CNN-ResLSTM-ConNet: A Novel Hybrid Deep Learning Model for Peak Electricity Demand Forecasting of National Grid

CNN-ResLSTM-ConNet: A Novel Hybrid Deep Learning Model for Peak Electricity Demand Forecasting of National Grid | 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 CNN-ResLSTM-ConNet: A Novel Hybrid Deep Learning Model for Peak Electricity Demand Forecasting of National Grid Dipayan Bhadra, Rumana Rois This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8706187/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract Precise forecasting of daily peak electricity demand is fundamental for ensuring the stability, economic operation, and strategic planning of a national power grid. Conventional statistical models frequently struggle to capture the complex, non-linear patterns inherent in grid load data, which are influenced by temporal dependencies, seasonality, and external factors. While deep learning offers a promising alternative, existing models frequently lack the integrated architecture to simultaneously extract spatial features and model long-term temporal dependencies effectively. This study proposes a novel hybrid deep learning model, termed CNN-ResLSTM-ConNet, to address this gap for forecasting the daily evening peak electricity demand of the Bangladesh national power grid. The model innovatively combines Convolutional Neural Networks (CNN) for extracting salient spatial features, Long Short-Term Memory (LSTM) networks with residual connections (ResLSTM) to capture long-term temporal patterns and mitigate vanishing gradient problems, and a concatenation network that synergistically integrates these features for final prediction. In this research, a comprehensive forecasting solution is presented using the historic data of the daily electricity demand of the power grid from 2016 to 2024. Its performance was benchmarked against a suite of models, including traditional statistical methods (ARIMA, ETS), base deep learning models (MLP, CNN, LSTM), and a state-of-the-art hybrid model (LSTM-mTrans-MLP). The proposed CNN-ResLSTM-ConNet model demonstrated better performance, achieving the lowest error rates across all metrics, including a Root Mean Square Error (RMSE) of 967.30, a Mean Absolute Error (MAE) of 657.08, and a Mean Absolute Percentage Error (MAPE) of 5.45%. These results signify a substantial improvement over all benchmarks, confirming the efficacy of the hybrid architecture. We conclude that the CNN-ResLSTM-ConNet model provides a robust and accurate tool for peak demand forecasting, with the potential to enhance grid reliability, optimize energy dispatch, and inform capacity planning for power utilities. Physical sciences/Energy science and technology Physical sciences/Engineering Physical sciences/Mathematics and computing Electricity Demand Forecasting Deep Learning Hybrid Neural Networks Peak Load Prediction National Power Grid Time Series Forecasting CNN-LSTM Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 11 May, 2026 Reviews received at journal 04 May, 2026 Reviewers agreed at journal 23 Apr, 2026 Reviewers agreed at journal 07 Apr, 2026 Reviews received at journal 26 Feb, 2026 Reviewers agreed at journal 26 Feb, 2026 Reviewers invited by journal 15 Feb, 2026 Editor invited by journal 30 Jan, 2026 Editor assigned by journal 28 Jan, 2026 Submission checks completed at journal 28 Jan, 2026 First submitted to journal 27 Jan, 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-8706187","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":580899925,"identity":"35eae310-758b-4df5-8dae-00f90fc10e5f","order_by":0,"name":"Dipayan Bhadra","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA9ElEQVRIiWNgGAWjYDCCwyCCTQLC+QDEfAwMzMRrYZzBwCDBRlDLAbAWCJuZhxgtfMd5j0kwlFnk8c9IfvbYpuJwHRt782EDhhqbaFxaJA/zpUkwnJMolriRZm6cc+awBBvPseQEhmNpuQ04tBgc5jGTYGyTSGy4kWAmndsG1CKRY3yAseEwYS3zb6R/k7YkScuGGzlm0oxQLQn4tAD9kmwB8ovhmTdlkj1n0iXbgH4xSMDjF77zZw/eYCiry5M7nr5N4keFNT8/MMQkPtTY4NTCwMDDwPyHgSGBQSABSTABh2KYFoga/gN4lY2CUTAKRsEIBgAHSVIYk3MdhgAAAABJRU5ErkJggg==","orcid":"","institution":"Jahangirnagar University","correspondingAuthor":true,"prefix":"","firstName":"Dipayan","middleName":"","lastName":"Bhadra","suffix":""},{"id":580899929,"identity":"2883d9bf-374b-4091-ad30-d52a49211313","order_by":1,"name":"Rumana Rois","email":"","orcid":"","institution":"Jahangirnagar University","correspondingAuthor":false,"prefix":"","firstName":"Rumana","middleName":"","lastName":"Rois","suffix":""}],"badges":[],"createdAt":"2026-01-27 05:53:28","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8706187/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8706187/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101292530,"identity":"b44dc821-dcb4-4e87-a411-08e0dba104ac","added_by":"auto","created_at":"2026-01-28 