Actor-critic based on Attention Model for Multi-robotCollaborative Backend Optimization

preprint OA: closed CC-BY-4.0
📄 Open PDF Full text JSON View at publisher
AI-generated deep summary by claude@2026-07, 2026-07-03 · read from full text

The paper studies collaborative backend optimization for multi-robot simultaneous localization and mapping (SLAM) in heterogeneous multi-robot systems, proposing a deep reinforcement learning framework with attention mechanisms (MAS-AA) to improve how map points and pose nodes are selected under constraints informed by robot states and attributes. Using a collaborative attention neural network plus a reinforcement-learning decision network, the authors define an “optimal collaborative chain,” derive a multi-robot bundle adjustment algorithm from it, and construct a cost and reward function tied to collaborative attention. They report a learning methodology that combines weight updates from both neural networks, and simulation experiments show improvements in localization accuracy and mapping performance. A key stated caveat is that the work targets limitations of attention models failing to perceive subsequent collaboration states; the study is presented as a Research Square preprint and not peer reviewed. The paper does not explicitly discuss endometriosis or adenomyosis; it was included in the corpus via a keyword match in the upstream search index.

Read from the paper's body, not the abstract. Not a substitute for reading the paper. No clinical advice. How this works

Abstract

Abstract Backend optimization is an essential component of simultaneous localization and mapping (SLAM).Collaborative backend optimization in multi-robot systems refers to the process of extending single-robot collaboration optimization to coordinate and optimize the backend processes of multiple robots working together, enhancing overall system performance and efficiency.In this paper, a deep reinforcement learning model based on attention mechanisms called MAS-AA specifically tailored for collaborative backend optimization in heterogeneous multi-robot systems is proposed, solved the problem of heterogeneous multi robot system collaborative backend optimization not being able to optimize the selection of map points and pose nodes based on constraints between map points and pose nodes considering robot states and attributes. Firstly, we introduce a collaborative attention neural network designed for multi-robot back-end optimization, along with a collaborative decision-making neural network based on deep reinforcement learning. Secondly, we delve into an optimization mechanism based on the optimal collaborative chain, as well as a multi-robot bundle adjustment algorithm derived from this mechanism. Lastly, we design and implement a cost function for the decision-making model based on collaborative attention, as well as a reward function for the collaborative model. We further present a learning methodology that combines the weight update processes of both neural networks.Simulation experiments validate the significant enhancements achieved by our algorithm in terms of localization accuracy and mapping performance in multi-robot collaborative backend optimization. Effectively addressing the limitation of improvement in collaboration performance caused by the inability to perceive subsequent collaboration states in the application of attention models in multi-robot collaborative backend optimization.
Full text 25,516 characters · extracted from preprint-html · click to expand
Actor-critic based on Attention Model for Multi-robotCollaborative Backend Optimization | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Actor-critic based on Attention Model for Multi-robotCollaborative Backend Optimization Zhaoyi Pei, Zhenghong Yu, Dunlu Lu, Qingqing Bian, Haijie Feng, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7797205/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Backend optimization is an essential component of simultaneous localization and mapping (SLAM).Collaborative backend optimization in multi-robot systems refers to the process of extending single-robot collaboration optimization to coordinate and optimize the backend processes of multiple robots working together, enhancing overall system performance and efficiency.In this paper, a deep reinforcement learning model based on attention mechanisms called MAS-AA specifically tailored for collaborative backend optimization in heterogeneous multi-robot systems is proposed, solved the problem of heterogeneous multi robot system collaborative backend optimization not being able to optimize the selection of map points and pose nodes based on constraints between map points and pose nodes considering robot states and attributes. Firstly, we introduce a collaborative attention neural network designed for multi-robot back-end optimization, along with a collaborative decision-making neural network based on deep reinforcement learning. Secondly, we delve into an optimization mechanism based on the optimal collaborative chain, as well as a multi-robot bundle adjustment algorithm derived from this mechanism. Lastly, we design and implement a cost function for the decision-making model based on collaborative attention, as well as a reward function for the collaborative model. We further present a learning methodology that combines the weight update processes of both neural