A shipping index board analysis method and system based on matrix mapping

Abstract

The application provides a shipping index board analysis method and system based on matrix mapping, first, ship and port information is acquired, each port is classified into a node dimension of a country, a trade zone or a continent based on latitude and longitude or a country to which the port belongs, and multi-dimensional attribution information of the port is constructed; then, ship AIS dynamic data is acquired in real time, and sailing records between ports are generated by processing; and a user-defined node pair matrix and analysis time range and statistical period are received, a mapping relationship between a starting node, a destination node and rows and columns of the matrix is established, and the row and column nodes of the matrix are analyzed into specific port sets; based on the unified analysis time range and the statistical period, sailing records of all cells in the matrix are batch-screened and periodically summarized, and complete and time-aligned index time series are generated for each cell; time sequence index curves of multiple directions are displayed in parallel on the same board interface according to the row and column layout of the matrix, and comparison analysis in the horizontal direction (same destination-multiple starting points) and the vertical direction (same starting point-multiple destinations) is realized.

Description

Technical Field

[0001] This invention relates to the fields of shipping informatization and intelligentization, specifically to a shipping indicator dashboard analysis method and system based on matrix mapping. Background Technology

[0002] The Automatic Identification System (AIS) is a ship navigation system implemented using wireless communication and computer technology. AIS employs a globally unique MMSI code as its identification method. Each ship is assigned a globally unique MMSI code from the start of construction to its dismantling. In conjunction with the Global Positioning System (GPS), AIS broadcasts dynamic ship data such as speed, position, rate of change of course, and course, along with static ship data such as name, call sign, draft, and dangerous cargo information, via VHF channels to nearby vessels and shore stations. This allows neighboring vessels and shore stations to promptly obtain information on the movements of all ships in the vicinity. AIS data is the collective term for all data transmitted by a ship each time it is used.

[0003] With the rapid development of global trade, the shipping industry handles approximately 90% of international trade volume. In-depth analysis of shipping data, particularly the monitoring and forecasting of cargo flows between ports, is crucial for shipping companies to optimize route planning, port operators to plan infrastructure construction, traders to assess market supply and demand, and regulatory agencies to understand trade dynamics. However, the massive amounts of AIS data, initially designed primarily for navigation safety, have complex structures and enormous volumes (generating massive position updates every second). Extracting commercially valuable shipping indicators from this raw data has become a pressing technical challenge in this field.

[0004] Currently, the analysis of AIS data mainly focuses on the statistical level of a single flow direction, such as counting the number of ships or deadweight tonnage on the "China-US" route. This single-flow analysis mode has the following technical shortcomings:

[0005] 1) Low analytical efficiency: When users need to compare cargo volume trends across multiple flows (e.g., from major Asian ports to major European ports), they must repeatedly configure query conditions and execute statistical tasks multiple times, making "one-stop" multi-flow comparative analysis impossible. 2) Inconsistent data structure: Statistical results for different flows are often based on different time ranges and statistical periods, resulting in data that cannot be strictly aligned on the timeline, making horizontal comparative analysis difficult. 3) Difficulty in anomaly identification: Single-flow analysis cannot quickly identify abnormal fluctuations and struggles to discover structural changes in trade flows. 4) Poor visualization: Multi-flow data is often presented in multiple independent charts, requiring users to repeatedly switch between, memorize, and compare charts, resulting in a heavy cognitive burden and making it difficult to form an overall judgment. Summary of the Invention

[0006] To address the problems of low efficiency in multi-direction shipping indicator analysis, inability to unify data comparison, and poor visualization in existing technologies, this invention provides a shipping indicator dashboard analysis method based on matrix mapping. This method is adaptable to complex shipping networks and massive AIS data application scenarios, featuring high analysis efficiency, strong multi-dimensional data comparability, and intuitive visualization. By constructing a multi-dimensional port attribution system, generating standardized inter-port navigation records, defining a node pair matrix with mapping relationships, and batch statistically obtaining time series of indicators for each flow direction, the method achieves synchronous comparison of multi-direction data through matrix-based data organization and visualization layout. This significantly improves the efficiency and intuitiveness of shipping data analysis, while efficiently and accurately displaying multi-direction cargo volume trends, providing reliable data support for shipping companies' route optimization, port operators' planning and construction, and traders' market analysis. This invention also relates to a shipping indicator dashboard analysis system based on matrix mapping.

