A port channel congestion monitoring and early warning method fusing multi-source heterogeneous data

Abstract

This invention belongs to the field of smart port and intelligent waterway traffic management technology, and discloses a method for monitoring and early warning of port and waterway congestion by integrating multi-source heterogeneous data. This invention constructs a unified spatiotemporal grid for the port by collecting data from the Automatic Identification System (AIS), shore-based radar, hydrological and meteorological data, and terminal production operation data, and then fuses the multi-source data. Based on the fused data, it extracts characteristic parameters such as waterway traffic flow density, actual flow rate, and basic capacity. A port collaborative dynamic model is established, using ship flow transfer equations to describe the traffic flow coupling relationship and congestion propagation mechanism between waterway networks. The time-varying capacity is dynamically corrected based on real-time hydrological and meteorological conditions, and a dynamic congestion index is calculated. Finally, tiered early warning information is triggered based on the current congestion status and the future trends predicted by the model. This invention achieves accurate perception of the traffic situation in the port and waterway network and forward prediction of congestion trends, improving the overall navigation efficiency and collaborative scheduling capabilities of the port.

Description

Technical Field

[0001] This invention relates to the field of smart port and intelligent waterway traffic management technology, specifically to a method for monitoring and early warning of port and waterway congestion by integrating multi-source heterogeneous data. Background Technology

[0002] With the rapid development of global shipping and trade, coastal ports have become the core hubs of the international logistics supply chain. The interwoven waterway networks and dense and heterogeneous ship traffic within ports have made waterway congestion a significant factor that restricts port operational efficiency and threatens navigation safety.

[0003] Current technologies for monitoring port waterways do not fully consider the traffic flow coupling relationships between multiple waterways within a port. In actual operation, ships engage in complex interactions such as divergence, merging, and convergence between different waterways. Congestion in one waterway often propagates, spreads, and even amplifies throughout the waterway network, thus affecting the operational efficiency of the entire port waterway system. However, existing methods still have certain shortcomings in modeling the relationship between waterway network topology and ship flow direction. They struggle to fully capture the cross-waterway propagation characteristics of congestion within the waterway network, resulting in fragmented and one-sided monitoring results. This makes it difficult to comprehensively and accurately reflect the overall traffic situation of port waterways and to provide sufficient and effective decision support for cross-regional, multi-entity collaborative scheduling.

[0004] Vessel traffic flow in port channels is influenced by multiple factors, including vessel dynamics, environmental conditions, and operational scheduling. A single data source cannot comprehensively and accurately depict the channel's operational status. While Automatic Identification Systems (AIS) provide vessel position and movement information, they suffer from signal blind spots, data gaps, and equipment malfunctions. Shore-based radar can compensate for blind spots but lacks vessel identification and operational information. Hydrological and meteorological data directly affect channel navigation capacity and vessel behavior; without inclusion in the monitoring system, congestion assessments will deviate from reality. Terminal operation data reflects vessel entry and exit demands and berth occupancy, serving as crucial preliminary information for predicting traffic flow changes. Therefore, relying solely on a single data source cannot achieve accurate perception and trend prediction of channel congestion. In contrast, integrating multi-source heterogeneous data can improve the completeness and continuity of vessel trajectories through data complementarity and verification, enhance the quantification of the impact of complex environments, and provide richer input features for traffic flow models. This makes congestion identification more accurate, real-time, and forward-looking, laying a reliable data foundation for subsequent collaborative scheduling and overall traffic management.

[0005] Traditional early warning methods rely heavily on human experience and fixed thresholds, making it difficult to dynamically adapt to real-time changes in traffic flow and predict the spread of congestion in advance. This results in delayed early warning responses and insufficient window space for scheduling decisions. Consequently, they cannot provide comprehensive congestion warning information to multiple management entities within the port, leading to fragmented scheduling decisions, hindering coordinated traffic flow management, and failing to effectively alleviate port congestion as a whole.

[0006] To address this, the present invention proposes a port and waterway congestion monitoring and early warning method that integrates multi-source heterogeneous data to reduce the probability of port and waterway congestion and improve the overall port navigation efficiency. Summary of the Invention

[0007] This invention aims to solve the aforementioned problems in the prior art and provides a port and waterway congestion monitoring and early warning method that integrates multi-source heterogeneous data. It can be applied to waterway traffic situation perception, congestion risk prediction, and collaborative scheduling decision support in coastal hub ports.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: a method for monitoring and early warning of port and waterway congestion by integrating multi-source heterogeneous data, comprising the following steps: Step 1, Multi-source data acquisition and spatiotemporal fusion: Collect data from the port's Automatic Identification System (AIS), shore-based radar, hydrological and meteorological data, and dock production operation data. Clean, interpolate, and perform spatiotemporal registration on the collected multi-source heterogeneous data, and map it to a pre-constructed unified spatiotemporal grid of the port to form a standardized fusion dataset. Step 2, Port traffic flow feature extraction: Based on the standardized fusion dataset, calculate the dynamic traffic flow parameters of each waterway spatiotemporal grid unit within the port; the dynamic traffic flow parameters include ship traffic flow density, actual traffic volume, and basic waterway capacity; Step 3, construct a port collaborative dynamic model: based on the dynamic traffic flow parameters of each waterway spatiotemporal grid unit and the ship flow direction relationship between port waterways, establish a ship flow transfer equation to describe the evolution law of traffic flow in the waterway network; quantify the traffic flow interaction and congestion propagation mechanism between different waterways through the ship flow transfer equation. Step 4, Real-time Congestion Index Calculation: Based on the quantitative results of the ship flow transfer equation and the impact of real-time hydrological and meteorological conditions on the navigation environment of the waterway, the time-varying capacity of each waterway is dynamically corrected, and based on the corrected time-varying capacity and the current actual traffic flow, a dynamic congestion index that reflects the overall operation of the waterway network is calculated. Step 5, Tiered Early Warning Mechanism: Based on the dynamic congestion index, obtain the current congestion status early warning and the congestion spread trend early warning in the future time domain. Compare with multiple preset early warning level thresholds to trigger corresponding tiered early warning information. The tiered early warning information includes mild congestion warning, moderate congestion warning and severe congestion warning, and is pushed to multiple management entities within the port to support the port's traffic flow coordinated scheduling and guidance decisions.