networks.Simulation experiments validate the significant enhancements achieved by our algorithm in terms of localization accuracy and mapping performance in multi-robot collaborative backend optimization. Effectively addressing the limitation of improvement in collaboration performance caused by the inability to perceive subsequent collaboration states in the application of attention models in multi-robot collaborative backend optimization. Backend optimization SLAM Multi-robot system Deep reinforcement learning Attention mechanism Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7797205","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":535022122,"identity":"735bb3e9-a32a-41c3-85db-18b3accf5a31","order_by":0,"name":"Zhaoyi Pei","email":"","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhaoyi","middleName":"","lastName":"Pei","suffix":""},{"id":535022124,"identity":"d6ab34eb-fe13-43ab-a685-af27ea7ab081","order_by":1,"name":"Zhenghong Yu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2klEQVRIiWNgGAWjYBACAwglIcfP33wAxJAhVouFseSMYwkgLTzEaqlI3HAgB8wmrMWcvffwi497JBJnNpz5/OpGjQUPA/vhoxvwabHsOZdmOeOZhHE/c+8265xjQIfxpKXdwOuwGzlmxjwHJGRnNpzdZpzDBtQiwWNGWMufAxKMQL88M875R5wW48cMByQUgVqYH+e2EaPlzBkzxp4DEqBANmPO7ZPgYSPol+M9xh9+HKgDReXjzznfgAz2w8fwagECNgkUBhsB5SDA/AGdMQpGwSgYBaMABQAA4xRMjAnTOU0AAAAASUVORK5CYII=","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Zhenghong","middleName":"","lastName":"Yu","suffix":""},{"id":535022126,"identity":"48f75601-48b4-466a-ac0f-5ffebf42a06b","order_by":2,"name":"Dunlu Lu","email":"","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Dunlu","middleName":"","lastName":"Lu","suffix":""},{"id":535022134,"identity":"36ebddc4-5080-421c-94ce-780e504a27f3","order_by":3,"name":"Qingqing Bian","email":"","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Qingqing","middleName":"","lastName":"Bian","suffix":""},{"id":535022135,"identity":"8733de46-0d5e-46ec-90db-3d867028197b","order_by":4,"name":"Haijie Feng","email":"","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Haijie","middleName":"","lastName":"Feng","suffix":""},{"id":535022143,"identity":"1bf0cd3e-c401-47fb-b586-9e803683e066","order_by":5,"name":"Hongyuan Li","email":"","orcid":"","institution":"Guangdong Polytechnic of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Hongyuan","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2025-10-07 08:08:27","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7797205/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7797205/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94719112,"identity":"022f0e32-bf7d-4c73-9d93-2e9bbb7abf4f","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"json","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8712,"visible":true,"origin":"","legend":"","description":"","filename":"e11e84298d4e418e819205eb0e27b1db.json","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/1fa939be7151f38895bc2ecf.json"},{"id":94719116,"identity":"c7c044e8-ed76-44c0-80d0-9e5fa611d46e","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"jpg","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6408964,"visible":true,"origin":"","legend":"","description":"","filename":"attentionmodel.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/e270ef3692ca9a4c46f35efe.jpg"},{"id":94719124,"identity":"df02d60f-67d3-4478-ab87-97d3d3060e3e","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10926995,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/d3bae55bfd35934f72db5e24.pdf"},{"id":94719111,"identity":"e9d62192-d64c-45a6-8f71-a0ad778ff443","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39805,"visible":true,"origin":"","legend":"","description":"","filename":"masmapamount.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/b9c04003faa1387ffe4acdc2.png"},{"id":94729460,"identity":"12af5d0f-7f90-4e72-8aa5-00a588251e56","added_by":"auto","created_at":"2025-10-30 07:04:59","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53088,"visible":true,"origin":"","legend":"","description":"","filename":"masmaptran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/e3f134ed2d3d660ed31c2297.png"},{"id":94719114,"identity":"b12f6dd9-5d44-4b1b-a151-ea6ba7034db5","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":65666,"visible":true,"origin":"","legend":"","description":"","filename":"masrobotrota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/e275033631d6144e14c7475f.png"},{"id":94729196,"identity":"00930091-6021-49c4-a7db-05e60e698dff","added_by":"auto","created_at":"2025-10-30 07:04:41","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51329,"visible":true,"origin":"","legend":"","description":"","filename":"masrobottran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/306c9354e9c8d577848dbb53.png"},{"id":94729756,"identity":"0a85ac0b-6e8b-4469-b108-b4038a2d935a","added_by":"auto","created_at":"2025-10-30 07:05:22","extension":"jpg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2319824,"visible":true,"origin":"","legend":"","description":"","filename":"modelrun.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/6fa18b460159f15733c47b5f.jpg"},{"id":94730152,"identity":"ffa32add-88f8-4fca-b1d4-00e274c8f3ec","added_by":"auto","created_at":"2025-10-30 07:05:42","extension":"jpg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":807977,"visible":true,"origin":"","legend":"","description":"","filename":"realenvironment.