[0007] The technical solution of the present invention is as follows:

[0008] A shipping indicator dashboard analysis method based on matrix mapping, characterized by the following steps:

[0009] The steps for constructing multidimensional port attribution information are as follows: First, acquire vessel information and port information. The vessel information includes the vessel's MMSI code and deadweight tonnage. The port information includes the port's latitude and longitude coordinates and the country to which the port belongs. Based on the port's latitude and longitude coordinates and the country to which the port belongs, each port is classified into the corresponding country, trade zone, or continent node dimension, thereby constructing multidimensional port attribution information.

[0010] The steps for generating inter-port navigation records are as follows: Real-time acquisition of vessel AIS dynamic data; based on the port latitude and longitude coordinates in the port multidimensional attribution information, voyage identification and association are performed on the AIS dynamic data of each vessel; complete voyage information from departure from the port of origin to berthing at the destination port is extracted to generate an inter-port navigation record; the inter-port navigation record includes the vessel's MMSI code, port of origin identifier, departure timestamp, destination port identifier, arrival timestamp, and deadweight tonnage;

[0011] The steps for defining a node pair matrix are as follows: Obtain the node pair matrix defined by the user on the visualization dashboard and the statistical parameters input by the user. The statistical parameters include the analysis time range and the statistical period. The matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the ending point of the analysis. Each cell of the node pair matrix corresponds to the starting node in the column and the destination node in the row, thereby establishing a mapping relationship between starting nodes, destination nodes, and matrix rows and columns.

[0012] Port set construction steps: For each cell in the node pair matrix, based on the multi-dimensional port affiliation information, query all ports belonging to the originating node in the column of that cell to form an originating port set; and query all ports belonging to the destination node in the row of that cell to form a destination port set.

[0013] Indicator time series generation steps: Based on the analysis time range and statistical period, perform the following operations on each cell in the node pair matrix:

[0014] The analysis time range is divided into a continuous periodic sequence according to the statistical period;

[0015] From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell;

[0016] The basic record set is periodically summarized according to the periodic sequence, and the deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value of that period is set to zero, thereby generating the index time series of that cell.

[0017] Steps for generating and displaying time series indicator curves: Organize the time series indicators of all cells according to their row and column positions in the node pair matrix to form the time series indicator curves of the corresponding cells; On the visualization dashboard interface, display the time series indicator curves corresponding to multiple cells in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node but different origin nodes for easy horizontal comparison; cells in the same column of the matrix correspond to the same origin node but different destination nodes for easy vertical comparison, so as to realize the comparative analysis of multi-direction shipping indicators corresponding to multiple cells in the matrix.

[0018] Preferably, in the port multidimensional attribution information construction step, berth information is also obtained, including berth name, port to which the berth belongs, and types of cargo that can be loaded and unloaded at each berth; in the port navigation record generation step, the port navigation record also includes the originating berth identifier, the destination berth identifier, and the empty / full load status; in the node pair matrix definition step, the statistical parameters also include cargo types.

[0019] In the step of generating the time series of indicators, after filtering out the basic record set, the method further includes: determining whether the statistical parameters input by the user contain cargo types; if so, filtering out and retaining records from the basic record set where both the originating berth and the destination berth can load and unload the cargo types based on the berth information; determining whether the statistical parameters input by the user contain empty / full load status; if so, filtering out and retaining records from the basic record set whose empty / full load status matches the empty / full load status specified by the user; and performing the periodic summary based on the navigation records obtained from the final filtering.

[0020] Preferably, in the indicator time series generation step, the generated indicator time series is further smoothed, and the smoothing process includes:

[0021] Get the number of moving average window periods N input by the user; for each cell's index time series, start from the starting period of the index time series and traverse it. Calculate the arithmetic mean based on the cumulative deadweight tonnage of the current period and the N-1 consecutive periods before it, and update the cumulative deadweight tonnage of the current period with this arithmetic mean, thereby generating a smoothed index time series.

