{"id":{"repo_id":"rowan","oai_identifier":"oai:rdw.rowan.edu:etd-1064"},"canonical_url":"https://search.dev.ndltd.org/etd/rowan/oai:rdw.rowan.edu:etd-1064","repository":{"repo_id":"rowan","name":"Rowan University","base_url":"https://rdw.rowan.edu/do/oai/"},"display":{"title":"Analysis of alternative methods for long-term wind speed and initial site assessment for purposes of wind energy production estimation","abstract":"A need exists for identification and evaluation of viable sites for wind energy. However, this is made difficult by the site-specific nature of the wind. A review of current industry practices shows that linear regression is the preferred method, although little work has been done to investigate alternative approaches. To address this, various methods were tested on 23 target sites with data measured over a long-term period. Each site was analyzed with the chosen best reference station, across 10 data periods for each method with results classified as mean absolute percentage error in energy density between the projected and measured results. Of primary concern was the comparative performance of these alternative methods to the current industry standard. The method that performed best was multiple regression analysis (MRA), which was 0.85% better than linear regression in absolute terms. The multilayer perceptron (MLP) artificial neural network (ANN) also exhibited promise in dealing with some of the non-linear aspects of the data sets, although its accuracy was found to be worse than linear regression by 0.79% in absolute terms. The affect of seasonality was also examined, showing extreme results in months of high and low wind speeds, but also results as close as 1.5% to the expected result when combining these extreme months to reach the expected mean wind speed.","abstract_html":"A need exists for identification and evaluation of viable sites for wind energy. However, this is made difficult by the site-specific nature of the wind. A review of current industry practices shows that linear regression is the preferred method, although little work has been done to investigate alternative approaches. To address this, various methods were tested on 23 target sites with data measured over a long-term period. Each site was analyzed with the chosen best reference station, across 10 data periods for each method with results classified as mean absolute percentage error in energy density between the projected and measured results. Of primary concern was the comparative performance of these alternative methods to the current industry standard. The method that performed best was multiple regression analysis (MRA), which was 0.85% better than linear regression in absolute terms. The multilayer perceptron (MLP) artificial neural network (ANN) also exhibited promise in dealing with some of the non-linear aspects of the data sets, although its accuracy was found to be worse than linear regression by 0.79% in absolute terms. The affect of seasonality was also examined, showing extreme results in months of high and low wind speeds, but also results as close as 1.5% to the expected result when combining these extreme months to reach the expected mean wind speed.","abstract_has_math":false,"creators":["McDevitt, Ryan"],"institution":null,"degree_name":"M.S. Engineering","degree_level":"Thesis","degree_discipline":"Electrical and Computer Engineering","degree_department":null,"school":null,"contributors":["Jansson, Peter"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2010,"date_issued":"2010-09-09T07:00:00Z","date_published":"2010-09-09T07:00:00Z","updated_at":"2026-07-24T04:12:49Z","subjects":["Wind power--Evaluation","Electrical and Computer Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://rdw.rowan.edu/etd/65","outbound_label":"Repository record","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Jansson, Peter"]},{"key":"dc:creator","label":"Author","values":["McDevitt, Ryan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2020-03-03T19:58:35Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Electrical and Computer Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["M.S. Engineering"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Wind power--Evaluation","Electrical and Computer Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://rdw.rowan.edu/etd/65"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["A need exists for identification and evaluation of viable sites for wind energy. However, this is made difficult by the site-specific nature of the wind. A review of current industry practices shows that linear regression is the preferred method, although little work has been done to investigate alternative approaches. To address this, various methods were tested on 23 target sites with data measured over a long-term period. Each site was analyzed with the chosen best reference station, across 10 data periods for each method with results classified as mean absolute percentage error in energy density between the projected and measured results. Of primary concern was the comparative performance of these alternative methods to the current industry standard. The method that performed best was multiple regression analysis (MRA), which was 0.85% better than linear regression in absolute terms. The multilayer perceptron (MLP) artificial neural network (ANN) also exhibited promise in dealing with some of the non-linear aspects of the data sets, although its accuracy was found to be worse than linear regression by 0.79% in absolute terms. The affect of seasonality was also examined, showing extreme results in months of high and low wind speeds, but also results as close as 1.5% to the expected result when combining these extreme months to reach the expected mean wind speed."]},{"key":"dc:title","label":"Title","values":["Analysis of alternative methods for long-term wind speed and initial site assessment for purposes of wind energy production estimation"]}]}],"canonical_facts":{"dc:contributor":["Jansson, Peter"],"dc:creator":["McDevitt, Ryan"],"dc:date.available":["2020-03-03T19:58:35Z"],"dc:description.abstract":["A need exists for identification and evaluation of viable sites for wind energy. However, this is made difficult by the site-specific nature of the wind. A review of current industry practices shows that linear regression is the preferred method, although little work has been done to investigate alternative approaches. To address this, various methods were tested on 23 target sites with data measured over a long-term period. Each site was analyzed with the chosen best reference station, across 10 data periods for each method with results classified as mean absolute percentage error in energy density between the projected and measured results. Of primary concern was the comparative performance of these alternative methods to the current industry standard. The method that performed best was multiple regression analysis (MRA), which was 0.85% better than linear regression in absolute terms. The multilayer perceptron (MLP) artificial neural network (ANN) also exhibited promise in dealing with some of the non-linear aspects of the data sets, although its accuracy was found to be worse than linear regression by 0.79% in absolute terms. The affect of seasonality was also examined, showing extreme results in months of high and low wind speeds, but also results as close as 1.5% to the expected result when combining these extreme months to reach the expected mean wind speed."],"dc:identifier":["https://rdw.rowan.edu/etd/65"],"dc:subject":["Wind power--Evaluation","Electrical and Computer Engineering"],"dc:title":["Analysis of alternative methods for long-term wind speed and initial site assessment for purposes of wind energy production estimation"],"thesis:degree_discipline":["Electrical and Computer Engineering"],"thesis:degree_level":["Thesis"],"thesis:degree_name":["M.S. Engineering"]},"updated_at":"2026-07-24T04:12:49Z"}