Abstract | ||
---|---|---|
yThe accurate prediction of vessel traffic flow plays an important role in practical applications such as port planning, maritime management and maritime security assurance, etc. The original vessel traffic flow data (i.e., one-dimensional time series) is rearranged as a two-dimensional matrix with dates grouped by month and year in this paper. To guarantee the prediction accuracy, we tend to propose a non-convex low-rank plus sparse decomposition model to separate the rearranged matrix into low-rank and sparse matrices. In particular, the low-rank matrix is essentially related to the stable change of traffic flow which most extensively occurs; whereas the sparse matrix corresponds to the drastic change of traffic flow which rarely occurs. The resulting non-convex optimization problem can be efficiently handled using the augmented Lagrange multiplier (ALM) algorithm. The autoregressive integrated moving average (ARIMA) and wavelet neural network (WNN) are introduced to respectively predict the low-rank and sparse matrices (not the whole vessel traffic flow). The final vessel traffic flow can be correspondingly obtained by integrating the predicted low-rank and sparse components. The proposed two-step method is more robust to undesirable artifacts (i.e., data collection errors) and can generate high-accuracy prediction. Experimental results have demonstrated the superior performance of the proposed method in terms of quantitative and qualitative evaluation. |
Year | Venue | Field |
---|---|---|
2017 | 2017 IEEE 20TH INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION SYSTEMS (ITSC) | Computer vision,Data collection,Traffic flow,Matrix (mathematics),Sparse approximation,Algorithm,Autoregressive integrated moving average,Artificial intelligence,Engineering,Augmented lagrange multiplier,Optimization problem,Sparse matrix |
DocType | ISSN | Citations |
Conference | 2153-0009 | 0 |
PageRank | References | Authors |
0.34 | 0 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ryan Wen Liu | 1 | 37 | 13.32 |
Jinwei Chen | 2 | 0 | 0.34 |
Zhao Liu | 3 | 25 | 10.73 |
Yan Li | 4 | 399 | 95.68 |
Yi Liu | 5 | 131 | 54.73 |
Jingxian Liu | 6 | 60 | 14.29 |