08:25:09","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2262416,"visible":true,"origin":"","legend":"","description":"","filename":"ManuscriptCNNResLSTMConNetDBRRv6.1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8706187/v1_covered_aca6d94e-c09e-4a57-915b-031026847ca0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"CNN-ResLSTM-ConNet: A Novel Hybrid Deep Learning Model for Peak Electricity Demand Forecasting of National Grid","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"[email protected]","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":"Electricity Demand Forecasting, Deep Learning, Hybrid Neural Networks, Peak Load Prediction, National Power Grid, Time Series Forecasting, CNN-LSTM","lastPublishedDoi":"10.21203/rs.3.rs-8706187/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8706187/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePrecise forecasting of daily peak electricity demand is fundamental for ensuring the stability, economic operation, and strategic planning of a national power grid. Conventional statistical models frequently struggle to capture the complex, non-linear patterns inherent in grid load data, which are influenced by temporal dependencies, seasonality, and external factors. While deep learning offers a promising alternative, existing models frequently lack the integrated architecture to simultaneously extract spatial features and model long-term temporal dependencies effectively. This study proposes a novel hybrid deep learning model, termed CNN-ResLSTM-ConNet, to address this gap for forecasting the daily evening peak electricity demand of the Bangladesh national power grid. The model innovatively combines Convolutional Neural Networks (CNN) for extracting salient spatial features, Long Short-Term Memory (LSTM) networks with residual connections (ResLSTM) to capture long-term temporal patterns and mitigate vanishing gradient problems, and a concatenation network that synergistically integrates these features for final prediction. In this research, a comprehensive forecasting solution is presented using the historic data of the daily electricity demand of the power grid from 2016 to 2024. Its performance was benchmarked against a suite of models, including traditional statistical methods (ARIMA, ETS), base deep learning models (MLP, CNN, LSTM), and a state-of-the-art hybrid model (LSTM-mTrans-MLP). The proposed CNN-ResLSTM-ConNet model demonstrated better performance, achieving the lowest error rates across all metrics, including a Root Mean Square Error (RMSE) of 967.30, a Mean Absolute Error (MAE) of 657.08, and a Mean Absolute Percentage Error (MAPE) of 5.45%. These results signify a substantial improvement over all benchmarks, confirming the efficacy of the hybrid architecture. We conclude that the CNN-ResLSTM-ConNet model provides a robust and accurate tool for peak demand forecasting, with the potential to enhance grid reliability, optimize energy dispatch, and inform capacity planning for power utilities.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e","manuscriptTitle":"CNN-ResLSTM-ConNet: A Novel Hybrid Deep Learning Model for Peak Electricity Demand Forecasting of National Grid","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-28 08:23:51","doi":"10.21203/rs.3.rs-8706187/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-12T02:01:07+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-04T11:34:38+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"300531576844754890014852904470150264144","date":"2026-04-23T08:51:47+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"95170590690555251654486491504196794375","date":"2026-04-08T02:23:39+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-02-27T04:47:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"304115312914472157425602912388684562553","date":"2026-02-27T04:14:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-02-15T15:08:21+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-01-30T10:03:06+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-28T13:21:45+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-28T13:17:53+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-01-27T05:35:49+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","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":"5dbd1230-daf2-4b95-b29e-1cc11b25fbb5","owner":[],"postedDate":"January 28th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-12T02:01:07+00:00","index":58,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-04T11:34:38+00:00","index":57,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":61794042,"name":"Physical sciences/Energy science and technology"},{"id":61794043,"name":"Physical sciences/Engineering"},{"id":61794044,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-02-15T15:23:08+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-28 08:23:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8706187","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8706187","identity":"rs-8706187","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}