initial ordered set of neurons consists of an initial number of neurons, and wherein the initial number is no more than one third the number of kernels in the set of kernels.
Claim 5
Dependent Claim
31 words
The method of claim 1 , wherein the multi-dimensional space has a number of dimensions that is determined by a number of logs, and wherein the number of logs exceeds two.
Claim 6
Dependent Claim
13 words
The method of claim 4 , wherein the logs include RhoB and Nphi.
Claim 7
Dependent Claim
74 words
The method of claim 1 , wherein the training comprises: reordering the set of kernels into a sequence; and for each kernel in the sequence: selecting a nearest neuron; adapting the position of the nearest neuron to reduce a distance between the kernel and the nearest neuron; adapting the positions of neurons adjacent to the nearest neuron in the ordered set of neurons to reduce a distance between the kernel and the adjacent neurons.
Claim 8
Dependent Claim
16 words
The method of claim 6 , wherein the acts of claim 6 are repeated multiple times.
Claim 9
Dependent Claim
33 words
The method of claim 1 , wherein the determining includes: determining for each kernel a nearest neuron; determining for each neuron a number of kernels for which the neuron is the nearest neuron.
Claim 10
Dependent Claim
29 words
The method of claim 1 , wherein the augmenting includes: determining initial positions of the replacement neurons by adding random perturbations to the position of the neuron being replaced.
Claim 11
Dependent Claim
32 words
The method of claim 1 , wherein the augmenting includes: ordering the augmented set by ordering the replacement sets in the order of the neurons they replaced to create an ordered set.
Claim 12
Independent Claim
109 words
A method of ordering electrofacies of a geological formation traversed by a borehole comprising: (a) training a one-dimensional linear self-organizing map to form a initial neural network that includes a plurality of neurons, said number of neurons small in comparison to the number of electrofacies kernels; (b1) selecting a neuron from the initial neural network; (b2) determining if more than one electrofacies kernel is attached to the neuron; (b3) splitting the neuron into a number of electrofacies kernels attached to the neuron if more than one electrofacies kernel is attached to the neuron; and (c) repeating steps (b1)-(b3) until all neurons in the initial neural network have been processed.
Claim 13
Independent Claim
109 words
A method of ordering electrofacies of a geological formation traversed by a borehole comprising: (a) training a one-dimensional linear self-organizing map to form a initial neural network that includes a plurality of neurons, said number of neurons small in comparison to the number of electrofacies kernels; (b1) selecting a neuron from the initial neural network; (b2) determining if more than one electrofacies kernel is attached to the neuron; (b3) splitting the neuron into a number of electrofacies kernels attached to the neuron if more than one electrofacies kernel is attached to the neuron; and (c) repeating steps (b1)-(b3) until all neurons in the initial neural network have been processed.
Claim 14
Dependent Claim
20 words
The method of claim 11 wherein said number of neurons is not greater than one-third the number of electrofacies kernels.
Claim 15
Dependent Claim
97 words
The method of claim 11 wherein said training a one-dimensional linear self-organizing map further includes: initializing the neurons so that the neurons are interconnected in a one-dimensional array; inputting electrofacies kernels for a given number of cycles to the initial neural network; calculating a winner neuron for each electrofacies kernel presented to the initial neural network, said winner neuron being the nearest neuron to the electrofacies kernel; and moving the winner neuron towards the electrofacies kernel, said winner neuron inducing with decreasing intensity its neighbor neurons in the initial neural network to move towards the electrofacies kernel.
Claim 16
Dependent Claim
76 words
The method of claim 11 , further comprising: (d) training a self-organizing map to form a final neural network using the split neurons in the initial neural network as initial state; (e) repeating steps (b1)-(d) if more than one electrofacies kernel is attached to a neuron with the initial neural network equal to the final neural network; and (f) correlating each electrofacies kernel corresponding to a neuron in the final neural network to an order number.
Claim 17
Independent Claim
51 words
An apparatus for performing electrofacies ordering, comprising: a memory unit configured to store log measurement points in multidimensional space; and a processing unit configured to retrieve the log measurement points from the memory unit, wherein said processing unit is configured to calculate an ordering of the electrofacies of a geological formation.
Claim 18
Independent Claim
51 words
An apparatus for performing electrofacies ordering, comprising: a memory unit configured to store log measurement points in multidimensional space; and a processing unit configured to retrieve the log measurement points from the memory unit, wherein said processing unit is configured to calculate an ordering of the electrofacies of a geological formation.
Claim 19
Dependent Claim
98 words
The apparatus of claim 15 , wherein ordering of the electrofacies of a geological formation is determined by: (a) training a one-dimensional linear self-organizing map to form an initial neural network that includes a plurality of neurons, said number of neurons small in comparison to the number of electrofacies kernels; (b1) selecting a neuron from the initial neural network; (b2) determining if more than one electrofacies kernel is attached to the neuron; and (b3) splitting the neuron into a number of electrofacies kernels attached to the neuron if more than one electrofacies kernel is attached to the neuron.
Claim 20
Dependent Claim
20 words
The apparatus of claim 16 , wherein the number of neurons is less than one-third the number of electrofacies kernels.
Claim 21
Dependent Claim
98 words
The apparatus of claim 16 , wherein said training a one-dimensional linear self-organizing map further includes: initializing the neurons so that the neurons are interconnected in a one-dimensional array; inputting electrofacies kernels for a given number of cycles to the initial neural network; calculating a winner neuron for each electrofacies kernel presented to the initial neural network, said winner neuron being the nearest neuron to the electrofacies kernel; and moving the winner neuron towards the electrofacies kernel, said winner neuron inducing with decreasing intensity its neighbor neurons in the initial neural network to move towards the electrofacies kernel.
Claim 22
Dependent Claim
102 words
The apparatus of claim 16 , wherein determining the order of the electrofacies of a geological formation further comprises: (c) repeating steps (b1)-(b3) until all neurons in the initial neural network have been processed; (d) training a self-organizing map to form a final neural network using the split neurons in the initial neural network as initial state; (e) repeating steps (b1)-(d) if more than one electrofacies kernel is attached to a neuron with the initial neural network equal to the final neural network; and (f) correlating each electrofacies kernel corresponding to a neuron in the final neural network to an order number.
References Cited (Backward Citations)
US4646240A
Cited
GB2215891A
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2005-11-04 | 2006-05-06 | 2006-11-05 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2009-11-04 | 2010-05-06 | 2010-11-05 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2013-11-04 | 2014-05-06 | 2014-11-05 | $7700.00 | Window Closed / Lapsed |