Wiener-NBEATS Hybrid Denoising and Deep Learning Framework for Multi-Horizon Hydroelectric Time Series Forecasting

Wiener-NBEATS Hybrid Denoising and Deep Learning Framework for Multi-Horizon Hydroelectric Time Series Forecasting | 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 Wiener-NBEATS Hybrid Denoising and Deep Learning Framework for Multi-Horizon Hydroelectric Time Series Forecasting Rafael Ninno Muniz, William Gouvêa Buratto, Gabriel Villarrubia Gonzalez, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9684105/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 Accurate forecasting of hydroelectric time series is challenging due to noise and nonlinear temporal dynamics. This paper proposes a hybrid framework combining Wiener filtering and the Neural Basis Expansion Analysis (NBEATS) model for short-term and multi-horizon turbine flow prediction. The Wiener filter is applied as a preprocessing step to optimally denoise the signal while preserving its intrinsic structure, enabling more effective learning. The proposed Wiener-NBEATS approach is evaluated using real data from the Colíder hydroelectric power plant and compared with state-of-the-art models. Results show that preprocessing plays a dominant role in performance. The proposed method achieves MAE values as low as 0.199 and RMSE of 0.239, outperforming transformer-based and recurrent models. Across different horizons, the method maintains strong performance, with MAE around 5.08 for longer horizons. These results demonstrate that the Wiener-NBEATS framework significantly improves forecasting accuracy while maintaining low computational cost, highlighting its effectiveness and applicability for hydroelectric system management. NBEATS prediction model hydroelectric power plants time series forecasting Wiener filter. Full Text Additional Declarations The authors declare no competing interests. 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-9684105","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":638549618,"identity":"e51bb7d1-2797-4575-bd75-e657ecff5ea2","order_by":0,"name":"Rafael Ninno Muniz","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAApUlEQVRIiWNgGAWjYJCCAwwVEiRrOUOqFgbGNlJU687IPXjw5zwLOX72BsbHFb+I0GJ2Iy/hgOQ2CWPJngPMhmf7iNKSY3DAcJtE4oYbCWySjT3EakmcI1G///4DUrQcbJBIMJBgYJNs+EGMljNvDA42HJMwnHEmsdmwsYEYLcdzjD/+qKmT528/fPBhwx8itCABxgYSIwgCSLRlFIyCUTAKRgYAAAIiOTy0Ib34AAAAAElFTkSuQmCC","orcid":"","institution":"Insight Energy","correspondingAuthor":true,"prefix":"","firstName":"Rafael","middleName":"Ninno","lastName":"Muniz","suffix":""},{"id":638549952,"identity":"6a384fb7-2d2d-4e09-a8f1-66660b8ed8d8","order_by":1,"name":"William Gouvêa Buratto","email":"","orcid":"","institution":"Universidade do Estado de Santa Catarina","correspondingAuthor":false,"prefix":"","firstName":"William","middleName":"Gouvêa","lastName":"Buratto","suffix":""},{"id":638549953,"identity":"7d669ed6-a8f3-46a0-b46c-d746f6e0132d","order_by":2,"name":"Gabriel Villarrubia Gonzalez","email":"","orcid":"","institution":"University of Salamanca","correspondingAuthor":false,"prefix":"","firstName":"Gabriel","middleName":"Villarrubia","lastName":"Gonzalez","suffix":""},{"id":638549954,"identity":"d9bc86f7-8fda-4243-bb92-4cfe50243e81","order_by":3,"name":"Ademir Nied","email":"","orcid":"","institution":"Universidade do Estado de Santa Catarina","correspondingAuthor":false,"prefix":"","firstName":"Ademir","middleName":"","lastName":"Nied","suffix":""},{"id":638549955,"identity":"9f437b4c-96e4-4313-a656-52b95ebee957","order_by":4,"name":"Erlon Cristian Finardi","email":"","orcid":"","institution":"Federal University of Santa Catarina","correspondingAuthor":false,"prefix":"","firstName":"Erlon","middleName":"Cristian","lastName":"Finardi","suffix":""},{"id":638549956,"identity":"f697013d-73fc-4993-8b75-c5e5ca56ef29","order_by":5,"name":"Sergio Fagundes","email":"","orcid":"","institution":"Insight Energy","correspondingAuthor":false,"prefix":"","firstName":"Sergio","middleName":"","lastName":"Fagundes","suffix":""}],"badges":[],"createdAt":"2026-05-11 20:59:46","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9684105/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9684105/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109174496,"identity":"6d4e5e6a-e85c-4a38-a98f-ccd25ef1c1c3","added_by":"auto","created_at":"2026-05-13 09:17:04","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":707986,"visible":true,"origin":"","legend":"","description":"","filename":"PaperSpringerRafaelScientificReports2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9684105/v1_covered_de25b5af-b062-40a1-9024-3fd6c0d1656f.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eWiener-NBEATS Hybrid Denoising and Deep Learning Framework for Multi-Horizon Hydroelectric Time Series Forecasting\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Universidade do Estado de Santa Catarina","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":"NBEATS prediction model, hydroelectric power plants, time series forecasting, Wiener filter.","lastPublishedDoi":"10.21203/rs.3.rs-9684105/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9684105/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAccurate forecasting of hydroelectric time series is challenging due to noise and nonlinear temporal dynamics. This paper proposes a hybrid framework combining Wiener filtering and the Neural Basis Expansion Analysis (NBEATS) model for short-term and multi-horizon turbine flow prediction. The Wiener filter is applied as a preprocessing step to optimally denoise the signal while preserving its intrinsic structure, enabling more effective learning. The proposed Wiener-NBEATS approach is evaluated using real data from the Colíder hydroelectric power plant and compared with state-of-the-art models. Results show that preprocessing plays a dominant role in performance. \u0026nbsp;The proposed method achieves MAE values as low as 0.199 and RMSE of 0.239, outperforming transformer-based and recurrent models. Across different horizons, the method maintains strong performance, with MAE around 5.08 for longer horizons. These results demonstrate that the Wiener-NBEATS framework significantly improves forecasting accuracy while maintaining low computational cost, highlighting its effectiveness and applicability for hydroelectric system management.\u003c/p\u003e","manuscriptTitle":"Wiener-NBEATS Hybrid Denoising and Deep Learning Framework for Multi-Horizon Hydroelectric Time Series Forecasting","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-13 09:13:09","doi":"10.21203/rs.3.rs-9684105/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":"583e1b95-8bb4-438a-8add-37c1d5319897","owner":[],"postedDate":"May 13th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-13T09:13:09+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-13 09:13:09","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9684105","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9684105","identity":"rs-9684105","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}