[0009] The construction of the unified spatiotemporal grid for the port in step 1 is specifically as follows: based on the geographical scope of the port, a two-dimensional planar grid is divided using a preset spatial resolution; in the time dimension, continuous time is discretized into time windows using a preset time resolution; and the collected multi-source heterogeneous data is mapped to the corresponding spatiotemporal grid units through interpolation or matching algorithms.

[0010] The vessel traffic flow density is calculated by statistically analyzing the ratio of the number of vessels in each spatiotemporal grid cell to the area of ​​the spatiotemporal grid cell; the actual traffic flow is calculated by statistically analyzing the number of vessels entering the spatiotemporal grid cell per unit time; the basic channel capacity is calculated based on the effective channel width, average vessel length, safe speed, channel curvature reduction factor, and safety zone coefficient.

[0011] The ship flow transfer equation described in step 3 is used to describe the dynamic transfer process of ship flow in time and space between adjacent channel spatiotemporal grid cells and between different channels; specifically, it includes the following steps: S3.1. Channel Network Topology Construction: Based on the dynamic traffic flow parameters of each channel's spatiotemporal grid unit, port geographical layout data, and historical AIS trajectory data of vessels, the two-dimensional spatiotemporal grid units in the unified port spatiotemporal grid are indexed. Simplified to a single index; adjacency relationships of spatiotemporal grid cells are established based on clustering results of port geographic layout data and historical AIS trajectory data of ships, and each spatiotemporal grid cell is defined. upstream grid set and downstream mesh sets Furthermore, by analyzing and statistically analyzing the start and end points of historical AIS trajectory data of ships, the basic steering ratio is determined. The basic steering ratio is derived from the spatiotemporal grid cell. Flow to spatiotemporal grid cells Ships occupy spatiotemporal grid cells A fixed percentage of the total outflow; S3.2. Establishment of Ship Flow Conservation Equations: Based on the law of flow conservation, ship flow conservation equations are constructed for each spatiotemporal grid cell within adjacent time steps. The equation expressions are as follows: In the formula, spatiotemporal grid unit In the The number of ships at each time step spatiotemporal grid unit In the The number of ships at each time step For the first Each time step starts from the upstream spatiotemporal grid cell Flow to spatiotemporal grid cells Ship traffic volume, in units of vessels / hour. For the first Each time step from the spatiotemporal grid cell Downstream spatiotemporal grid cells Ship traffic volume, in units of vessels / hour. The preset time step; S3.3. Calculation of Ship Flow Transfer Between Grids: Determine the transmitting and receiving capacities of each spatiotemporal grid cell, and calculate the actual ship flow transfer between upstream and downstream spatiotemporal grid cells based on the basic steering ratio. Specifically, this includes: Sending capacity calculation, sending capacity spatiotemporal grid unit In the The maximum number of ships that can be sent downstream in a given time step is calculated using the following formula: In the formula, spatiotemporal grid unit The maximum outflow rate is taken as the basic channel capacity corresponding to the upstream spatiotemporal grid unit. Receive capability calculation, receive capability spatiotemporal grid unit In the The maximum number of ships that can be received from upstream at each time step is calculated using the following formula: In the formula, spatiotemporal grid unit In the The time-varying passage capacity of the waterway at each time step spatiotemporal grid unit The blockage density is equivalent to the maximum number of ships that a spatiotemporal grid cell can accommodate. The adjustment factor is set to 1. The actual transfer flow is calculated based on the principle of matching supply and demand between upstream and downstream areas. The calculation formula is as follows: Simultaneously satisfy the total outflow constraint of a single grid: In the formula, For the first Each time step from the spatiotemporal grid cell Flow to spatiotemporal grid cells The steering ratio, with the initial value being the base steering ratio. ; spatiotemporal grid unit In the The maximum number of ships that can be sent downstream in each time step; if the upstream sending demand exceeds the total downstream receiving capacity, the flow is allocated according to the proportion of each downstream receiving capacity. S3.4. Dynamic Correction of Steering Ratio: Based on the predicted congestion index of downstream spatiotemporal grid cells, the basic steering ratio is dynamically corrected to simulate the path selection behavior of ships in congested scenarios. The correction formula is as follows: In the formula, For downstream spatiotemporal grid units In the Predicted congestion index at each time step This is a sensitivity coefficient used to control how sensitive a ship's turning behavior is to congestion. Indicates from upstream spatiotemporal grid cell Downstream spatiotemporal grid cells The basic steering ratio, Represents downstream spatiotemporal grid cells In the Predicted congestion index at each time step; At the current moment Number of ships in each spatiotemporal grid unit As initial conditions, combined with spatiotemporal grid cells In the Time-varying channel traffic capacity at each time step According to time steps By recursively solving the ship flow conservation equations, the traffic flow state of each spatiotemporal grid cell in any future time period can be obtained. To quantify the interaction and congestion propagation mechanisms of traffic flow in future time periods.

