The Temperature Forecast of Ship Propulsion Devices from Sensor Data
Abstract
:1. Introduction
2. Materials and Methods
2.1. Datasets
2.2. Data Preprocessing and Feature Selection
2.3. Prediction Models
2.3.1. ARDL Model
2.3.2. Stepwise Regression Model
2.3.3. Neural Network Model
3. Results and Discussion
3.1. Preprocessing and Feature Selection
3.2. Results of GTCOAT Temperature Prediction.
- LR1: Time-Series Model with 1 lag.
- NN: Neural Network Model.
- DNN: Deep Neural Network Model.
- SR: Stepwise Regression Model.
3.3. Results of HPTET Temperature Prediction.
3.4. Discussions
4. Conclusions
Author Contributions
Funding
Conflicts of Interest
References
- Ji, Y.L.; Sun, Y.Q.; Zhang, H.P.; Zhang, Y.D.; Chen, H.Q. Novel Ship Propulsion System. Chin. J. Mech. Engin. 2009, 22, 198–206. [Google Scholar] [CrossRef]
- Onwuegbuchunam, D.E.; Ebe, T.E.; Okoroji, L.I. An Analysis of Ship-Source Marine Pollution in Nigeria Seaports. J. Mar. Sci. Eng. 2017, 5, 39. [Google Scholar] [CrossRef]
- Iodice, P.; Langella, G.; Amoresano, A. A numerical approach to assess air pollution by ship engines in manoeuvring mode and fuel switch conditions. Energy Environ. 2017, 28, 827–845. [Google Scholar] [CrossRef]
- Kim, M.; Park, D.W. A study on the green ship design for ultra large container ship. J. Mar. Sci. Eng. 2015, 21, 558–570. [Google Scholar] [CrossRef]
- Gudehus, T.; Kotzab, H. Economic Ship Travel Speed and Consequences for Operating Strategies of Container Shipping Companies. In Proceedings of the 4th International Conference on Dynamics in Logistics (LDIC), Bremen, Germany, 10–14 February 2014; pp. 407–415. [Google Scholar]
- Feng, D.H.; Li, Y.H. Research on Intelligent Diagnosis Method for Large-Scale Ship Engine Fault in Non-Deterministic Environment. Pol. Marit. Res. 2017, 24, 200–206. [Google Scholar] [CrossRef] [Green Version]
- Tang, G.G.; Fu, X.D.; Chen, N. Ship Power Plant Remote Fault Diagnosis System Based on B/S and C/S Architecture. Ship Eng. 2018, 40, 66–71. [Google Scholar]
- Yang, Y.F.; Feng, J. Fault Pattern Recognition and State Prediction Research of Ship Power Equipment Based on HMM-SVR. Ship Eng. 2018, 40, 68–72, 97. [Google Scholar]
- Wu, J.P.; Peng, H.; Ohtsu, K. Ship’s tracking control based on nonlinear time series model. Appl. Ocean Res. 2012, 36, 1–11. [Google Scholar] [CrossRef]
- McCullough, M.; Sakellariou, K.; Stemler, T.; Stemler, T. Regenerating time series from ordinal networks. Chaos. 2017, 27. [Google Scholar] [CrossRef]
- Cinar, Y.G.; Mirisaee, H.; Goswami, P. Period-aware content attention RNNs for time series forecasting with missing values. Neurocomput. 2018, 312, 177–186. [Google Scholar] [CrossRef]
