Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Prediction of Vessel Arrival Time to Pilotage Area Using Multi-Data Fusion and Deep Learning

Xiaocai Zhang, Xiuju Fu, Zhe Xiao, Haiyan Xu, Xiaoyang Wei, Jimmy Koh, Daichi Ogawa, Zheng Qin

TL;DR

This paper investigates the prediction of vessels' arrival time to the pilotage area using multi-data fusion and deep learning approaches, and promising results have been obtained.

Abstract

This paper investigates the prediction of vessels' arrival time to the pilotage area using multi-data fusion and deep learning approaches. Firstly, the vessel arrival contour is extracted based on Multivariate Kernel Density Estimation (MKDE) and clustering. Secondly, multiple data sources, including Automatic Identification System (AIS), pilotage booking information, and meteorological data, are fused before latent feature extraction. Thirdly, a Temporal Convolutional Network (TCN) framework that incorporates a residual mechanism is constructed to learn the hidden arrival patterns of the vessels. Extensive tests on two real-world data sets from Singapore have been conducted and the following promising results have been obtained: 1) fusion of pilotage booking information and meteorological data improves the prediction accuracy, with pilotage booking information having a more significant impact; 2) using discrete embedding for the meteorological data performs better than using continuous embedding; 3) the TCN outperforms the state-of-the-art baseline methods in regression tasks, exhibiting Mean Absolute Error (MAE) ranging from 4.58 min to 4.86 min; and 4) approximately 89.41% to 90.61% of the absolute prediction residuals fall within a time frame of 10 min.

Prediction of Vessel Arrival Time to Pilotage Area Using Multi-Data Fusion and Deep Learning

TL;DR

This paper investigates the prediction of vessels' arrival time to the pilotage area using multi-data fusion and deep learning approaches, and promising results have been obtained.

Abstract

This paper investigates the prediction of vessels' arrival time to the pilotage area using multi-data fusion and deep learning approaches. Firstly, the vessel arrival contour is extracted based on Multivariate Kernel Density Estimation (MKDE) and clustering. Secondly, multiple data sources, including Automatic Identification System (AIS), pilotage booking information, and meteorological data, are fused before latent feature extraction. Thirdly, a Temporal Convolutional Network (TCN) framework that incorporates a residual mechanism is constructed to learn the hidden arrival patterns of the vessels. Extensive tests on two real-world data sets from Singapore have been conducted and the following promising results have been obtained: 1) fusion of pilotage booking information and meteorological data improves the prediction accuracy, with pilotage booking information having a more significant impact; 2) using discrete embedding for the meteorological data performs better than using continuous embedding; 3) the TCN outperforms the state-of-the-art baseline methods in regression tasks, exhibiting Mean Absolute Error (MAE) ranging from 4.58 min to 4.86 min; and 4) approximately 89.41% to 90.61% of the absolute prediction residuals fall within a time frame of 10 min.
Paper Structure (11 sections, 15 equations, 6 figures, 6 tables)

This paper contains 11 sections, 15 equations, 6 figures, 6 tables.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. METHOD
  4. Vessel Arrival Contour Extraction
  5. TCN-Based Model Using Multi-Data Fusion
  6. EXPERIMENTS AND EVALUATION
  7. Data Sets
  8. Experimental Settings
  9. Results and Analyses
  10. CONCLUSIONS
  11. ACKNOWLEDGEMENT

Figures (6)

  • Figure 1: The framework of prediction of the vessel arrival time to pilotage area.
  • Figure 2: Illustration of the dilated casual convolution.
  • Figure 3: The overall architecture of TCN for vessel arrival time prediction, where $d$ represents the dilation rate, $ks$ is the kernel size, $f$ stands for the filter size, and $L$ denotes the number of layers.
  • Figure 4: Vessel arrival contour extraction for PEBGA. (a) heatmap of the closest points to PEBGA. (b) refined area after clustering and removing outlier clusters. (c) extracted contour (bounding polygon).
  • Figure 5: Vessel arrival contour extraction for PEBGC. (a) heatmap of the closest points to PEBGC. (b) refined area after clustering and removing outlier clusters. (c) extracted contour (bounding polygon).
  • ...and 1 more figures