[0012] The dynamic correction of the time-varying capacity of each waterway in step 4 is specifically as follows: based on the basic waterway capacity under standard conditions, a comprehensive capacity reduction coefficient calculated from real-time hydrological and meteorological data is introduced; the comprehensive capacity reduction coefficient takes into account factors such as wind speed, visibility, flow direction and flow velocity, and the corrected time-varying capacity is the product of the basic waterway capacity and the comprehensive reduction coefficient.

[0013] The dynamic congestion index mentioned in step 4 is the ratio of the current actual traffic flow to the corrected time-varying capacity of the waterway spatiotemporal grid unit. At the same time, based on the dynamic congestion index of each waterway spatiotemporal grid unit, the overall weighted average congestion index of the port is calculated in combination with the preset weights, which is used to characterize the overall operation status of the port waterway network.

[0014] The multiple preset warning level thresholds mentioned in step 5 are dynamically calibrated and optimized based on historical port congestion data and navigation events.

[0015] The congestion status warning at the current moment and the congestion spread trend warning in the future time domain mentioned in step 5 are based on the current traffic status as the initial condition, combined with future hydrological and meteorological forecast data, and rolling deduction according to the ship flow transfer equation to calculate the dynamic congestion index. The corresponding graded warning information is triggered by comparing it with multiple preset warning level thresholds.

[0016] The graded early warning information mentioned in step 5 is pushed to multiple management entities within the port, including the maritime vessel traffic service system, port dispatch center, pilot station, and port and shipping enterprise dispatch room, through a standardized data interface. At the same time, based on the electronic nautical chart, the spatiotemporal grid units of the waterway at different early warning levels are displayed with differentiated colors, and the trend of congestion spread is dynamically simulated and visualized.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. By constructing a unified spatiotemporal grid and a port collaborative dynamic model, this invention can effectively capture the traffic flow coupling relationship and congestion propagation characteristics between multiple waterways, realizing the transformation from single-point monitoring to networked global monitoring of ports and waterways, and the monitoring results are more comprehensive and closer to the actual operation status.

[0018] 2. This invention effectively solves the problems of insufficient data utilization and low monitoring accuracy in the prior art by deeply integrating multi-source heterogeneous data such as AIS, meteorology, and dock operations, and using ship flow transfer equations for dynamic simulation. It significantly improves the perception accuracy and real-time performance of waterway operation status in complex traffic flow environments.

[0019] 3. This invention uses the ship flow transfer equation to perform rolling deduction and calculate the dynamic congestion index to obtain the congestion status warning at the current moment and the congestion spread trend warning in the future time domain. This enables the early prediction of port and waterway congestion trends, reserves sufficient response time for scheduling decisions, and effectively improves the timeliness and foresight of the warning.

[0020] 4. The hierarchical early warning mechanism and its information push method for multiple management entities constructed by this invention provide unified decision-making information support for various ports and management departments within the port, which helps to break down information silos, realize the coordinated scheduling and overall guidance of port traffic flow, improve the overall navigation efficiency of the port, and reduce the probability of congestion. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall framework of the port and waterway congestion monitoring and early warning method that integrates multi-source heterogeneous data provided in this embodiment of the invention; Figure 2 This is a schematic diagram illustrating the specific implementation process of the port and waterway congestion monitoring and early warning method that integrates multi-source heterogeneous data provided in this embodiment of the invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the scope of protection of the present invention. Unless otherwise defined, the technical or scientific terms used herein should have the ordinary meaning understood by those skilled in the art to which this invention pertains.

[0023] The following steps detail the specific implementation process, with each step including clear inputs, processing logic, and output results to ensure the clarity and feasibility of the technical solution.

[0024] like Figure 1 As shown, a method for monitoring and early warning of port and waterway congestion by integrating multi-source heterogeneous data includes the following steps: Step 1, Multi-source data acquisition and spatiotemporal fusion: Collect data from the port's Automatic Identification System (AIS), shore-based radar, hydrological and meteorological data, and dock production operation data. Clean, interpolate, and perform spatiotemporal registration on the collected multi-source heterogeneous data, and map it to a pre-constructed unified spatiotemporal grid of the port to form a standardized fusion dataset. Step 2, Port traffic flow feature extraction: Based on the standardized fusion dataset, calculate the dynamic traffic flow parameters of each waterway spatiotemporal grid unit in the port. The dynamic traffic flow parameters include at least the ship traffic flow density, actual traffic volume, and basic waterway capacity. Step 3, construct a port collaborative dynamic model: based on the traffic flow parameters of each waterway spatiotemporal grid unit and the ship flow direction relationship between port waterways, establish a ship flow transfer equation to describe the evolution law of traffic flow in the waterway network, and quantify the traffic flow interaction and congestion propagation mechanism between different waterways through the ship flow transfer equation. Step 4, Real-time Congestion Index Calculation: Based on the quantitative results of the ship flow transfer equation and the impact of real-time hydrological and meteorological conditions on the navigation environment of the waterway, the time-varying capacity of each waterway is dynamically corrected, and based on the corrected time-varying capacity and the current actual traffic flow, a dynamic congestion index that reflects the overall operation of the waterway network is calculated. Step 5, Tiered Early Warning Mechanism: Based on the calculated dynamic congestion index and its predicted trend in the future time domain, and in accordance with multiple preset early warning level thresholds, corresponding tiered early warning information is triggered; the tiered early warning information includes at least mild congestion warning, moderate congestion warning, and severe congestion warning, and is pushed to multiple management entities within the port to support the port's traffic flow coordination scheduling and guidance decisions.

