{"paper_id":"2b02f7c5-dc6e-4325-a0be-985b9f2ea391","body_text":"Intelligent Diagnosis Model for Rolling Bearing Faults Based on Multi-Modal Data Fusion | 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 Article Intelligent Diagnosis Model for Rolling Bearing Faults Based on Multi-Modal Data Fusion Lin Sun, Jing Shen, Lei Zhao, Ju Zhang, Tongtong Wang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9473421/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract The rolling bearings of high-speed trains are in long-term service under high-speed, variable-load and intense noise conditions, and the collected vibration signals exhibit prominent non-stationarity and non-linearity. This makes it difficult for fault diagnosis methods based on a single feature or single modality to stably characterize fault information. To address the above issues, this paper proposes an intelligent diagnosis model for rolling bearing faults based on multi-modal data fusion, namely MultiFuseAttenNet. First, experimental samples are constructed using the 48 kHz vibration signals from the drive end of the bearing test rig, and locally stationary samples are obtained through data cleaning and a sliding window segmentation strategy (with a window length of 1024 points and an overlap ratio of 50%). Second, starting from the bearing fault mechanism, vibration features, time-domain statistical features and frequency-domain statistical features are extracted, and time-frequency images are generated via the Synchrosqueezing Wavelet Transform (WSST), so as to construct a dual-modality representation of \"numerical features + time-frequency images\". Furthermore, a dual-branch fusion network consisting of a Swin Transformer-based image branch and a SENet-based numerical branch is designed to achieve collaborative learning of local impact information, global time-frequency patterns and statistical feature information. Finally, the performance of the model is comprehensively evaluated with accuracy, precision, recall, F1-score, confusion matrix and t-SNE visualization, and a comparison is performed with mainstream baseline models including 1DCNN, LSTM, GRU, Transformer and VGG16. The experimental results show that the proposed model achieves an accuracy of 99.95% on the test set, and its overall diagnostic performance outperforms the compared baseline models, which verifies the effectiveness of the multi-modal feature fusion strategy for rolling bearing fault identification. Physical sciences/Engineering Physical sciences/Mathematics and computing Rolling bearings Multi-modal fusion Synchrosqueezing Wavelet Transform Swin Transformer Fault diagnosis Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 21 May, 2026 Reviewers agreed at journal 10 May, 2026 Reviewers invited by journal 07 May, 2026 Editor invited by journal 07 May, 2026 Editor assigned by journal 29 Apr, 2026 Submission checks completed at journal 29 Apr, 2026 First submitted to journal 20 Apr, 2026 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-9473421\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":false,\"archivedVersions\":[],\"articleType\":\"Article\",\"associatedPublications\":[],\"authors\":[{\"id\":641477574,\"identity\":\"c0c08de5-b56f-44c9-a1c0-39de1a59f1d4\",\"order_by\":0,\"name\":\"Lin Sun\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8klEQVRIie3PsUoEMRCA4QkDe80saW9B3FcYsRDh0FfJIljt+QoGFuIrZN/CSq6c5cBq0fbgqusF9zotRHNubfZKi/yQwMB8kACkUv8zJQbgNJs1AsouSGs7bQ7kXNOzCeT2pPByBAmnan3NgawXbE18vfRLkd0K1KMYxnz1Sgyihn39N+HNnZGqB2QRg3m/pQu0WLRPETKvWSoHGXdWcOm2dGklwzxCSj8S4rWygbwQi4kT2IxkXjiEQGSacP/2S1hTBvjlbqjwXRP9S/lQn+0+Hdy78n1Qrbu61rrphn3sYWPfh2v2MQ7KTu6nUqlUKt4P6+RUdE8PtHkAAAAASUVORK5CYII=\",\"orcid\":\"\",\"institution\":\"Wenhua College\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Lin\",\"middleName\":\"\",\"lastName\":\"Sun\",\"suffix\":\"\"},{\"id\":641477575,\"identity\":\"1e1b491e-2e8c-48a7-98cb-4a458f0557e1\",\"order_by\":1,\"name\":\"Jing Shen\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Wenhua College\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Jing\",\"middleName\":\"\",\"lastName\":\"Shen\",\"suffix\":\"\"},{\"id\":641477576,\"identity\":\"570a993a-bee0-4432-baa8-9ac41a5298a4\",\"order_by\":2,\"name\":\"Lei Zhao\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Wenhua College\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Lei\",\"middleName\":\"\",\"lastName\":\"Zhao\",\"suffix\":\"\"},{\"id\":641477577,\"identity\":\"3aa94749-529a-46fb-bc6d-3b6225d68302\",\"order_by\":3,\"name\":\"Ju Zhang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Wenhua College\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ju\",\"middleName\":\"\",\"lastName\":\"Zhang\",\"suffix\":\"\"},{\"id\":641477578,\"identity\":\"a4b75876-1aa4-48ee-8b4d-4dd93ece6570\",\"order_by\":4,\"name\":\"Tongtong Wang\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Wenhua College\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Tongtong\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2026-04-20 14:11:16\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-9473421/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-9473421/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":109462471,\"identity\":\"2f741ebc-bfc7-4688-a263-a98d7aa52782\",\"added_by\":\"auto\",\"created_at\":\"2026-05-18 11:11:06\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":1382204,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"newmanuscript01.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-9473421/v1_covered_83990a3e-4deb-4042-8765-006a93acf021.