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/a920aafbad38ccd2deab0a83.jpg"},{"id":94719126,"identity":"08c5b879-4609-4a75-a56f-67d0be899f10","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42303,"visible":true,"origin":"","legend":"","description":"","filename":"realmappoint.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/b25bd9488afba6b01f9c2cd0.png"},{"id":94719119,"identity":"68e0b877-83d0-4b52-8089-a49ec95a7293","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":47870,"visible":true,"origin":"","legend":"","description":"","filename":"realrota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/b3e248f5b025c4697b475c0d.png"},{"id":94719117,"identity":"511254eb-b1e4-46a6-8e6c-3ad72be85fe6","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39848,"visible":true,"origin":"","legend":"","description":"","filename":"realtran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/c1d9cce4633b0d7ca4de4cc8.png"},{"id":94719125,"identity":"78ff8938-b7cf-464a-b8b6-5c512cc6376f","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":50652,"visible":true,"origin":"","legend":"","description":"","filename":"robot1map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/d2f58f642cb229e448b8b0ff.png"},{"id":94719140,"identity":"783290e7-6683-4832-8126-0589d27ce5b3","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58970,"visible":true,"origin":"","legend":"","description":"","filename":"robot1rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/5e4f5af0e2d6171d43018311.png"},{"id":94729574,"identity":"06dff33d-a0ab-4601-aee7-b55f235fb9a0","added_by":"auto","created_at":"2025-10-30 07:05:10","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":38449,"visible":true,"origin":"","legend":"","description":"","filename":"robot1tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/537873c5e3d2ece902b1dbd7.png"},{"id":94719130,"identity":"3ad5446b-c49f-45aa-b6ed-0b23c2ec27ca","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53787,"visible":true,"origin":"","legend":"","description":"","filename":"robot2map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/3121225b933097c91628e7f6.png"},{"id":94719123,"identity":"37b40058-dcc8-4ce7-9293-818ddc67b5e5","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54747,"visible":true,"origin":"","legend":"","description":"","filename":"robot2rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/be466736788ac21f11cfa01f.png"},{"id":94729399,"identity":"b14dfdb2-b96a-4bc1-9898-3616b619d672","added_by":"auto","created_at":"2025-10-30 07:04:54","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":45633,"visible":true,"origin":"","legend":"","description":"","filename":"robot2tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/85a6d609c68cfe34a09911d7.png"},{"id":94719142,"identity":"908a0b0e-5a8c-4ce2-9015-ed3938d3ea8a","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":58590,"visible":true,"origin":"","legend":"","description":"","filename":"robot3map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/72b0010932e4f2cd941b0333.png"},{"id":94719135,"identity":"b6f971a3-8b02-4997-942e-03fe2f569475","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":60184,"visible":true,"origin":"","legend":"","description":"","filename":"robot3rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/2f1a2fecc21cfaeac0116d48.png"},{"id":94719132,"identity":"1ea00728-b06c-493f-a9ed-4d08105413da","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40423,"visible":true,"origin":"","legend":"","description":"","filename":"robot3tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/14b45038a0ff56966661bc46.png"},{"id":94719137,"identity":"491fb51f-c3db-4dbb-a247-e5f88c39dbd5","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55137,"visible":true,"origin":"","legend":"","description":"","filename":"robot4map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/0b1434535744f19edde217b8.png"},{"id":94729218,"identity":"1fb0ec05-7fca-44ba-a3c4-1fe20bafad71","added_by":"auto","created_at":"2025-10-30 07:04:42","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62591,"visible":true,"origin":"","legend":"","description":"","filename":"robot4rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/f77a7a350e119db84dc4eb1e.png"},{"id":94729963,"identity":"57e9ed0f-5d03-4464-9352-5be0713233c0","added_by":"auto","created_at":"2025-10-30 07:05:32","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":39071,"visible":true,"origin":"","legend":"","description":"","filename":"robot4tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/ff1b4124ed1c925c81e48140.png"},{"id":94719118,"identity":"b999112e-47c5-4783-a2a0-427091c793ac","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53269,"visible":true,"origin":"","legend":"","description":"","filename":"robot5map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/9dcdc7cc3765bea23824ccf0.png"},{"id":94719127,"identity":"a8b4a570-ebe8-4d23-be99-355280ba22d0","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56566,"visible":true,"origin":"","legend":"","description":"","filename":"robot5rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/2588cdded324e18c91841627.png"},{"id":94719138,"identity":"ab93ac42-add7-4ba6-b45a-592240062f71","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":41580,"visible":true,"origin":"","legend":"","description":"","filename":"robot5tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/c2209230eedfb031b96f787c.png"},{"id":94719136,"identity":"55b5760d-bced-42c2-83c2-490665494a5a","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51072,"visible":true,"origin":"","legend":"","description":"","filename":"robot6map.