[0025] As a preferred embodiment of the present invention, the construction of the unified spatiotemporal grid for the port in step 1 is specifically as follows: according to the geographical spatial range of the port, a two-dimensional planar grid is divided using a preset spatial resolution; in the time dimension, continuous time is discretized into time windows using a preset time resolution; and the collected multi-source heterogeneous data is mapped to the corresponding spatiotemporal grid units through interpolation or matching algorithms.

[0026] As a preferred embodiment of the present invention, the ship flow transfer equation in step 3 is constructed based on cellular automata theory and is used to describe the dynamic transfer process of ship flow in time and space between adjacent spatiotemporal grid cells of waterways and between different waterways.

[0027] As a preferred embodiment of the present invention, the dynamic correction of the time-varying traffic capacity of each waterway in step 4 specifically involves: based on the basic traffic capacity of the waterway under standard conditions, introducing a traffic capacity reduction coefficient calculated from real-time hydrological and meteorological data, wherein the traffic capacity reduction coefficient considers at least wind speed, visibility, flow direction and flow velocity factors.

[0028] As a preferred embodiment of the present invention, the preset multiple warning level thresholds in step 5 are dynamically calibrated and optimized based on historical congestion data and navigation events; and the graded warning mechanism includes not only the congestion status warning at the current moment, but also the congestion spread trend warning within a future preset time window derived from the ship flow transfer equation.

[0029] like Figure 2As shown, the specific steps of the port and waterway congestion monitoring and early warning method that integrates multi-source heterogeneous data according to the present invention are as follows: Step 1: Multi-source Data Acquisition and Spatiotemporal Fusion. The inputs for this step are Automatic Identification System (AIS) data, shore-based radar data, hydrological and meteorological data, and terminal operation data within the port area. The processing logic involves accessing, preprocessing, and uniformly dividing the spatiotemporal grid and mapping the data across multiple sources. The output is a standardized fusion dataset based on the port's unified spatiotemporal grid. The core objective of this step is to address the data fusion problems existing in current technologies by unifying data from different sources, formats, and time frequencies under the same spatiotemporal reference, laying the foundation for subsequent traffic flow analysis. The specific implementation process is as follows: S1.1: Data Access and Preprocessing. First, data is accessed in real-time from various data sources within the port area via standardized interfaces to ensure data integrity and timeliness. AIS data includes static ship information (MMSI, ship name, length, beam, ship type, draft) and dynamic information (longitude, latitude, ground speed, heading, bow direction, UTC timestamp), and is the core source of ship traffic perception, with a data frequency typically ranging from a few seconds to a few minutes. Shore-based radar data provides the coordinates, speed, and heading of radar targets, primarily used to supplement AIS signal blind spots (such as channel bends, bridge areas, etc.) or missing time periods caused by AIS equipment malfunction or shutdown, ensuring full coverage of ship targets. Hydrological and meteorological data comes from port weather stations, buoys, radar, or numerical weather prediction models, including wind speed (…). ),wind direction( ),visibility( ), tide height ( ), flow rate ( ), flow direction ( Data such as berth occupancy status, planned berthing and departure times, and estimated vessel arrival times (ETA) are obtained from the operating systems (TOS) of each terminal to predict future vessel traffic flow and provide a basis for trend forecasting. Raw data often contains noise, missing data, and anomalies, therefore requiring cleaning: data with speeds exceeding the preset maximum speed (e.g., exceeding 30 knots), locations outside the port's geographical boundaries, and obviously incorrect timestamps are discarded; short-term missing AIS tracks (e.g., data lost due to signal obstruction) are filled using linear interpolation or Kalman filtering. For example, linear interpolation is used to fill in missing time points. The ship's position, if the forward and backward positions are known. and Then the interpolated coordinates are: This method ensures the continuity of ship trajectories and avoids omissions in subsequent grid statistics due to missing data. Meteorological data usually comes from discrete observation stations, but meteorological values ​​for each two-dimensional plane grid are required later. Therefore, Kriging interpolation or inverse distance weighting is used to interpolate the discrete station data to the center point of each subsequently divided two-dimensional plane grid, ensuring that each two-dimensional plane grid has corresponding meteorological and environmental data.

[0030] S1.2: Unified Spatiotemporal Grid Partitioning. To discretize continuous spatiotemporal space for easier computer processing, a two-dimensional geographic grid covering the entire waterway of the port is defined: Each spatiotemporal grid cell This corresponds to a real-world rectangular geographic area. The spatial resolution setting needs to balance computational accuracy and efficiency; a typical range is [range missing]. to This can be adjusted according to the channel width and ship density. In the time dimension, continuous time is discretized into time windows: For the time step, a typical value can be taken as follows: , or The choice of time step needs to balance real-time requirements and computational load.

[0031] S1.3: Spatiotemporal Mapping of Data. After gridding is completed, the preprocessed data needs to be mapped to the corresponding spatiotemporal grid cells. For each time window... Iterate through each ship's data, if the ship's time... satisfy Then, based on its latitude and longitude Determine the spatiotemporal grid cell Add the ship information to the spatiotemporal grid unit In the time window The list of ships. For hydro-meteorological data, the interpolated value of the center point of the grid is taken as the value of that grid in the list. Environmental conditions at any given moment. Ultimately, each spatiotemporal grid cell... Obtain a standardized fusion dataset: Ship List ; This completes the fusion of multi-source data into a unified spatiotemporal benchmark, providing standardized input for subsequent feature extraction.

