An example of the application of neural networks of a simple architecture to unfocused well electrometry probes
An effective method of finding stable solutions of inverse problems of electric and induction logging along the well is proposed, which allows avoiding the influence of the resistance values of the neighboring formations on the determination of the geoelectrical parameters of the object under study....
Збережено в:
| Дата: | 2024 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Kyiv National University of Construction and Architecture
2024
|
| Теми: | |
| Онлайн доступ: | https://es-journal.in.ua/article/view/314101 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Environmental safety and natural resources |
Репозитарії
Environmental safety and natural resources| id |
es-journalinua-article-314101 |
|---|---|
| record_format |
ojs |
| institution |
Environmental safety and natural resources |
| baseUrl_str |
|
| datestamp_date |
2024-10-28T07:19:42Z |
| collection |
OJS |
| language |
English |
| topic |
geophysical exploration of wells resistivity oil and gas wells Shoulder effect inverse problem vertical resolution |
| spellingShingle |
geophysical exploration of wells resistivity oil and gas wells Shoulder effect inverse problem vertical resolution Миронцов, М.Л. An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| topic_facet |
геофізичне досліження свердловин питомий опір нафтогазові свердловини обернена задача вертикальна роздільна здатність geophysical exploration of wells resistivity oil and gas wells Shoulder effect inverse problem vertical resolution |
| format |
Article |
| author |
Миронцов, М.Л. |
| author_facet |
Миронцов, М.Л. |
| author_sort |
Миронцов, М.Л. |
| title |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_short |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_full |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_fullStr |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_full_unstemmed |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_sort |
example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| title_alt |
An example of the application of neural networks of a simple architecture to unfocused well electrometry probes |
| description |
An effective method of finding stable solutions of inverse problems of electric and induction logging along the well is proposed, which allows avoiding the influence of the resistance values of the neighboring formations on the determination of the geoelectrical parameters of the object under study. A highly efficient method was proposed for solving such an unstable inverse problem. This method is based on the application of a neural network with inverse error propagation of a simple architecture. Namely three-layer. The mathematical statement of the problem is given, both the topology of the neural network and all its parameters are described in detail. In the course of the numerical experiment, they were selected as optimal. The process of building a base for training a neural network is described in detail. Namely, how each of the examples of the learning base is built by solving a direct problem. With this cut parameter, the training for each example is chosen arbitrarily, which guarantees a comprehensive range for training the neural network. The number of examples in the training base is one hundred thousand examples. As the activation function, the sigmoid is chosen due to the fact that it is differentiable everywhere. The results of testing the written program are given. The learning rate was estimated to obtain the required small error. It is shown that this approach is stably convergent. For testing, the parameters of the layers of the cut, which are inherent to the geophysical parameters of the cuts of the Dnipro-Donetsk depression, were chosen. A complex of lateral logging sounding was chosen as the electrical logging equipment. Four-probe low-frequency induction logging equipment was chosen as induction logging equipment. Examples for induction and electrical logging are given separately. The obtained results are analyzed in detail. Ways of further improvement of the obtained neural network and its use for other problems of geophysics are given. |
| publisher |
Kyiv National University of Construction and Architecture |
| publishDate |
2024 |
| url |
https://es-journal.in.ua/article/view/314101 |
| work_keys_str_mv |
AT mironcovml anexampleoftheapplicationofneuralnetworksofasimplearchitecturetounfocusedwellelectrometryprobes AT mironcovml exampleoftheapplicationofneuralnetworksofasimplearchitecturetounfocusedwellelectrometryprobes |
| first_indexed |
2024-12-15T20:49:48Z |
| last_indexed |
2024-12-15T20:49:48Z |
| _version_ |
1818749611189207040 |
| spelling |