- Yang, S.; Liu, J. Time-Series Forecasting Based on High-Order Fuzzy Cognitive Maps and Wavelet Transform. IEEE Trans. Fuzzy Syst. 2018, 26, 3391–3402. [Google Scholar] [CrossRef]
- Coraddu, A.; Oneto, L.; Ghio, A. Machine learning approaches for improving condition- based maintenance of naval propulsion plants. Proc. Inst. Mech. Eng. Part M J. Eng. Marit. Environ 2016, 230, 136–153. [Google Scholar] [CrossRef]
- Cipollini, F.; Oneto, L.; Coraddu, A.; Murphy, A.J.; Anguita, D. Condition-based maintenance of naval propulsion systems with supervised data analysis. Ocean Eng. 2018, 149, 268–278. [Google Scholar] [CrossRef]
- Yerel, S.; Ersen, T. Prediction of the Calorific Value of Coal Deposit Using Linear Regression Analysis. Energy Sour. Part A Recov. Util. Environ. Eff. 2013, 35, 976–980. [Google Scholar] [CrossRef]
- Liu, X.Q.; Gao, F.; Xu, J.W. Linearized Restricted Ridge Regression Estimator in Linear Regression. Commun. Stat.-Theory Methods 2012, 41, 4503–4514. [Google Scholar] [CrossRef]
- Chang, L.; Roberts, S.; Welsh, A. Robust Lasso Regression Using Tukey’s Biweight Criterion. Technometrics 2018, 60, 36–47. [Google Scholar] [CrossRef]
- Scornet, E. Random Forests and Kernel Methods. IEEE Trans. Inf. Theory 2016, 62, 1485–1500. [Google Scholar] [CrossRef] [Green Version]
- Díaz-Uriarte, R.; DeAndres, S.A. Gene selection and classification of microarray data using random forest. BMC Bioinform. 2006, 7, 1–13. [Google Scholar] [CrossRef]
- Zhang, Y.; Liu, T.; Li, K.F.; Zhang, J.W. Improved visual correlation analysis for multidimensional data. J. Vis. Lang. Computing 2017, 41, 121–132. [Google Scholar]
- Wen, M.L.; Qin, Z.F.; Kang, R. Sensitivity and stability analysis in fuzzy data envelopment analysis. Fuzzy Optim. Decis. Mak. 2011, 10, 1. [Google Scholar] [CrossRef]
- Acikkar, M.; Sivrikaya, O. Prediction of gross calorific value of coal based on proximate analysis using multiple linear regression and artificial neural networks. Turk. J. Electr. Eng. Comput. Sci. 2018, 26, 2541–2552. [Google Scholar] [CrossRef]
- Yue, S.P. Modeling of Engine Parameters for Condition-Based Maintenance of the MTU Series 2000 Diesel Engine. Master’s Thesis, Naval Postgraduate School, Monterey, CA, USA, September 2016. [Google Scholar]
- Agostinelli, C. Robust stepwise regression. J. Appl. Stat. 2002, 29, 825–840. [Google Scholar] [CrossRef] [Green Version]
Features | Linear Regression | Ridge Regression | Lasso Regression | Random Forest | Correlation | Stability | AVG |
---|---|---|---|---|---|---|---|