[0032] Step 2: Port Traffic Flow Feature Extraction. The input to this step is the port unified spatiotemporal grid fusion dataset output from Step 1; the processing logic is to calculate the core traffic flow feature parameters of each waterway spatiotemporal grid unit based on the standardized fusion dataset; the output is the vessel traffic flow density, actual traffic volume, and basic waterway capacity of each spatiotemporal grid unit. The core objective of this step is to extract core parameters that characterize the waterway traffic state from the standardized fusion dataset, providing a foundation for subsequent model construction and congestion index calculation. The specific implementation process is as follows: S2.1: Traffic Flow Density Calculation. Traffic flow density reflects the density of ships within a region and is an important indicator for assessing congestion. Statistical Spatiotemporal Grid Units In the time window Number of ships inside The grid area is The formula for calculating traffic flow density is: The higher this value, the more ships there are per unit area, and the more congested the waterway. When the density exceeds a certain threshold, the mutual interference between ships will increase significantly, and the navigation risk will rise.

[0033] S2.2: Actual Traffic Flow Calculation. Actual traffic flow reflects the number of vessels passing through a cross-section or waterway per unit time, and is used to measure the busyness of the waterway. Actual Flow Defined as the number of ships passing through a cross-section of a spatiotemporal grid cell per unit time. For ease of calculation in the grid model, the cross-sectional statistical method is typically used. in In the time window The number of ships entering this spatiotemporal grid cell from upstream.

[0034] S2.3: Calculation of basic channel capacity. Basic capacity Basic navigation capacity refers to the maximum number of vessels that can safely pass through a waterway per unit time under standard environmental conditions (good visibility, no wind or waves, no traffic interference). It depends on the waterway's physical properties (width, depth, curvature) and the composition of the vessels, and is a relatively stable inherent property. Basic navigation capacity can be determined using the following methods: in, The effective width of the channel, The average length of ships passing through this waterway. For safe speed (usually the design speed of the waterway). To account for reduction factors such as channel curvature and intersections, Factors related to ship safety need to be considered. Specific parameters need to be calibrated through on-site observation and simulation experiments based on the actual conditions of the target port.

[0035] This step yields the density of each spatiotemporal grid cell at each time step. Actual traffic and basic traffic capacity This provides the foundational data for subsequent models.

[0036] Step 3: Construct a port collaborative dynamic model. The inputs to this step are the traffic flow parameters of each spatiotemporal grid unit output from Step 2, port geographical layout data, and historical AIS trajectory data of vessels. The processing logic involves constructing the waterway network topology, establishing vessel flow conservation equations, calculating flow transfer values, and dynamically adjusting turning ratios. The output is the traffic flow projection data for each grid in future time periods. This step aims to address the shortcomings of existing technologies in capturing the coupling and correlation of traffic flows between waterway networks. By establishing a waterway network traffic flow evolution model, it quantifies the interaction and congestion propagation mechanisms between different waterways and different ports. This invention draws on the idea of ​​the one-dimensional cellular transport model (CTM), treating each waterway grid as a cell, and using conservation equations and flow transfer rules to project the spatiotemporal evolution of traffic flow. The specific implementation process is as follows: S3.1: Channel Network Topology Construction. First, the connectivity between grid cells needs to be defined. Based on the port's geographical layout and actual ship navigation habits (obtainable through historical AIS trajectory clustering), the adjacency relationships between grid cells are established. To simplify the representation, the two-dimensional grid index is simplified to a single index. Each spatiotemporal grid cell is defined. upstream spatiotemporal grid cell set and downstream spatiotemporal grid cell set For channel intersections or forks, the turning ratio needs to be defined. , indicating from spatiotemporal grid cell Flow to spatiotemporal grid cells Ships occupy spatiotemporal grid cells The proportion of total outflow. The turning ratio can be obtained by analyzing and statistically analyzing historical trajectory OD (origin and destination points).

[0037] S3.2: Ship Flow Conservation Equation. Based on the law of conservation of flow, the change in the number of ships within each spatiotemporal grid cell should equal the inflow minus the outflow. For each spatiotemporal grid cell... In time step arrive Between these points, the number of ships meets the following requirements: in, For upstream spatiotemporal grid cells Flow to spatiotemporal grid cells Flow rate (unit: ships / h) To obtain from spatiotemporal grid cells Flowing downstream spatiotemporal grid cells The equation forms the basis of the model, ensuring that the total number of ships is conserved throughout the network.

[0038] S3.3: Flow Transfer Calculation. The conservation equations alone are insufficient; the actual transferred flow rate also needs to be determined. The size depends on the transmitting capability of the upstream spatiotemporal grid cell and the receiving capability of the downstream spatiotemporal grid cell. Transmitting capability Spatiotemporal grid cell At time step The maximum number of ships that can be sent downstream is limited by the current number of ships within the spatiotemporal grid cell and the maximum outflow rate of the spatiotemporal grid cell itself. in spatiotemporal grid unit The maximum possible outflow rate is usually taken as the basic throughput capacity. Reception capability Spatiotemporal grid cell At time step The maximum number of vessels that can be received from upstream is limited by the remaining capacity of the downstream spatiotemporal grid cells and the time-varying traffic capacity of the downstream waterway: in This is the time-varying capacity that will be calculated in step 4; This refers to the grid blocking density, which is the maximum number of ships the grid can accommodate. The adjustment factor is typically set to 1. Actual transfer flow: for each pair of upstream and downstream spatiotemporal grid cells. The actual transferred traffic is the minimum value of the upstream transmitting capacity multiplied by the diversion ratio and the downstream receiving capacity, reflecting the principle of supply and demand matching. At the same time, a constraint must also be satisfied: from the spatiotemporal grid cell The total flow of traffic to all downstream spatiotemporal grid cells cannot exceed their total transmission capacity: If the upstream transmission demand exceeds the total downstream receiving capacity, the demand must be allocated according to the proportion of each downstream receiving capacity to avoid exceeding the downstream capacity.