es-journalinua-article-3141012024-10-28T07:19:42Z An example of the application of neural networks of a simple architecture to unfocused well electrometry probes An example of the application of neural networks of a simple architecture to unfocused well electrometry probes Миронцов, М.Л. геофізичне досліження свердловин питомий опір нафтогазові свердловини обернена задача вертикальна роздільна здатність geophysical exploration of wells resistivity oil and gas wells Shoulder effect inverse problem vertical resolution An effective method of finding stable solutions of inverse problems of electric and induction logging along the well is proposed, which allows avoiding the influence of the resistance values of the neighboring formations on the determination of the geoelectrical parameters of the object under study. A highly efficient method was proposed for solving such an unstable inverse problem. This method is based on the application of a neural network with inverse error propagation of a simple architecture. Namely three-layer. The mathematical statement of the problem is given, both the topology of the neural network and all its parameters are described in detail. In the course of the numerical experiment, they were selected as optimal. The process of building a base for training a neural network is described in detail. Namely, how each of the examples of the learning base is built by solving a direct problem. With this cut parameter, the training for each example is chosen arbitrarily, which guarantees a comprehensive range for training the neural network. The number of examples in the training base is one hundred thousand examples. As the activation function, the sigmoid is chosen due to the fact that it is differentiable everywhere. The results of testing the written program are given. The learning rate was estimated to obtain the required small error. It is shown that this approach is stably convergent. For testing, the parameters of the layers of the cut, which are inherent to the geophysical parameters of the cuts of the Dnipro-Donetsk depression, were chosen. A complex of lateral logging sounding was chosen as the electrical logging equipment. Four-probe low-frequency induction logging equipment was chosen as induction logging equipment. Examples for induction and electrical logging are given separately. The obtained results are analyzed in detail. Ways of further improvement of the obtained neural network and its use for other problems of geophysics are given. Запропоновано ефективний метод знаходження стійких розв'язків обернених задач електричного та індукційного каротажу вздовж свердловини, який дозволяє уникнути впливу значень опору сусідніх пластів на визначення геоелектричних параметрів досліджуваного об'єкта. Для розв’язання такої нестійкої оберненої задачі було запропоновано високоефективний метод. Такий метод заснований на застосуванні нейронної мережі з оберненим розповсюдженням похибки простої архітектури. А саме тришарової. Дано математичну постановку задачі, детально описано як топологію нейронної мережі, так і всі її параметри. В ході числового експерименту вони обрані оптимальними. Детально описано процес побудови бази для навчання нейронної мережі. А саме як за допомогою розв’язання прямої задачі будується кожен з прикладів бази навчання. При цьому параметри розрізу для кожного прикладу навчання обираються довільним чином, що гарантує всеохоплюючий діапазон для навчання нейронної мережі. Кількість прикладів в базі навчання складає сто тисяч прикладів. В якості функції активації обрано сигмоїду через те, що вона всюди диференційована. Наведено результати тестування написаної програми. Оцінена швидкість навчання для отримання необхідної малої похибки. Показано, що такий підхід є стабільно збіжним. Для тестування обрано параметри пластів розрізу, що притаманні геофізичним параметрам розрізів Дніпровсько-Донецької западини. В якості апаратури електричного каротажу обрано комплекс бокового каротажного зондування. В якості апаратури індукційного каротажу обрано чотиризондову апаратуру низькочастотного індукційного каротажу. Наведено окремо приклади для індукційного та електричного каротажу. Детально проаналізовано отримані результати. Наведено шляхи подальшого вдосконалення отриманої нейронної мережі та використання її для інших задач геофізики. Kyiv National University of Construction and Architecture 2024-09-30 Article Article application/pdf https://es-journal.in.ua/article/view/314101 10.32347/2411-4049.2024.3.177-182 Environmental safety and natural resources; Vol. 51 No. 3 (2024): Environmental safety and natural resources; 177-182 Екологічна безпека та природокористування; Том 51 № 3 (2024): Екологічна безпека та природокористування; 177-182 2616-2121 2411-4049 10.32347/2411-4049.2024.3 en https://es-journal.in.ua/article/view/314101/305030 Авторське право (c) 2024 М.Л. Миронцов http://creativecommons.org/licenses/by/4.0 |