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Intelligent Diagnosis Model for Rolling Bearing Faults Based on Multi-Modal Data Fusion\",\"fulltext\":[],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":false,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":false,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":true,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":true,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"scientific-reports\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"scirep\",\"sideBox\":\"Learn more about [Scientific Reports](http://www.nature.com/srep/)\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Scientific Reports\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"stoa\",\"reportingPortfolio\":\"Scientific Reports\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Rolling bearings, Multi-modal fusion, Synchrosqueezing Wavelet Transform, Swin Transformer, Fault diagnosis\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-9473421/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-9473421/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eThe rolling bearings of high-speed trains are in long-term service under high-speed, variable-load and intense noise conditions, and the collected vibration signals exhibit prominent non-stationarity and non-linearity. This makes it difficult for fault diagnosis methods based on a single feature or single modality to stably characterize fault information. To address the above issues, this paper proposes an intelligent diagnosis model for rolling bearing faults based on multi-modal data fusion, namely MultiFuseAttenNet. First, experimental samples are constructed using the 48 kHz vibration signals from the drive end of the bearing test rig, and locally stationary samples are obtained through data cleaning and a sliding window segmentation strategy (with a window length of 1024 points and an overlap ratio of 50%). Second, starting from the bearing fault mechanism, vibration features, time-domain statistical features and frequency-domain statistical features are extracted, and time-frequency images are generated via the Synchrosqueezing Wavelet Transform (WSST), so as to construct a dual-modality representation of \\\"numerical features\\u0026thinsp;+\\u0026thinsp;time-frequency images\\\". Furthermore, a dual-branch fusion network consisting of a Swin Transformer-based image branch and a SENet-based numerical branch is designed to achieve collaborative learning of local impact information, global time-frequency patterns and statistical feature information. Finally, the performance of the model is comprehensively evaluated with accuracy, precision, recall, F1-score, confusion matrix and t-SNE visualization, and a comparison is performed with mainstream baseline models including 1DCNN, LSTM, GRU, Transformer and VGG16. The experimental results show that the proposed model achieves an accuracy of 99.95% on the test set, and its overall diagnostic performance outperforms the compared baseline models, which verifies the effectiveness of the multi-modal feature fusion strategy for rolling bearing fault identification.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Intelligent Diagnosis Model for Rolling Bearing Faults Based on Multi-Modal Data Fusion\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2026-05-18 11:09:18\",\"doi\":\"10.21203/rs.3.rs-9473421/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0},{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2026-05-21T08:48:55+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"335206794603309428690702290501840528118\",\"date\":\"2026-05-10T08:22:05+00:00\",\"index\":\"hide\",\"fulltext\":\"\"},{\"type\":\"reviewersInvited\",\"content\":\"\",\"date\":\"2026-05-07T09:10:04+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorInvited\",\"content\":\"\",\"date\":\"2026-05-07T08:35:43+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorAssigned\",\"content\":\"\",\"date\":\"2026-04-29T07:42:32+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"checksComplete\",\"content\":\"\",\"date\":\"2026-04-29T07:42:20+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"submitted\",\"content\":\"Scientific Reports\",\"date\":\"2026-04-20T14:04:24+00:00\",\"index\":\"\",\"fulltext\":\"\"}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"scientific-reports\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":false,\"externalIdentity\":\"scirep\",\"sideBox\":\"Learn more about [Scientific Reports](http://www.nature.com/srep/)\",\"snPcode\":\"\",\"submissionUrl\":\"\",\"title\":\"Scientific Reports\",\"twitterHandle\":\"\",\"acdcEnabled\":true,\"dfaEnabled\":true,\"editorialSystem\":\"stoa\",\"reportingPortfolio\":\"Scientific Reports\",\"inReviewEnabled\":true,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"a4cddcd0-a61f-4969-bec7-c706739ba486\",\"owner\":[],\"postedDate\":\"May 18th, 2026\",\"published\":true,\"recentEditorialEvents\":[{\"type\":\"editorInvitedReview\",\"content\":\"\",\"date\":\"2026-05-21T08:48:55+00:00\",\"index\":21,\"fulltext\":\"\"},{\"type\":\"reviewerAgreed\",\"content\":\"335206794603309428690702290501840528118\",\"date\":\"2026-05-10T08:22:05+00:00\",\"index\":20,\"fulltext\":\"\"},{\"type\":\"reviewersInvited\",\"content\":\"2\",\"date\":\"2026-05-07T09:10:04+00:00\",\"index\":\"\",\"fulltext\":\"\"},{\"type\":\"editorInvited\",\"content\":\"\",\"date\":\"2026-05-07T08:35:43+00:00\",\"index\":\"\",\"fulltext\":\"\"}],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"under-review\",\"subjectAreas\":[{\"id\":68235763,\"name\":\"Physical sciences/Engineering\"},{\"id\":68235764,\"name\":\"Physical sciences/Mathematics and computing\"}],\"tags\":[],\"updatedAt\":\"2026-05-18T11:09:18+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2026-05-18 11:09:18\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-9473421\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-9473421\",\"identity\":\"rs-9473421\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}