{ "article_id": 1123, "public_id": "c2936c33-a3fe-4ae5-a5e7-4514c1c6b747", "internal_identifier": null, "source_identifier": "42:128", "status": "published", "type": "research_article", "title": "GAUSS FONKSİYONLARI İLE PENCERELENMİŞ POLİNOM AKTİVASYON FONKSİYONU KULLANAN DALGACIK SİNİR AĞI İLE YAPAY SİNİR AĞLARININ, ÇOK KISA VADELİ RÜZGAR HIZI TAHMİNİNDE, TAHMİN BAŞARIMLARININ KARŞILAŞTIRILMASI", "title_secondary": null, "abstract": "Fosil bazlı yakıtların giderek ömrünü tamamladığı dünyamızda alternatif enerji kaynaklarının önemi hızla artmaktadır. Bir alternatif enerji şekli olan rüzgar enerjisinde, rüzgar türbinleri kurulu olduğu alanın sadece %1’ini işgal etmesi ve kurulduğu kırsal alanlarda tarım, hayvancılık ve ormancılık gibi faaliyetlere engel olmadığı için güneş enerji santrallerine göre daha fazla tercih edilebilir konumdadır. Ancak rüzgar enerjisinin değişkenliği ve süreksizliği, elektrik şebekesi içine rüzgar enerjisini entegre etmenin en büyük zorluğudur. Bu durumun önüne geçebilmenin en akıllıca yolu, rüzgar enerjisi üretiminin gelecekteki değerlerini tahmin etmektir. Yüksek doğrulukları ile rüzgar hızı tahmin yöntemleri bu sorunları en aza indirmek için kullanılabilecek etkili bir araçtır. Bu çalışma, geleneksel fonksiyonlu Yapay Sinir Ağları (YSA) ve Gauss Fonksiyonları ile Pencerelenmiş Polinom fonksiyonlarından POLYWOG1 aktivasyon fonksiyonu kullanan Dalgacık Sinir Ağları (DSA) kullanılarak çok kısa süreli rüzgar hızı tahminini ve bu ağların başarımlarının karşılaştırmasını sunmaktadır. Veri, bir yıllık bir süre için on dakikalık çözünürlükte Ondokuz Mayıs Üniversitesi'nde bulunan bir hava istasyonundan toplanmıştır. Rüzgar hızı tahminler önümüzdeki 10 dakika için 24 saatlik bir süre içinde sunulmaktadır. YSA ve DSA için aynı topoloji kullanılmasına rağmen, DSA için önerilen tahmin sisteminin başarımı, YSA’ya göre daha yüksektir. Kök Hata Kareler Ortalaması (RMSE) ve Ortalama Karesel Hata (MSE) başarım etmeni olarak seçilmiştir. Rüzgar enerjisi sağlayan sistemlerin, elde edilecek tahminler doğrultusunda, değişen atmosferik şartlara doğru ve hızlı şekilde uyumluğunun sağlanması hedeflenmektedir.", "language": "tr", "english_title": "COMPARISON OF THE PREDICTION PERFORMANCE OF WAVELET NEURAL NETWORKS AND ARTIFICIAL NEURAL NETWORKS USING POLYNOMIAL ACTIVATION FUNCTIONS WITH GAUSSIAN FUNCTIONS IN VERY SHORT-TERM WIND SPEED FORECASTING", "english_abstract": "In our world where fossil fuels are increasingly reaching the end of their lifespan, the importance of alternative energy sources is rapidly increasing. Wind energy, as an alternative energy form, is more preferable than solar power plants because wind turbines occupy only 1% of the area where they are installed and do not interfere with agricultural, livestock, and forestry activities in rural areas. However, the variability and discontinuity of wind energy is the biggest challenge in integrating it into the electricity grid. The smartest way to overcome this is to predict the future values ​​of wind energy production. High-accuracy wind speed prediction methods are an effective tool that can be used to minimize these problems. This study presents and compares the performance of Artificial Neural Networks (ANNs) with traditional functions and Wavelet Neural Networks (WNNs) using the POLYWOG1 activation function from windowed polynomial functions with Gaussian functions for very short-term wind speed prediction. Data was collected from a weather station at Ondokuz Mayıs University for a one-year period with ten-minute resolution. Wind speed predictions are presented for the next 10 minutes within a 24-hour period. Although the same topology is used for ANN and WNN, the performance of the proposed prediction system for WNN is higher than that of ANN. Root Mean Square Error (RMSE) and Mean Squared Error (MSE) were selected as performance factors. The aim is to ensure that wind energy systems adapt accurately and quickly to changing atmospheric conditions in line with the predictions obtained.", "doi": null, "doi_url": null, "resource_url": "https://www.set-science.com/a/c2936c33-a3fe-4ae5-a5e7-4514c1c6b747", "pdf_url": "https://www.set-science.com/a/c2936c33-a3fe-4ae5-a5e7-4514c1c6b747/pdf", "og_image": "https://www.set-science.com/assets/img/og-card-v2.png", "authors": [ { "public_id": "ae272939-ac3b-42d2-8edf-c70023c626bf", "given": "Serkan   ", "family": "Şenkal", "full": "Serkan    Şenkal", "literal": null, "orcid": null, "orcid_url": null, "email": "serkan.senkal@gop.edu.tr", "corresponding": true, "sequence": "first", "affiliations": [ { "name": "Tokat Gaziosmanpaşa Üniversitesi  ", "department": null, "country": null, "country_code": null, "ror": null, "ror_url": null } ] }, { "public_id": "47c549f3-7805-4d10-84c1-8c2a7a411142", "given": "Cem   ", "family": "Emeksiz", "full": "Cem    Emeksiz", "literal": null, "orcid": null, "orcid_url": null, "email": "serkan.senkal@gop.edu.tr", "corresponding": false, "sequence": "additional", "affiliations": [ { "name": "Tokat Gaziosmanpaşa Üniversitesi  ", "department": null, "country": null, "country_code": null, "ror": null, "ror_url": null } ] } ], "keywords": [ "Dalgacık Sinir Ağı", "Yapay Sinir Ağı", "Rüzgar Hızı Tahmini" ], "translations": [ { "language": "en", "title": "COMPARISON OF THE PREDICTION PERFORMANCE OF WAVELET NEURAL NETWORKS AND ARTIFICIAL NEURAL NETWORKS USING POLYNOMIAL ACTIVATION FUNCTIONS WITH GAUSSIAN FUNCTIONS IN VERY SHORT-TERM WIND SPEED FORECASTING", "abstract": "In our world where fossil fuels are increasingly reaching the end of their lifespan, the importance of alternative energy sources is rapidly increasing. Wind energy, as an alternative energy form, is more preferable than solar power plants because wind turbines occupy only 1% of the area where they are installed and do not interfere with agricultural, livestock, and forestry activities in rural areas. However, the variability and discontinuity of wind energy is the biggest challenge in integrating it into the electricity grid. The smartest way to overcome this is to predict the future values ​​of wind energy production. High-accuracy wind speed prediction methods are an effective tool that can be used to minimize these problems. This study presents and compares the performance of Artificial Neural Networks (ANNs) with traditional functions and Wavelet Neural Networks (WNNs) using the POLYWOG1 activation function from windowed polynomial functions with Gaussian functions for very short-term wind speed prediction. Data was collected from a weather station at Ondokuz Mayıs University for a one-year period with ten-minute resolution. Wind speed predictions are presented for the next 10 minutes within a 24-hour period. Although the same topology is used for ANN and WNN, the performance of the proposed prediction system for WNN is higher than that of ANN. Root Mean Square Error (RMSE) and Mean Squared Error (MSE) were selected as performance factors. The aim is to ensure that wind energy systems adapt accurately and quickly to changing atmospheric conditions in line with the predictions obtained.", "keywords": [ "Wavelet Neural Network", "Artificial Neural Network", "Wind Speed ​​Forecasting" ] } ], "subjects": [], "subject_names": [], "published_date": "2019-06-01", "published_year": "2019", "received_date": null, "accepted_date": null, "volume": "4", "issue": null, "page_first": "651", "page_last": "655", "article_number": null, "conference": { "short_name": "ISAS2019-ENS", "full_title": "3rd International Symposium on Innovative Approaches in Scientific Studies (Engineering and