a1 | 0.0240 | 0.0071 | 0.0248 | 0.0022 | 0.0529 | 0.2518 | 0.0605 |
a2 | 0.0102 | 0.0608 | 0.0943 | 0.0019 | 0.0489 | 0.1027 | 0.0531 |
a3 | 0.0061 | 0.1091 | 0.0000 | 0.0007 | 0.0746 | 0.0489 | 0.0399 |
a4 | 0.0928 | 0.0402 | 0.1069 | 0.0150 | 0.0697 | 0.0440 | 0.0614 |
a5 | 0.0002 | 0.0520 | 0.0000 | 0.0168 | 0.0554 | 0.0073 | 0.0219 |
a6 | 0.3179 | 0.0031 | 0.0000 | 0.0000 | 0.0688 | 0.0122 | 0.0670 |
a7 | 0.0121 | 0.0402 | 0.0000 | 0.1065 | 0.0697 | 0.0245 | 0.0422 |
a8 | 0.0023 | 0.0351 | 0.2630 | 0.0080 | 0.0256 | 0.0685 | 0.0671 |
a9 | 0.0084 | 0.1084 | 0.2995 | 0.5773 | 0.0314 | 0.0000 | 0.1708 |
a10 | 0.0097 | 0.1017 | 0.0000 | 0.0916 | 0.0313 | 0.0000 | 0.0391 |
a11 | 0.0003 | 0.1120 | 0.2115 | 0.0828 | 0.1571 | 0.2689 | 0.1388 |
a12 | 0.0371 | 0.1986 | 0.0000 | 0.0025 | 0.0567 | 0.0049 | 0.0500 |
a13 | 0.0514 | 0.0398 | 0.0000 | 0.0004 | 0.1019 | 0.1222 | 0.0526 |
a14 | 0.0012 | 0.0046 | 0.0000 | 0.0827 | 0.0199 | 0.0024 | 0.0185 |
a15 | 0.0015 | 0.0007 | 0.0000 | 0.0000 | 0.0026 | 0.0000 | 0.0008 |
a16 | 0.0952 | 0.0400 | 0.0000 | 0.0116 | 0.0697 | 0.0220 | 0.0398 |
a17 | 0.3296 | 0.0467 | 0.0000 | 0.0000 | 0.0639 | 0.0196 | 0.0766 |
Features | Linear Regression | Ridge Regression | Lasso Regression | Random Forest | Correlation | Stability | AVG |
---|---|---|---|---|---|---|---|
a1 | 0.0165 | 0.0147 | 0.0000 | 0.0000 | 0.0244 | 0.0748 | 0.0217 |
a2 | 0.0057 | 0.0173 | 0.0159 | 0.0000 | 0.0230 | 0.0262 | 0.0147 |
a3 | 0.1262 | 0.1930 | 0.2302 | 0.0000 | 0.1218 | 0.2037 | 0.1458 |
a4 | 0.0241 | 0.0067 | 0.0550 | 0.0000 | 0.0486 | 0.0037 | 0.0230 |
a5 | 0.0025 | 0.0345 | 0.0000 | 0.0000 | 0.0680 | 0.0093 | 0.0191 |
a6 | 0.2952 | 0.1279 | 0.0617 | 0.0000 | 0.0831 | 0.0131 | 0.0968 |
a7 | 0.0763 | 0.0066 | 0.0027 | 0.0000 | 0.0486 | 0.0019 | 0.0227 |
a8 | 0.0074 | 0.0153 | 0.0463 | 0.0000 | 0.0151 | 0.1364 | 0.0368 |
a9 | 0.0501 | 0.1755 | 0.2563 | 0.9221 | 0.0873 | 0.2056 | 0.2828 |
a10 | 0.0041 | 0.0001 | 0.0959 | 0.0000 | 0.0593 | 0.0206 | 0.0300 |
a11 | 0.0414 | 0.0952 | 0.0227 | 0.0000 | 0.0996 | 0.2037 | 0.0771 |
a12 | 0.0565 | 0.1154 | 0.1982 | 0.0012 | 0.0519 | 0.0000 | 0.0705 |
a13 | 0.0092 | 0.0624 | 0.0000 | 0.0000 | 0.0983 | 0.0486 | 0.0364 |
a14 | 0.0008 | 0.0005 | 0.0000 | 0.0765 | 0.0390 | 0.0486 | 0.0276 |
a15 | 0.0003 | 0.0093 | 0.0053 | 0.0000 | 0.0024 | 0.0000 | 0.0029 |
a16 | 0.1010 | 0.0066 | 0.0000 | 0.0000 | 0.0486 | 0.0000 | 0.0261 |
a17 | 0.1826 | 0.1192 | 0.0099 | 0.0000 | 0.0809 | 0.0037 | 0.0660 |
Feature | Error Measures | ARDL | LR | NN | DNNc | SR | |||
---|---|---|---|---|---|---|---|---|---|
LR1 | LR5 | LR10 | LR15 | ||||||