[0022] Preferably, in the port navigation record generation step, the AIS dynamic data includes the ship's draft, real-time location, timestamp, and navigation status;

[0023] The empty / full load status is determined based on the comparison between the ship's current draft and its empty and full load drafts: if the current draft is greater than or equal to the full load draft, the empty / full load status is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load status is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load status is determined to be half load.

[0024] Preferably, in the node pair matrix definition step, the statistical period includes day, week, month, quarter, and year.

[0025] A shipping indicator dashboard analysis system based on matrix mapping is characterized by comprising, in sequence, a port multidimensional attribution information construction module, a port navigation record generation module, a node pair matrix definition module, a port set construction module, an indicator time series generation module, and a time series indicator curve generation and display module.

[0026] The port multidimensional attribution information construction module acquires ship information and port information; the ship information includes the ship's MMSI code and deadweight tonnage; the port information includes the port's latitude and longitude coordinates and the country to which the port belongs; based on the port's latitude and longitude coordinates and the country to which the port belongs, each port is classified into the corresponding country, trade zone or continent node dimension, thereby constructing the port multidimensional attribution information;

[0027] The inter-port navigation record generation module acquires the AIS dynamic data of ships in real time. Based on the port latitude and longitude coordinates in the port multidimensional attribution information, it identifies and associates the voyage data of each ship's AIS dynamic data, extracts the complete voyage information of the ship from departure from the port of origin to berthing at the port of destination, and generates an inter-port navigation record. The inter-port navigation record includes the ship's MMSI code, port of origin identifier, departure timestamp, port of destination identifier, arrival timestamp, and deadweight tonnage.

[0028] The node pair matrix definition module obtains the node pair matrix defined by the user on the visualization dashboard and the statistical parameters input by the user. The statistical parameters include the analysis time range and the statistical period. The matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the ending point of the analysis. Each cell of the node pair matrix corresponds to the starting node in the column and the destination node in the row, thereby establishing a mapping relationship between the starting nodes, destination nodes and the matrix rows and columns.

[0029] The port set construction module, for each cell in the node pair matrix, queries all ports belonging to the originating node in the column of the cell according to the multidimensional port affiliation information, to form an originating port set; and queries all ports belonging to the destination node in the row of the cell, to form a destination port set.

[0030] The indicator time series generation module performs the following operations on each cell in the node pair matrix based on the analysis time range and statistical period:

[0031] The analysis time range is divided into a continuous periodic sequence according to the statistical period;

[0032] From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell;

[0033] The basic record set is periodically summarized according to the periodic sequence, and the deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value of that period is set to zero, thereby generating the index time series of that cell.

[0034] The time-series indicator curve generation and display module organizes the time series indicators of all cells according to their row and column positions in the node pair matrix, forming the time-series indicator curves of the corresponding cells. On the visualization dashboard interface, the time-series indicator curves corresponding to multiple cells are displayed in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node but different origin nodes, which facilitates horizontal comparison. Cells in the same column of the matrix correspond to the same origin node but different destination nodes, which facilitates vertical comparison, so as to realize the comparative analysis of multi-direction shipping indicators corresponding to multiple cells in the matrix.

[0035] Preferably, the port multidimensional attribution information construction module also acquires berth information, which includes berth name, port to which the berth belongs, and types of cargo that can be loaded and unloaded at each berth; in the port navigation record generation module, the port navigation record also includes the originating berth identifier, the destination berth identifier, and the empty / full load status; in the node pair matrix definition module, the statistical parameters also include cargo types.

[0036] The indicator time series generation module, after filtering out the basic record set, further includes: determining whether the statistical parameters input by the user contain cargo types; if so, filtering out and retaining records from the basic record set where both the originating berth and the destination berth can load and unload the cargo type based on the berth information; determining whether the statistical parameters input by the user contain empty / full load status; if so, filtering out and retaining records from the basic record set whose empty / full load status matches the empty / full load status specified by the user; and performing the periodic summary based on the finally filtered navigation records.

[0037] Preferably, the indicator time series generation module further performs smoothing processing on the generated indicator time series, the smoothing processing including:

[0038] Get the number of moving average window periods N input by the user; for each cell's index time series, start from the starting period of the index time series and traverse it. Calculate the arithmetic mean based on the cumulative deadweight tonnage of the current period and the N-1 consecutive periods before it, and update the cumulative deadweight tonnage of the current period with this arithmetic mean, thereby generating a smoothed index time series.

