2026 Vol.9 Aug.N04

• Title: Real-Time Detection Algorithm for Abnormal Fuel Consumption Based on Multi-Source Ship Data Fusion
• Name: Yuting Jiang
• Company: School of Information Engineering,Jiangsu Maritime Institute,Nanjing,211170,China
• Abstract:

  Existing methods for real-time detection of abnormal fuel consumption on ships rely solely on data from a single fuel sensor in the engine room. Since these methods fail to account for the coupled constraints between various operating parameters and fuel consumption, their detection accuracy is poor. To address this, we propose a real-time detection algorithm for abnormal fuel consumption based on the fusion of multi-source ship data.The correlation strength between the two types of data sources is calculated group by group, and the results of all data correlations are integrated to solve for the global joint correlation coefficient, thereby quantifying the degree of coupling among different vessel operational data.Multidimensional voyage operating condition parameters are fused and converted into a dynamic reference fuel consumption rate that aligns with the real-time environment. Combining this with the vessel’s inherent equipment wear parameters and a baseline fuel consumption value, the dynamic reference fuel consumption rate is derived through an operating condition–fuel consumption mapping model.Using the dynamic reference fuel consumption as a benchmark, the magnitude of fuel consumption deviation is calculated. By integrating correlation coefficients from multiple data sources, a graded anomaly determination index is constructed. Based on preset intervals, normal, mild, and severe anomaly states are distinguished in real time, enabling dynamic identification of vessel fuel consumption anomalies and graded early warning.In the experiment, the detection accuracy of the proposed method was verified. The test comparison results clearly show that when using the proposed method for real-time detection of abnormal fuel consumption, the identification accuracy for abnormal samples remains consistently stable within the range of 0.79 to 0.95, demonstrating relatively ideal detection performance.

• Keyword: multi-source ship data; data fusion; abnormal fuel consumption; real-time detection algorithm;
• DOI: 10.12250/jpciams2026090802
• Citation form: Yuting Jiang.Real-Time Detection Algorithm for Abnormal Fuel Consumption Based on Multi-Source Ship Data Fusion[J]. Computer Informatization and Mechanical System,2026,Vol.9,pp.

[1] Chen Y, Sun B, Xie X, Li Y, Zhao Y, Li X. Short-term forecasting for ship fuel consumption based on deep learning [J]. Ocean Engineering, 2024, 301(Jun. 1): 1.1–1.11.

[2] Maliarenko O., Maistrenko N., Eutukhova T. Factors Influencing the Forecast of a Country’s Energy Consumption Under Wartime Conditions and the Magnitude of Greenhouse Gas Emissions Reductions [J]. System Research in Energy, 2024, 2024(4):65–76.

[3] Huang W. Forecast of Natural Gas Consumption in the U.S. Based on Time Series Analysis and the ARIMA Model [J]. Theoretical and Natural Science, 2024, 51(1):157-164.

[4] Jha S., Agrawal H., Rai P. AI-Powered Prediction for Estimating Specific Fuel Consumption in Heavy-Duty Dump Trucks in Coal Mines [J]. Journal of the Institution of Engineers (India): Series D, 2026, 107(1):657-666.

[5] Chen Z. A Forecast of CO₂ Emissions Based on the Fuel Consumption Rating of a Particular Vehicle [J]. Highlights in Science, Engineering and Technology, 2024, 92:87–93.

[6] Rizzo G., Tiano F. A. An Automated Method for Obtaining Engine Fuel Consumption and Efficiency Models from Road Test Data [J]. International Journal of Powertrains, 2025, 14(2):161–183.

[7] Kim Y. C., Ku S. K. Simulation of Aircraft Fuel Consumption in Domestic Airspace Using Flight Trajectory Data and AEDT [J]. Journal of Advanced Navigation Technology, 2025, 29(1):52–60.

[8] Cai Z, Li L, Sun L M. Diversity, Quality, and Quantity of Real Ship Data for Black-Box and Gray-Box Prediction Models of Ship Fuel Consumption [J]. Ocean Engineering, 2024, 291(Jan. 1):116434.1–116434.14.

[9] Katyare P, Joshi S S, Rajapurkar S. Real-time data modeling for forecasting fuel consumption of construction equipment using an integrated approach of IoT and ML techniques [J]. Journal of Information and Optimization Sciences, 2023, 44(3):427–437.

[10] Ansorena I L. Service Anomaly Detection in Dry Bulk Terminals: A Machine Learning Approach [J]. International Journal of Shipping and Transport Logistics: IJSTL, 2023, 17(3):281–302.

[11] Wang Z, Dahouda M K, Hwang H, et al. Explanatory LSTM-AE-Based Anomaly Detection for Time Series Data in Marine Transportation [J]. IEEE Access, 2025, 13(000):23195–23208.

[12] Ilias L , Blika A , Michalitsi-Psarrou A ,et al.Predicting Ship Main Engine Fuel Oil Consumption Using a Mixture of Experts Approach[J].IEEE Access, 2026, 14:45464-45474.

[13] Sanchez P. D., Garcia H. J., Molina L. J. M. Context Learning from a Ship Trajectory Cluster for Anomaly Detection [J]. Neurocomputing, 2024, 563(Jan. 1):1.1–1.14.

[14] Zhang J, Wu E, Lv T. Research on Anomaly Detection of Ship Behavior Based on Data Mining[J]. Frontiers in Science and Engineering, 2024, 4(10):133–141.

[15] Kexin L. I., Guo J., Ranchong L. I., et al. A Ship Trajectory Anomaly Detection Method Based on an Encoder-Decoder Architecture Composed of Transformer-LSTM Modules [J]. Chinese Journal of Ship Research, 2024, 19(2):223–232.