[0039] S3.4: Dynamic Adjustment of Steering Ratio (Optional). In actual navigation, when severe congestion occurs in a downstream channel, some vessels may choose alternative routes. To simulate this driver behavior, the steering ratio can be dynamically adjusted. For example, the adjusted steering ratio can be defined as: in, Based on the proportion of basic shifts (historical statistics). The predicted congestion index for downstream spatiotemporal grid cells (which can be the model value from the previous time step or the current real-time value). The sensitivity coefficient controls how well steering behavior responds to congestion. This formula uses a softmax function to reduce the proportion of traffic allocated to downstream directions with high congestion levels, thus simulating path selection behavior.

[0040] By combining the above conservation equations and the formula for calculating the transfer flow, and then iteratively solving the problem, the traffic flow status of each grid at multiple future time steps can be deduced, making it possible to predict the spread trend of congestion.

[0041] Step 4: Real-time calculation of congestion index; The inputs to this step are the basic traffic flow parameters output from step 2, the traffic flow data output from step 3, and the real-time hydrological and meteorological data from step 1. The processing logic is to correct the time-varying navigation capacity of the waterway based on the real-time hydrological and meteorological data, and calculate the dynamic congestion index in combination with the actual traffic flow. The outputs are the dynamic congestion index of each spatiotemporal grid unit and the overall weighted average congestion index of the port. The core objective of this step is to transform the traffic flow status into an intuitive and quantifiable congestion indicator, while considering the impact of real-time environmental factors on navigation capacity, making the congestion assessment closer to the actual navigation situation. The specific implementation process is as follows: S4.1: Time-Varying Capacity Correction. The basic capacity calculated in Step 2 is a theoretical value under standard conditions. However, in actual navigation, hydrological and meteorological conditions significantly affect the channel's capacity. Therefore, it is necessary to reduce the basic capacity based on real-time meteorological data. Define a comprehensive reduction factor. It is a function of the reduction factors for each environmental factor. Calculate the individual reduction factors separately: wind speed reduction factor. ,in Critical wind speed; visibility reduction factor ,in Minimum visibility required for normal navigation; flow direction and velocity reduction factor ,in For the direction of the waterway, The critical flow velocity is given. The overall reduction factor can be taken as the minimum value of each individual coefficient: Then the spatiotemporal grid unit At time step The time-varying passage capability is: .

[0042] S4.2: Dynamic Congestion Index Calculation. With actual traffic flow and time-varying capacity, the congestion index can be calculated. This invention defines the congestion index. This is the ratio of current traffic load to capacity, reflecting the saturation level of the waterway. in This represents the current actual traffic flow (which can be obtained directly from step 2 or extracted from the model extrapolation results in step 3). This index is dimensionless, and its value has a clear physical meaning: when... When, it indicates that the waterway is unobstructed; when When, it indicates mild congestion; when When, it indicates moderate congestion; when A value of 1 indicates severe congestion. The above thresholds are for illustrative purposes only; specific values ​​need to be determined based on historical data and the actual conditions of the port. To assess the overall congestion level of the port, a weighted average congestion index can be calculated: Weight This indicator can be set based on factors such as waterway grade and historical traffic volume, and can be used to provide port management with a quick overview of the overall operational status.

[0043] Step 5: Tiered early warning mechanism; The inputs to this step are the single-grid dynamic congestion index output from step 4, the port's overall weighted average congestion index, and the congestion trend projection data output from step 3. The processing logic involves setting dynamic early warning thresholds, triggering real-time status early warnings, issuing trend prediction early warnings, and pushing early warning information through multiple channels. The outputs are tiered early warning information, early warning visualization results, and early warning reports. The core objective of this step is to transform the calculated congestion index into actionable early warning information and release it to multiple port management entities, providing decision support for collaborative scheduling and solving the problem of delayed early warnings in existing technologies. The specific implementation process is as follows: S5.1: Warning Level Threshold Setting. Based on historical port operation data and operational experience, a three-level warning threshold is set. The threshold setting needs to consider historical congestion events, accident data, and management requirements. Example values ​​are as follows: Mild Congestion Threshold moderate congestion threshold Severe congestion threshold These thresholds are not static; the system should have dynamic optimization capabilities. For example, it should recalculate the quantiles of the congestion index quarterly based on the latest historical data, using these quantiles as new thresholds to make the warnings more aligned with actual operational patterns.

[0044] S5.2: Real-time status alerts. At each time step... For each spatiotemporal grid cell Calculate the current congestion index And compare it with the threshold: if This triggers a mild congestion warning; if This triggers a moderate congestion warning; if This triggers a severe congestion warning. Each warning should include detailed information: grid location (e.g., waterway name, latitude and longitude range), congestion index value, warning level, time of occurrence, potential impact area (based on model simulation results), and recommended measures (e.g., speed limits, off-peak travel, suspension of some waterway operations, etc.).

[0045] S5.3: Trend Prediction and Early Warning. This is a significant improvement of the present invention compared to traditional methods, achieving a leap from post-event identification to pre-event prediction. Utilizing the collaborative dynamic model established in step 3, with the current time... Using the current traffic conditions as initial conditions and combining them with weather forecast data for future periods, the model is continuously simulated to predict future traffic conditions. time steps (e.g.) For the next hour, if Congestion index During the simulation, the system continuously checks the predicted value for each future moment. If it detects a future moment... of If a certain warning threshold is exceeded, but a real-time warning of that level has not yet been triggered at the current moment, the system will issue a trend warning in advance. The lead time for trend warnings can be dynamically adjusted based on the reliability of the prediction model. This proactive warning provides a valuable response window for scheduling decisions, enabling managers to take preventative measures before congestion occurs.

