[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122801-en":3,"doc-seo-122801-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},122801,1649267921044,"Ava Thompson","https://us-avatar.wpscdn.com/avatar/1800007509477c92dfb?_k=1782875107921204101",8,"Research & Report","Learned multiphysics inversion with differentiable programming and machine learning","Learned multiphysics inversion with differentiable programming and machine learning presents the SLIM open-source framework for computational geophysics and broader inverse problems governed by the wave equation, covering seismic imaging and medical ultrasound. The work combines regularization with learned priors and neural surrogates for multiphase-flow simulations through layered abstraction that supports readability, scalability, and exploitation of high-performance computing advances. A scalable permeability-inversion prototype on time-lapse crosswell seismic data demonstrates integration of wave physics and machine-learning approaches.","arXiv :2304 .05592v1 [ cs .MS] 12 Apr 2023  \nLearned multiphysics inversion with differentiable programming  \nand machine learning  \nMathias Louboutin 1,* , Ziyi Yin 1,* , Rafael Orozco 1 , Thomas J. Grady II 1 , Ali Siahkoohi2 , Gabrio Rizzuti3 , Philipp A. Witte4 , Olav Møyner5 , Gerard J. Gorman6 , Felix J. Herrmann 1  \n1 Georgia Institute of Technology, 2 Rice University, 3 University Medical Center Utrecht,  \n4 Microsoft, 5 SINTEF Digital, 6 Imperial College London  \nSummary  \nWe present the Seismic Laboratory for Imaging and Modeling/Monitoring (SLIM) open-source software framework for computational geophysics and, more generally, inverse problems involving the wave-equation (e.g., seismic and medical ultrasound), regularization with learned priors, and learned neural surrogates for multiphase flow simulations. By integrating multiple layers of abstraction, our software is designed to be both readable and scalable. This allows researchers to easily formulate their problems in an abstract fashion while exploiting the latest developments in high-performance computing. We illustrate and demonstrate our design principles and their benefits by means of building a scalable prototype for permeability inversion from time-lapse crosswell seismic data, which aside from coupling of wave physics and multiphase flow, involves machine learning.  \nMotivation  \nThanks to major advancements in high-performance computing (HPC) techniques, computational (exploration) geophysics has made giant leaps over the past decades. These developments have, for instance, led to the adoption of wave-equation-based inversion technologies such as full-waveform inversion (FWI) and reverse-time migration (RTM) that, thanks to their adherence to wave physics, have resulted in superior imaging in complex geologies. While these techniques certainly rank amongst the most sophisticated imaging technologies, their implementation relies with few exceptions—most notably iWave++ (Sun and Symes 2010), Julia Devito Inversion framework (JUDI.jl of the Seismic Laboratory for Imaging and Modeling (SLIM), P. A. Witte, Louboutin, Kukreja, et al. (2019); Mathias Louboutin et al. (2023)), and Chevron’s COFII (Washbourne et al. 2021)—on monolithic low-level (C/Fortran) implementations. As a consequence, due to their lack of abstraction and modern programming constructs, these low-level implementations are difficult and very costly to maintain, especially when performance considerations prevail over best software practices. While these implementation design choices lead to performant code for specific problems, such as FWI, they often hinder the implementation of new algorithms, e.g., based on different objective functions or constraints, as well as coupling existing code bases with external software libraries. For instance, combining wave-equation-based inversion with machine learning frameworks or coupling wave-physics with multiphase fluid-flow solvers are considered challenging and costly. Thus, our industry runs the risk of losing its ability to innovate, a situation that is exacerbated by the challenges we face as a result of the energy transition.  \n*  \nThese authors contributed equally. Corresponding [author: mlouboutin3@gatech.edu](author: mlouboutin3@gatech.edu)  \nDesign principles  \nTo address these important shortcomings of current software implementations that impede progress, we have embarked on the development of a performant software framework. For instance, our wave propagators, implemented in Devito (M. Louboutin et al. 2019; Luporini et al. 2020), are used in production by contractors and Oil & Gas majors while enabling rapid, low-cost, scalable, and interoperable algorithm development for multiphysics and machine learning problems that runs on a variety of different chipsets (e.g. , ARM, Intel, POWER) and graphics accelerators (e.g., NVIDIA) . To achieve this, we adopt contemporary software design practices that include high-level abstractions, software design princi","cbCaibkgWA0FH6vr","https://ap.wps.com/l/cbCaibkgWA0FH6vr","pdf",11058746,1,20,"English","en",105,"# Summary\n# Motivation\n# Design principles","[{\"question\":\"SLIM框架主要用于哪些类型的问题？\",\"answer\":\"SLIM面向由波动方程描述的计算地球物理及更一般的逆问题，包括地震成像与医学超声等场景。\"},{\"question\":\"文档如何将学习能力引入多物理场反演？\",\"answer\":\"通过在逆问题中使用带学习先验的正则化，并引入用于多相流仿真的学习型神经替代模型。\"},{\"question\":\"作者希望解决的现有软件实现痛点是什么？\",\"answer\":\"现有实现往往采用低层级、缺乏抽象与现代编程构造的单体式代码，导致难以维护，并且在更换目标函数/约束或与外部库耦合时成本高、创新受限。\"}]","Learned multiphysics inversion with differentiable programming and machine learning | PDF",1785812973,50,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"learned-multiphysics-inversion-with-differentiable-programming-and-machine-learning","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/learned-multiphysics-inversion-with-differentiable-programming-and-machine-learning/122801/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"SLIM框架主要用于哪些类型的问题？","Question",{"text":75,"@type":76},"SLIM面向由波动方程描述的计算地球物理及更一般的逆问题，包括地震成像与医学超声等场景。","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"文档如何将学习能力引入多物理场反演？",{"text":80,"@type":76},"通过在逆问题中使用带学习先验的正则化，并引入用于多相流仿真的学习型神经替代模型。",{"name":82,"@type":73,"acceptedAnswer":83},"作者希望解决的现有软件实现痛点是什么？",{"text":84,"@type":76},"现有实现往往采用低层级、缺乏抽象与现代编程构造的单体式代码，导致难以维护，并且在更换目标函数/约束或与外部库耦合时成本高、创新受限。","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,114,119,122,126,129,133],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":29,"slug":113},6,"Technology","technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":21,"slug":125},9,"Religion & Spirituality","religion-spirituality",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":127,"show_sort_weight":21,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":46,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":46,"category_name":135,"show_sort_weight":106,"slug":136},19,"General","general"]