VERİYE DAYALI YÖNTEMLER YARDIMI İLE KİMYASAL TANKERDE YAKIT TÜKETİMİ TAHMİNİ

Yıl 2022, , 190 - 205, 30.12.2022

Tayfun Uyanık

https://doi.org/10.18613/deudfd.1015260

Cited By: 1

Öz

Ticari gemilerde yakıt tüketimi denizcilik işletmelerinde en önemli gider kalemini oluşturmaktadır. Aynı zamanda enerji verimliliği ile de yakından alakalı olan bu konu denizcilik sektörü açısından son derece önem arz etmektedir. Uluslararası Denizcilik Örgütü kuralları gereği denizcilik sektöründe emisyon azaltma konusunun gündemdeki yerini koruduğu da göz önünde bulundurulduğunda gemilerde yakıt tüketimi ve ortaya çıkan emisyonlar denizcilik otoriteleri tarafından ciddi olarak takip edilmektedir.
Bu çalışmada bir kimyasal tanker gemisinin yakıt tüketimi gerçek sefer verilerinden hareketle veriye dayalı yöntemler yardımıyla modellenip tahmin edilmiştir. Öncelikle gemiden alınan sefer verileri işlenip algoritmaların üzerinde çalışabileceği hale getirilmiştir. Algoritmalar veri seti üzerinde çalıştırılmış ve yakıt tüketimi tahmin başarımları incelenmiştir. İlk aşamada bazı algoritmaların başarısı yetersiz bulunmuştur. Tahmin başarımları yetersiz bulunan algoritmaların parametreleri ayarlanarak tahmin işlemi tekrar edilmiştir. Son olarak hata metrikleri kullanılarak algoritmaların yaptığı tahminler karşılaştırılmıştır. Sonuçlar incelendiğinde Çok Katmanlı Derin Sinir Ağı yönteminin kimyasal tanker yakıt tüketimi tahmini problemi kapsamında ele alınan diğer yöntemlerden daha başarılı olduğu tespit edilmiştir.

Anahtar Kelimeler

Kimyasal tanker, enerji verimliliği, yakıt tüketimi tahmini, veriye dayalı yöntemler.

FUEL CONSUMPTION PREDICTION IN CHEMICAL TANKER WITH DATA-DRIVEN METHODS

Öz

Fuel consumption in commercial ships constitutes the most important expense item in maritime enterprises. This issue, which is also closely related to energy efficiency, is extremely important for the maritime industry. Considering that the issue of emission reduction remains on the agenda in the maritime sector as per the International Maritime Organization rules, fuel consumption and the emissions on ships are followed seriously by the maritime authorities.
In this study, the fuel consumption of a chemical tanker ship was modeled and estimated with the help of data-driven methods based on real voyage data. First of all, the voyage data taken from the ship was processed and made into a way that algorithms can work on. Algorithms were run on the data and fuel consumption prediction performances were examined. The success of some models established in the first stage was found to be inadequate. The estimation process was repeated by tuning the algorithm parameters with unsatisfactory estimation performance. Finally, the predictions were compared using error metrics. When the results are examined, it has been determined that the Multi-Layer Deep Neural Network method is more successful than the other methods discussed for the chemical tanker fuel consumption estimation problem.

Anahtar Kelimeler

Chemical tanker, energy efficiency, fuel consumption estimation, data-driven methods.