[0039] Preferably, in the port navigation record generation module, the AIS dynamic data includes the ship's draft, real-time location, timestamp, and navigation status;

[0040] The empty / full load status is determined based on the comparison between the ship's current draft and its empty and full load drafts: if the current draft is greater than or equal to the full load draft, the empty / full load status is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load status is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load status is determined to be half load.

[0041] Preferably, in the node pair matrix definition module, the statistical period includes day, week, month, quarter, and year.

[0042] The beneficial effects of this invention are as follows:

[0043] This invention provides a matrix mapping-based shipping indicator dashboard analysis method. It constructs a multi-dimensional port attribution system, generates standardized voyage records (standardized inter-port navigation records), defines a node pair matrix with mapping relationships, and batch-generates time series of various flow direction indicators. These are then uniformly displayed on a single dashboard interface according to a matrix layout, enabling efficient comparative analysis of multi-flow direction shipping indicators. First, vessel and port information is acquired. Based on latitude and longitude or country of origin, each port is categorized into nodes representing countries, trade zones, or continents, thus constructing a multi-dimensional port attribution system and obtaining multi-dimensional port attribution information. This allows users to select specific ports as analysis nodes, or macro-level trade zones or continents, achieving multi-granularity analysis capabilities from micro to macro levels, overcoming the technical limitation of traditional methods that can only analyze single-port flow directions.

[0044] Then, by acquiring real-time ship AIS dynamic data, and based on the port latitude and longitude coordinates in the port multidimensional attribution information, the AIS dynamic data of each ship is used to identify and associate voyages, extract the complete voyage information of the ship from departure from the port of origin to berthing at the port of destination, and generate inter-port navigation records. This transforms massive, disordered raw AIS point data into voyage records with clear business significance, greatly reducing data redundancy, improving statistical efficiency, and ensuring the accuracy of statistical results.

[0045] Furthermore, the node pair matrix defined by the user on the visualization dashboard is obtained. This matrix has multiple originating nodes arranged in its columns and multiple destination nodes arranged in its rows. Each cell corresponds to a single originating-destination node combination to be analyzed, thus establishing a mapping relationship between originating nodes, destination nodes, and matrix rows and columns. This completely changes the traditional "single-flow-single-direction" configuration mode. Users only need to configure once to define analysis tasks with M×N flow directions, significantly improving configuration efficiency and solving the technical problem of low configuration efficiency in traditional methods. Moreover, for each cell in the node pair matrix, all ports belonging to the originating and destination nodes are queried based on the multi-dimensional port attribution information, forming the corresponding port set, thus transforming the abstract node dimension into a concrete port list.

[0046] Based on this, and using a unified analysis time range and statistical period, periodic summaries are performed on each cell in the node pair matrix. The analysis time range is divided into continuous periodic sequences according to the statistical period. A basic record set that meets the conditions is selected from the navigation records between ports. The deadweight tonnage is accumulated according to the period. Periods with no data are assigned a value of zero. This ensures that the indicator time series of each flow direction are strictly aligned on the time axis. It also ensures that the data of multiple flows have the same time span, the same time interval, and the same start and end time. This gives the horizontal and vertical comparisons a strict mathematical basis and solves the technical problem that traditional methods cannot directly compare data of multiple flows due to inconsistent statistical standards.

[0047] Finally, the time series of all indicators in each cell are organized according to their row and column positions in the node pair matrix and displayed in parallel on the visualization dashboard in a matrix layout. By organizing the data of M×N flows into a two-dimensional matrix visualization dashboard, users can compare the trends of all flows in a single interface. Cells in the same row of the matrix correspond to the same destination node, and different starting nodes facilitate horizontal comparison; cells in the same column of the matrix correspond to the same starting node, and different destination nodes facilitate vertical comparison. This eliminates the need to switch repeatedly between multiple charts, significantly reducing cognitive burden and improving analysis efficiency.

