Multi-scale Maritime Network Dynamics Modeling And Prediction Based On Massive Automatic Identification System Ship Trajectory Data

Marine economic cooperation has become an important theme in this era of globalization.Especially in the context of the 21st Century Maritime Silk Road,the interconnection structure and spatiotemporal dynamic of shipping networks have received more and more attention.Maritime ports face many challenges such as navigation safety,operational efficiency,and management difficulty.With the development of Internet of Things,Artificial Intelligence Simulation Technology,Geographic Information System,Cloud Computing Technology,and Maritime Navigation Aided and Decision Support Systems,intelligent modern port construction is possible.This can improve navigation safety and operational efficiency,as well as optimize management system through intelligent services,such as autonomous navigation,self-loading and unload.Intelligent port building and shipping routes improvement are essential to serve growing maritime trade,thereby promoting economic viability and maritime transportation sustainability.Maritime network structure and transportation flow dynamics closely relate to economy,energy,geopolitics,and ocean transport strategy.Investigating the impact of international events on the global shipping network is a challenging task that must comprehensively incorporate geographical,political and maritime transportation science.Understanding global maritime network dynamics is an initial and critical step in this investigation.This paper collected qualitative and quantitative data,including massive ship trajectory,oil price fluctuation,geopolitical policy,military conflict,and economic sanctions.A comprehensive framework integrating maritime network adaptively construction algorithm,maritime network dynamic modeling,influence assessment,and trend prediction has been proposed.The main research contents are as follows.In the port-level maritime network dynamic research,Exploratory Spatial Data Analysis technology is used to analyze spatiotemporal traffic dynamics of the navigation network in port.Furthermore,spatiotemporal risk inside the navigation network has been evaluated base on space-time geography theory integrating behavior pattern of ships.These results are valuable in unmanned ship path optimization and intelligent deployment,as well as smart,green and safe modern ports construction.This research firstly introduced space-time geography theory into the multi-ship collision risk assessment and used direction-constraint space-time prisms to characterize multi-ship potential collision risk.Also,an automatic identification system-based approach for adaptively calibrating near-miss collision risk model,and assessing ships near-miss collision risk using the speed and course patterns of vessels to obtain a robust estimate of the collision risk,have been proposed.The collision risk assessment provided guidance for unmanned ship path optimization driven by safety objective.These results can also help port administration to plan monitoring areas to ensure safe traffic flow in the port.In the national coastal-level maritime network dynamic research,this paper reveals the inter-port linkage dynamic structure of Chinese maritime ports from a complex multilayer perspective based on linkage intensity between subnetworks,the linkage tightness within subnetworks,the spatial isolation between high-intensity backbones and tight skeleton networks,and linkage concentration indices.Automatic Information Systems(AIS)data supports finer-grained spatiotemporal maritime network dynamics analysis,quantitative rigorous methodology development,and expanding the application of graph theory and complex network theory in maritime shipping studies.The presented research fully considers geographical location,development scale and nature,and the range of economic hinterlands related to national coastal-level maritime network dynamic.The external competitiveness and internal cohesion of each subnetwork have been analyzed.The results revealed problems in port management and planning,such as unclear divisions in port operations and weak complementary relationships between the backbone and skeleton networks.In the regional level maritime network dynamic research,a spatial-temporal framework is introduced to analyze shipping network from multi-layers(bulk,container,and tanker)and multidimensional(e.g.,point,link,and network)structure perspectives by means of big trajectory data.Transport capacity and stability are exploited to infer spatial-temporal dynamics of system nodes and links.This study extends the timeline method by taking traffic flow into account to characterize maritime network structure.Maritime network structure changes and traffic flow dynamics grouping including countries along 21^st Century Maritime Silk Road,BRICS,and Union of the United States,Japan,and South Korea are then possible to extract.The shipping dynamics exhibit interesting geographical and spatial variations.Study results indicate that certain countries,such as China,Singapore,Republic of Korea,Australia,and United Arab Emirates,build new corresponding shipping relationships with some ports of countries along the Silk Road and these new linkages carry significant traffic flow.The result shows that internal multi-layer maritime network structural dynamics for countries along 21st Century Maritime Silk Road are higher than those of BRICS and Union of the United States,Japan,and South Korea.However,it is still premature whether countries along 21^st Century Maritime Silk Road will become more competitive than other countries,and possibly hold a better position in the maritime trade.This indicates that the maritime transportation infrastructure,operational efficiency,and shipping routes for countries along 21st Century Maritime Silk Road can be further improved.The results are meaningful for comprehensive policy development integrating cross-regional cooperation,marine economy,and transportation systems.In the global level maritime network dynamic research,this study proposes a comprehensive methodology to investigate the influence of international events on global maritime networks.The methodology incorporates automatic network construction algorithm based on AIS data,spatiotemporal sequence trend modeling,similarity measurement,and vector autoregressive method and vector error correction model.Furthermore,four case studies of international events,military conflict,lifted economic sanctions,government elections,and oil price fluctuation were used to investigate the observed network dynamics possibly affected by international events.The results indicate that container,tanker,and bulk shipping between India and its connected countries all declined significantly after military conflicts between India and Pakistan in August 2015.Tanker shipping between Iran and the United Arab Emirates increased obviously after economic sanctions on Iran were lifted.Container shipping between Sri Lanka and Singapore,Malaysia,and India increased expressively after the general election in Sri Lanka.Evidence suggests that there are different two-way linkages between oil price fluctuations,maritime network structure and traffic flow changes in oil import-dependent and export-dependent countries.International crude oil price fluctuations contribute to maritime network structure changes for many of the countries examined and showed rapid growth peaks in the second or third month.The results of these studies should be of much interest to corporate executives,financial managers,regulators and policymakers and could help domestic and potential foreign investors understand the effect of diversified factors on maritime network changes to effectively manage their portfolio.This study is meaningful to policy formulation,such as shipping modes diversifying and adjustment,evaluating the adaptability of a changing traffic flow and navigation environment,integration of the maritime economy and transportation systems,and countermeasures and counterstrategies development to cope with potential influence.