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/734e7ba26ac8eb13649f30b1.png"},{"id":94719131,"identity":"68949926-96d7-4c1b-8cef-814a1e4b52ba","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62283,"visible":true,"origin":"","legend":"","description":"","filename":"robot6rota.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/4083f4ecff217bcad843ae6d.png"},{"id":94719139,"identity":"975accd6-eec3-40bd-80a1-2b74035897c1","added_by":"auto","created_at":"2025-10-30 04:50:25","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40589,"visible":true,"origin":"","legend":"","description":"","filename":"robot6tran.png","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/dbceaab7549c76d1f2b5344c.png"},{"id":94719129,"identity":"af81febf-a830-4611-95a5-ab85cd16fb69","added_by":"auto","created_at":"2025-10-30 04:50:24","extension":"jpg","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2010913,"visible":true,"origin":"","legend":"","description":"","filename":"trainsample.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/f3577b5ebf989e78f60ba66f.jpg"},{"id":94730060,"identity":"cd30adab-6eb0-4ae8-8fe0-fac4fe1e977c","added_by":"auto","created_at":"2025-10-30 07:05:37","extension":"xml","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106676,"visible":true,"origin":"","legend":"","description":"","filename":"e11e84298d4e418e819205eb0e27b1db1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1/989025eaff69533c5d3ef94e.xml"},{"id":101610210,"identity":"5741dfe7-5cd1-419a-bf6a-f8043b317bd9","added_by":"auto","created_at":"2026-02-01 17:10:07","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10393782,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7797205/v1_covered_26b02037-8c90-47e6-b405-9d2640b35bf5.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Actor-critic based on Attention Model for Multi-robotCollaborative Backend Optimization","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Backend optimization, SLAM,Multi-robot system, Deep reinforcement learning, Attention mechanism","lastPublishedDoi":"10.21203/rs.3.rs-7797205/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7797205/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Backend optimization is an essential component of simultaneous localization and mapping (SLAM).Collaborative backend optimization in multi-robot systems refers to the process of extending single-robot collaboration optimization to coordinate and optimize the backend processes of multiple robots working together, enhancing overall system performance and efficiency.In this paper, a deep reinforcement learning model based on attention mechanisms called MAS-AA specifically tailored for collaborative backend optimization in heterogeneous multi-robot systems is proposed, solved the problem of heterogeneous multi robot system collaborative backend optimization not being able to optimize the selection of map points and pose nodes based on constraints between map points and pose nodes considering robot states and attributes. Firstly, we introduce a collaborative attention neural network designed for multi-robot back-end optimization, along with a collaborative decision-making neural network based on deep reinforcement learning. Secondly, we delve into an optimization mechanism based on the optimal collaborative chain, as well as a multi-robot bundle adjustment algorithm derived from this mechanism. Lastly, we design and implement a cost function for the decision-making model based on collaborative attention, as well as a reward function for the collaborative model. We further present a learning methodology that combines the weight update processes of both neural networks.Simulation experiments validate the significant enhancements achieved by our algorithm in terms of localization accuracy and mapping performance in multi-robot collaborative backend optimization. Effectively addressing the limitation of improvement in collaboration performance caused by the inability to perceive subsequent collaboration states in the application of attention models in multi-robot collaborative backend optimization.","manuscriptTitle":"Actor-critic based on Attention Model for Multi-robotCollaborative Backend Optimization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-30 04:50:19","doi":"10.21203/rs.3.rs-7797205/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a04277cd-6f13-47ee-a466-0f0dc3d68099","owner":[],"postedDate":"October 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-01T17:09:43+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-30 04:50:19","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7797205","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7797205","identity":"rs-7797205","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}

Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.

My notes (saved in your browser only)

Ask this paper AI returns verbatim quotes from the full text · source: preprint-html

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2025) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

Source provenance

europepmc
last seen: 2026-05-20T01:45:00.602351+00:00
unpaywall
last seen: 2026-05-29T02:00:03.542394+00:00
License: CC-BY-4.0