Kaynakça

• Ahlgren F., Mondejar M.E., Thern M. (2019). Predicting Dynamic Fuel Oil Consumption on Ships with Automated Machine Learning. Energy Procedia, 158, 6126-6131, 1876-6102.
• Aline F. S., Nicolau A. C., André D. S. B., José E. S., Amauri G., Noé C., Bismarck L. S. (2021). Multiple linear regression approach to predict tensile properties of Sn-Ag-Cu (SAC) alloys. Materials Letters, 304, 130587, 0167-577X.
• Bui-Duy L., Vu-Thi-Minh N. (2021). Utilization of a deep learning-based fuel consumption model in choosing a liner shipping route for container ships in Asia. The Asian Journal of Shipping and Logistics. 37, 1,1-11, 2092-5212.
• Cheliotis M., Lazakis I, Theotokatos G. (2020). Machine learning and data-driven fault detection for ship systems operations. Ocean Engineering, 216, 107968, 0029-8018.
• Chen C., Ruiz M.T., Delefortrie G., Mei T., Vantorre M., Lataire E. (2019). Parameter estimation for a ship's roll response model in shallow water using an intelligent machine learning method, Ocean Engineering, 191, 106479, 0029-8018.
• Chen C.H., Wu J.C., Chen J.C. (2008). Prediction of flutter derivatives by artificial neural networks, Journal of Wind Engineering and Industrial Aerodynamics, 96, 10–11, 1925-1937, 0167-6105.
• Chen L, Gao X., Li X. (2021). Using the motor power and XGBoost to diagnose working states of a sucker rod pump, Journal of Petroleum Science and Engineering, 199, 108329, 0920-4105.
• Choi S., Kim Y.J. (2021). Artificial neural network models for airport capacity prediction, Journal of Air Transport Management, 97, 102146, 0969-6997.
• Desai M., Shah M. (2021). An anatomization on breast cancer detection and diagnosis employing multi-layer perceptron neural network (MLP) and Convolutional neural network (CNN), Clinical eHealth, 4, 1-11, 2588-9141.
• Fabregat A., Vázquez L., Vernet A. (2021). Using Machine Learning to estimate the impact of ports and cruise ship traffic on urban air quality: The case of Barcelona. Environmental Modelling & Software. 139,104995, 1364-8152.
• Gkerekos C., Lazakis I., Theotokatos G. (2019). Machine learning models for predicting ship main engine Fuel Oil Consumption: A comparative study. Ocean Engineering. 188, 106282,0029-8018.
• López J.E.F., Mancini C. (2019). Optimum thresholding using mean and conditional mean squared error. Journal of Econometrics, 208, 1, 179-210, 0304-4076.
• Olsen A.A., McLaughlin J.E., Harpe S.E. (2020). Using multiple linear regression in pharmacy education scholarship. Currents in Pharmacy Teaching and Learning. 12, 10, 1258-1268, 1877-1297.
• Peng Y., Liu H., Li X., Huang J., Wang W. (2020). Machine learning method for energy consumption prediction of ships in port considering green ports. Journal of Cleaner Production, 264, 121564,0959-6526.
• Paredes L.F., Mallor F., Romeo M.G., León T. (2018). Dynamic mean absolute error as a new measure for assessing forecasting errors, Energy Conversion and Management, 162, 176-188, 0196-8904.
• Rawson A., Brito M., Sabeur Z., Tran-Thanh L. (2021). A machine learning approach for monitoring ship safety in extreme weather events. Safety Science, 141, 105336, 0925-7535.
• Sahu P., Raghavan S., Chandrasekaran K. (2021). Ensemble deep neural network-based quality of service prediction for cloud service recommendation. Neurocomputing, 465, 476-489, 0925-2312.
• Salman M.S., Kukrer O., Hocanin A. (2017). Recursive inverse algorithm: Mean-square-error analysis, Digital Signal Processing, 66, 10-17, 1051-2004.
• Schubert A.L., Hagemann D., Voss A., Bergmann K. (2017). Evaluating the model fit of diffusion models with the root mean square error of approximation, Journal of Mathematical Psychology, 77, 29-45, 0022-2496.
• Shehadeh A., Alshboul O., Mamlook R.E. A, Hamedat O. (2021). Machine learning models for predicting the residual value of heavy construction equipment: An evaluation of modified decision tree, LightGBM, and XGBoost regression. Automation in Construction. 129, 103827, 0926-5805.
• Sun W., Su F., Wang L. (2018). Improving deep neural networks with multi-layer maxout networks and a novel initialization method. Neurocomputing. 278, 34-40, 0925-2312.
• Thiangchanta S., Chaichana C. (2020). The multiple linear regression models of heat load for an air-conditioned room. Energy Reports, 6, 9, 972-977, 2352-4847.
• Tien-Anh T. (2021). Comparative analysis on the fuel consumption prediction model for bulk carriers from ship launching to current states based on sea trial data and machine learning technique. Journal of Ocean Engineering and Science, 2468-0133.
• Ueki M. (2021). Testing conditional mean through regression model sequence using Yanai’s generalized coefficient of determination, Computational Statistics & Data Analysis, 158, 107168, 0167-9473.
• Zhou T., Hu Q., Hu Z., Zhen R. (2021). An adaptive hyper-parameter tuning model for ship fuel consumption prediction under complex maritime environments. Journal of Ocean Engineering and Science. 2468-0133.