Natural Sciences)", "name": "3rd International Symposium on Innovative Approaches in Scientific Studies (Engineering and Natural Sciences)", "acronym": "ISAS2019-ENS", "location": "Ankara, Türkiye", "start_date": "2019-04-19", "end_date": null, "public_id": "ed3b36e9-38cf-4aeb-a135-833c969cc89f" }, "proceedings": { "title": "3rd International Symposium on Innovative Approaches in Scientific Studies (Engineering and Natural Sciences)", "series_title": "Setsci conference indexing system", "publisher": "SET Technology", "isbn": null, "eisbn": null, "issn": "2618-6446", "eissn": null, "doi": null, "doi_prefix": null, "slug": "setsci-volume-4-2019-legacy-42", "public_id": "24d215ef-a545-4351-9725-6e9bdb167b3c", "url": "https://www.set-science.com/p/setsci-volume-4-2019-legacy-42" }, "license": { "name": "CC BY 4.0", "url": "https://creativecommons.org/licenses/by/4.0/", "code": "CC-BY-4.0", "open_access": true }, "copyright": { "holder": "SET Technology", "year": "2019" }, "funding_statement": null, "funders": [], "references": [ { "order": 1, "raw": "[1] Tasçıkaraoğlu A., Uzunoglu M., “Dalgacık Dönüşümü ve Yapay Sinir Ağları ile Rüzgar Hızı Tahmini”, Elektrik-Elektronik ve Bilgisayar Sempozyumu FEEB 2011, Fırat Üniversitesi-Elazığ, 5-7 Ekim 2011.\r\n\r\n[3] Soman S. S., Zareipour H., Malik O., and MandaI P., \"A review of wind power forecasting methods with different time horizons,\" in North American Power Symposium, Arlington, TX, Sep. 26-28, 2010. Tahminlerinin Zaman Bazında Sınıflandırılması.\r\n\r\n[47] Kalogirou, S. A., 2000. Artificial neural networks in renewable energy systems applications: a review, Renewable and Sustainable Energy Reviews, 5, 373 – 401.\r\n\r\n[48] Kalogirou, S.A., 1999. Applications of artificial neural networks in energy systems a review, Energy Conversion & Management, 40, 1073-1087.\r\n\r\n[49] Fu, L.M., 1994. Neural Networks in Computer Intelligence, McGrawHill International Editions.\r\n\r\n[50] Rumelhart D.E., Hinton, G.E., ve Williams, R.J.1986. Learning internal representation by error propagation. Parallel distributed processing, Vol. 1: Foundations, MIT Press, Cambridge, Mass.\r\n\r\n[4] Veitch D., 2005, Wavelet Neural Networks and their application in the study of dynamical systems, Yüksek Lisans Tezi, York Üniversitesi, Matematik Bölümü, York /İngiltere.\r\n\r\n[5] Baily, D., Thomson, D., “Developing Neural Network Application”. AI Expert, 33-41, 1990.\r\n\r\n[6] Katz, J. O., “Developing Neural Network Forecasters for Trading”, Technical Analysis of Stocks and Commodities, 58-70, 1992.\r\n\r\n[7] Doig, G., “Severity of Illness Scoring in the Intensive Care Unit: A Comparison of Logistic Regression and Artificial Neural Networks”, PhD , The University of Western Ontario, 1999.\r\n\r\n[8] Jayalakshmi, T. and Santhakumaran A.; (2011), “Statistical Normalization and Back Propagation for Classification’’, International Journal of Computer Theory and Engineering, 3(1), pp. 1793-8201.\r\n\r\n[9] Öztemel, E; (2003), Yapay Sinir Ağları, Birinci Baskı, İstanbul: Papatya Yayıncılık.\r\n\r\n[10] Doğan, E., Işık S. ve Mehmet Sandalcı M., (2007), “ Günlük Buharlaşmanın Yapay Sinir Ağları Kullanarak Tahmin Edilmesi’’, İMO Teknik Dergi, ss. 4119 -4131.\r\n\r\n[11] Challagulla, V.U.B. , Bastani, F.B., I-Ling Yen, Paul, (2005) “Empirical assessment of machine learning based software defect prediction techniques”, 10th IEEE International Workshop on ObjectOriented Real-Time Dependable Systems, WORDS 2005, 2-4 Feb 2005, sayfa 263-270.", "doi": null, "title": null, "authors": null, "source": null, "year": null, "volume": null, "pages": null } ], "statements": { "acknowledgements": null, "conflict_of_interest": null, "data_availability": null, "ai_disclosure": null }, "credit": [], "history": { "published": "2019-06-01" } }