AVG | MAE MARE | 2.3686 0.0037 | 1.3055 0.0020 | 1.0542 0.0016 | 1.0491 0.0016 | 2.4770 0.0039 | 9.3936 0.0141 | 17.9556 0.0271 | 2.7024 0.0042 |
Corr | MAE MARE | 3.3785 0.0050 | 3.1014 0.0046 | 2.7143 0.0040 | 2.7087 0.0039 | 3.4136 0.0051 | 52.1559 0.0683 | 17.9556 0.0271 | |
Lasso | MAE MARE | 3.4321 0.0050 | 2.9558 0.0043 | 2.5779 0.0037 | 2.5703 0.0037 | 3.4511 0.0050 | 53.3080 0.0699 | 15.5496 0.0233 | |
Lr | MAE MARE | 3.6609 0.0053 | 3.3412 0.0049 | 2.7627 0.0040 | 2.7579 0.0040 | 3.7152 0.0054 | 46.1879 0.0575 | 20.2862 0.0306 | |
RF | MAE MARE | 20.9895 0.0312 | 12.5970 0.0183 | 6.0679 0.0088 | 6.0698 0.0088 | 21.8452 0.0328 | 41.7166 0.0572 | 28.8141 0.0460 | |
Ridge | MAE MARE | 3.3291 0.0049 | 2.9402 0.0043 | 2.5979 0.0038 | 2.5888 0.0038 | 3.3844 0.0049 | 21.3081 0.0335 | 27.7896 0.0405 | |
Stability | MAE MARE | 7.7299 0.0111 | 5.4886 0.0077 | 4.9152 0.0069 | 4.9091 0.0068 | 8.5276 0.0122 | 47.2233 0.0600 | 17.8013 0.0268 |
Feature | Error Measures | ARDL | LR | NN | DNN | SR | |||
---|---|---|---|---|---|---|---|---|---|
LR1 | LR5 | LR10 | LR15 | ||||||
AVG | MAE MARE | 3.3385 0.0049 | 2.8086 0.0041 | 2.5311 0.0037 | 2.5255 0.0037 | 3.3548 0.0049 | 66.4863 0.0855 | 50.5350 0.0645 | 20.8675 0.0317 |
Corr | MAE MARE | 3.3785 0.0050 | 3.1014 0.0046 | 2.7143 0.0040 | 2.7087 0.0039 | 3.4136 0.0051 | 52.1559 0.0683 | 33.5700 0.0502 | |
Lasso | MAE MARE | 3.4321 0.0050 | 2.9560 0.0043 | 2.5779 0.0037 | 2.5703 0.0037 | 3.4511 0.0050 | 53.3080 0.0699 | 54.1260 0.0700 | |
Lr | MAE MARE | 3.6609 0.0053 | 3.3412 0.0049 | 2.7627 0.0040 | 2.7579 0.0040 | 3.7152 0.0054 | 46.1879 0.0575 | 38.9521 0.0615 | |
RF | MAE MARE | 20.9895 0.0312 | 12.5970 0.0183 | 6.0679 0.0089 | 6.0699 0.0088 | 21.8452 0.0328 | 41.7166 0.0572 | 326.0747 0.4034 | |
Ridge | MAE MARE | 3.3290 0.0049 | 2.9402 0.0043 | 2.5979 0.0038 | 2.5888 0.0038 | 3.3844 0.0049 | 21.3081 0.0335 | 42.6881 0.0575 | |
Stability | MAE MARE | 7.7299 0.0111 | 5.4886 0.0077 | 4.9152 0.0069 | 4.9091 0.0069 | 8.5276 0.0122 | 47.2233 0.0600 | 48.1805 0.0612 |
© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Share and Cite
Li, T.; Hua, M.; Yin, Q. The Temperature Forecast of Ship Propulsion Devices from Sensor Data. Information 2019, 10, 316. https://doi.org/10.3390/info10100316
Li T, Hua M, Yin Q. The Temperature Forecast of Ship Propulsion Devices from Sensor Data. Information. 2019; 10(10):316. https://doi.org/10.3390/info10100316
Chicago/Turabian StyleLi, Taoying, Miao Hua, and Qian Yin. 2019. "The Temperature Forecast of Ship Propulsion Devices from Sensor Data" Information 10, no. 10: 316. https://doi.org/10.3390/info10100316