[0046] S5.4: Warning Information Release and Visualization. Warning information needs to be delivered to decision-makers through effective channels and presented in an intuitive manner. For information dissemination, standardized data interfaces (such as Web services and message queues) are used to push warning information in real time to the systems of various port management entities, including the Maritime VTS (Vessel Traffic Service), port dispatch center, pilot station, and port and shipping company dispatch rooms. For visualization, different colors are used to highlight congestion grids on electronic charts or the large screen system of the dispatch and command center; for example, green represents unobstructed traffic (…). Yellow indicates mild congestion. Orange indicates moderate congestion. Red indicates severe congestion. For trend warnings, the system can use flashing or gradient colors to display the future evolution of congestion, helping managers intuitively understand the propagation path and trend of congestion. Regarding report generation, the system also supports the automatic generation of warning reports, including text descriptions, statistical charts, and screenshots of the warning area, facilitating archiving and post-event analysis.

[0047] This specific implementation method achieves real-time monitoring and early warning of port and waterway congestion through the above steps: Step 1 ensures data spatiotemporal consistency; Step 2 extracts key traffic flow indicators; Step 3 constructs a port collaborative dynamic model; Step 4 calculates the dynamic congestion index; Step 5 triggers tiered early warnings based on predicted trends, forming a closed loop of data collection, fusion, extraction, modeling, calculation, and early warning, thus realizing continuous monitoring and dynamic early warning of port and waterway congestion. This method avoids the limitations of existing technologies, significantly improves the timeliness and accuracy of port traffic flow collaborative scheduling, and provides feasible technical support for intelligent traffic management platforms.

[0048] It should be noted that the specific values ​​of formulas, parameters, and thresholds involved in the above specific embodiments are merely illustrative. In practical applications, appropriate parameter values ​​should be determined through experiments or calibration based on the actual geographical environment, waterway conditions, ship composition, and historical operational data of the target port. Those skilled in the art can easily conceive of adjustments or optimizations within the technical scope disclosed in this invention, but all such changes should be covered within the protection scope of this invention.

[0049] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for monitoring and early warning of port and waterway congestion by integrating multi-source heterogeneous data, characterized in that, Includes the following steps: Step 1, Multi-source data acquisition and spatiotemporal fusion: Collect data from the port's Automatic Identification System (AIS), shore-based radar, hydrological and meteorological data, and dock production operation data. Clean, interpolate, and perform spatiotemporal registration on the collected multi-source heterogeneous data, and map it to a pre-constructed unified spatiotemporal grid of the port to form a standardized fusion dataset. Step 2, Port traffic flow feature extraction: Based on the standardized fusion dataset, calculate the dynamic traffic flow parameters of each waterway spatiotemporal grid unit within the port; the dynamic traffic flow parameters include ship traffic flow density, actual traffic volume, and basic waterway capacity; Step 3, construct a port collaborative dynamic model: based on the dynamic traffic flow parameters of each waterway spatiotemporal grid unit and the ship flow direction relationship between port waterways, establish a ship flow transfer equation to describe the evolution law of traffic flow in the waterway network; quantify the traffic flow interaction and congestion propagation mechanism between different waterways through the ship flow transfer equation. Step 4, Real-time Congestion Index Calculation: Based on the quantitative results of the ship flow transfer equation and the impact of real-time hydrological and meteorological conditions on the navigation environment of the waterway, the time-varying capacity of each waterway is dynamically corrected, and based on the corrected time-varying capacity and the current actual traffic flow, a dynamic congestion index that reflects the overall operation of the waterway network is calculated. Step 5, graded early warning mechanism: Based on the dynamic congestion index, obtain the congestion status early warning at the current moment and the congestion spread trend early warning in the future time domain, and trigger the corresponding graded early warning information by comparing with multiple preset early warning level thresholds; The tiered early warning information includes mild congestion warnings, moderate congestion warnings, and severe congestion warnings, and is pushed to multiple management entities within the port to support the port's coordinated traffic flow scheduling and management decisions.

2. The port and waterway congestion monitoring and early warning method integrating multi-source heterogeneous data according to claim 1, characterized in that, The construction of the unified spatiotemporal grid for the port in step 1 is specifically as follows: based on the geographical scope of the port, a two-dimensional planar grid is divided using a preset spatial resolution; in the time dimension, continuous time is discretized into time windows using a preset time resolution; and the collected multi-source heterogeneous data is mapped to the corresponding spatiotemporal grid units through interpolation or matching algorithms.

3. The port and waterway congestion monitoring and early warning method according to claim 2, characterized in that, The vessel traffic flow density is calculated by statistically analyzing the ratio of the number of vessels in each spatiotemporal grid cell to the area of ​​the spatiotemporal grid cell; the actual traffic flow is calculated by statistically analyzing the number of vessels entering the spatiotemporal grid cell per unit time; the basic channel capacity is calculated based on the effective channel width, average vessel length, safe speed, channel curvature reduction factor, and safety zone coefficient.

4. The port and waterway congestion monitoring and early warning method according to claim 1, characterized in that, The ship flow transfer equation described in step 3 is used to describe the dynamic transfer process of ship flow in time and space between adjacent channel spatiotemporal grid cells and between different channels; specifically, it includes the following steps: S3.

1. Channel Network Topology Construction: Based on the dynamic traffic flow parameters of each channel's spatiotemporal grid unit, port geographical layout data, and historical AIS trajectory data of vessels, the two-dimensional spatiotemporal grid units in the unified port spatiotemporal grid are indexed. Simplified to a single index; adjacency relationships of spatiotemporal grid cells are established based on clustering results of port geographic layout data and historical AIS trajectory data of ships, and each spatiotemporal grid cell is defined. upstream grid set and downstream mesh sets Furthermore, by analyzing and statistically analyzing the start and end points of historical AIS trajectory data of ships, the basic steering ratio is determined. ; The basic steering ratio is derived from the spatiotemporal grid unit. Flow to spatiotemporal grid cells Ships occupy spatiotemporal grid cells A fixed percentage of the total outflow; S3.

2. Establishment of Ship Flow Conservation Equations: Based on the law of flow conservation, ship flow conservation equations are constructed for each spatiotemporal grid cell within adjacent time steps. The equation expressions are as follows: In the formula, spatiotemporal grid unit In the The number of ships at each time step spatiotemporal grid unit In the The number of ships at each time step For the first Each time step starts from the upstream spatiotemporal grid cell Flow to spatiotemporal grid cells Ship traffic volume, in units of vessels / hour. For the first Each time step from the spatiotemporal grid cell Downstream spatiotemporal grid cells Ship traffic volume, in units of vessels / hour. The preset time step; S3.

3. Calculation of Ship Flow Transfer Between Grids: Determine the transmitting and receiving capacities of each spatiotemporal grid cell, and calculate the actual ship flow transfer between upstream and downstream spatiotemporal grid cells based on the basic steering ratio. Specifically, this includes: Sending capacity calculation, sending capacity spatiotemporal grid unit In the The maximum number of ships that can be sent downstream in a given time step is calculated using the following formula: In the formula, spatiotemporal grid unit The maximum outflow rate is taken as the basic channel capacity corresponding to the upstream spatiotemporal grid unit. Receive capability calculation, receive capability spatiotemporal grid unit In the The maximum number of ships that can be received from upstream at each time step is calculated using the following formula: In the formula, spatiotemporal grid unit In the The time-varying passage capacity of the waterway at each time step spatiotemporal grid unit The blockage density is equivalent to the maximum number of ships that a spatiotemporal grid cell can accommodate. The adjustment factor is set to 1. The actual transfer flow is calculated based on the principle of matching supply and demand between upstream and downstream areas. The calculation formula is as follows: Simultaneously satisfy the total outflow constraint of a single grid: In the formula, For the first Each time step from the spatiotemporal grid cell Flow to spatiotemporal grid cells The steering ratio, with the initial value being the base steering ratio. ; spatiotemporal grid unit In the The maximum number of ships that can be sent downstream in each time step; if the upstream sending demand exceeds the total downstream receiving capacity, the flow is allocated according to the proportion of each downstream receiving capacity. S3.

4. Dynamic Correction of Steering Ratio: Based on the predicted congestion index of downstream spatiotemporal grid cells, the basic steering ratio is dynamically corrected to simulate the path selection behavior of ships in congested scenarios. The correction formula is as follows: In the formula, For downstream spatiotemporal grid units In the Predicted congestion index at each time step This is a sensitivity coefficient used to control how sensitive a ship's turning behavior is to congestion. Indicates from upstream spatiotemporal grid cell Downstream spatiotemporal grid cells The basic steering ratio, Represents downstream spatiotemporal grid cells In the Predicted congestion index at each time step; At the current moment Number of ships in each spatiotemporal grid unit As initial conditions, combined with spatiotemporal grid cells In the Time-varying channel traffic capacity at each time step According to time steps By recursively solving the ship flow conservation equations, the traffic flow state of each spatiotemporal grid cell in any future time period can be obtained. To quantify the interaction and congestion propagation mechanisms of traffic flow in future time periods.

5. The port and waterway congestion monitoring and early warning method integrating multi-source heterogeneous data according to claim 1, characterized in that, The dynamic correction of the time-varying capacity of each waterway in step 4 is specifically as follows: based on the basic waterway capacity under standard conditions, a comprehensive capacity reduction coefficient calculated from real-time hydrological and meteorological data is introduced; the comprehensive capacity reduction coefficient takes into account factors such as wind speed, visibility, flow direction and flow velocity, and the corrected time-varying capacity is the product of the basic waterway capacity and the comprehensive reduction coefficient.

6. The port and waterway congestion monitoring and early warning method according to claim 1, characterized in that, The dynamic congestion index mentioned in step 4 is the ratio of the current actual traffic flow to the corrected time-varying capacity of the waterway spatiotemporal grid unit. At the same time, based on the dynamic congestion index of each waterway spatiotemporal grid unit, the overall weighted average congestion index of the port is calculated in combination with the preset weights, which is used to characterize the overall operation status of the port waterway network.

7. The port and waterway congestion monitoring and early warning method according to claim 1, characterized in that, The multiple preset warning level thresholds mentioned in step 5 are dynamically calibrated and optimized based on historical port congestion data and navigation events.

8. The port and waterway congestion monitoring and early warning method according to claim 1, characterized in that, The congestion status warning at the current moment and the congestion spread trend warning in the future time domain mentioned in step 5 are based on the current traffic status as the initial condition, combined with future hydrological and meteorological forecast data, and rolling deduction according to the ship flow transfer equation to calculate the dynamic congestion index. The corresponding graded warning information is triggered by comparing it with multiple preset warning level thresholds.

9. The port and waterway congestion monitoring and early warning method according to claim 1, characterized in that, The graded early warning information mentioned in step 5 is pushed to multiple management entities within the port, including the maritime vessel traffic service system, port dispatch center, pilot station, and port and shipping enterprise dispatch room, through a standardized data interface. At the same time, based on the electronic nautical chart, the spatiotemporal grid units of the waterway at different early warning levels are displayed with differentiated colors, and the trend of congestion spread is dynamically simulated and visualized.

Images