[0048] In summary, this invention constructs a complete multi-directional shipping indicator analysis system through matrix-based data organization, batch statistical processing logic, and visualized matrix layout. Based on multi-dimensional attribution technology using port latitude and longitude coordinates and their respective countries, it achieves flexible mapping and dynamic analysis of ports at the country, trade zone, and continent levels. Based on unified time axis alignment and zero-padding technology for missing periods, it ensures strict comparability of multi-directional data over time. Based on node-to-matrix cell row and column position mapping technology, it enables batch statistics and organization of multi-directional data. Based on parallel display technology using matrix row and column layout, it enables intuitive comparison of multi-directional time-series indicators. This invention can provide shipping companies with high-confidence data-driven decision support in scenarios such as route optimization, capacity allocation, and trade flow analysis, significantly improving the utilization efficiency and decision-making quality of shipping data assets, and has significant engineering application value and economic benefits.

[0049] Furthermore, this invention acquires berth information and records the originating berth identifier, destination berth identifier, and empty / full load status in the navigation log, allowing users to filter by cargo type and empty / full load status. When a user specifies a cargo type, the system filters records from the basic record set that can load and unload that cargo type at both the originating and destination berths based on the berth information; when a user specifies an empty / full load status, the system filters records whose empty / full load status matches the specified status. This technique enables users to perform refined analysis on specific cargo types (such as containers, iron ore, and crude oil) and specific loading statuses, further enhancing the analytical depth and application value of this invention. This invention fully utilizes the ship's draft information in AIS dynamic data, comparing it with the ship's empty and full load drafts to automatically determine the empty / full load status without manual annotation.

[0050] Furthermore, this invention supports user input of moving average window parameters to perform moving average smoothing on the generated index time series. By calculating the arithmetic mean based on the cumulative deadweight tonnage of each period and the preceding several consecutive periods, and updating the value of the current period with this arithmetic mean, short-term random fluctuations are effectively smoothed, highlighting long-term trends. This allows users to more clearly identify the cyclical changes and structural trends in the shipping market, providing more valuable references for strategic decision-making.

[0051] This invention also relates to a shipping indicator dashboard analysis system based on matrix mapping. This system corresponds to the aforementioned shipping indicator dashboard analysis method based on matrix mapping and can be understood as a system that implements the aforementioned method. It includes a sequentially connected module for constructing multi-dimensional port attribution information, a module for generating inter-port navigation records, a module for defining node pairs in a matrix, a module for constructing port sets, a module for generating indicator time series, and a module for generating and displaying time series indicator curves. These modules work collaboratively, employing data processing techniques that combine matrix mapping with batch statistics. Through core technologies such as multi-dimensional attribution construction based on port latitude and longitude coordinates and their respective countries, AIS voyage identification and association, node pair matrix definition and port set parsing, alignment and zero-padding periodic summarization along a unified time axis, and parallel display of matrix-based time series curves, it achieves efficient comparative analysis of shipping indicators across multiple flow directions. This effectively solves the problems of traditional solutions, such as reliance on repeated manual configuration, inability to uniformly compare multi-flow data, and poor visualization effects. It is suitable for various scenarios, including shipping company route optimization, port operator planning and construction, and trader market judgment. Attached Figure Description

[0052] Figure 1 This is a flowchart of the shipping indicator dashboard analysis method based on matrix mapping of the present invention.

[0053] Figure 2 This is a schematic diagram of a shipping indicator dashboard displayed in a matrix layout according to the present invention. Detailed Implementation

[0054] The present invention will now be described with reference to the accompanying drawings.

[0055] This invention relates to a shipping indicator dashboard analysis method based on matrix mapping. By processing massive amounts of vessel AIS data, a multi-dimensional port attribution system is constructed, standardized voyage records are generated, and batch statistics are performed on the matrix based on user-defined nodes. Finally, the trend curves of indicators for multiple flow directions are displayed on a single dashboard interface according to the matrix layout. In other words, this invention uses processed shipping data from massive amounts of vessel AIS data to statistically analyze indicators (deadweight tonnage, number of vessels, TEU, ton-miles, etc.) of vessels navigating between two nodes (ports, countries, trade zones, continents, etc.) on a periodic basis (day, week, month, quarter, year, etc.), and displays multiple statistics in chart form (line graphs, bar charts, etc.) on the same interface for easy comparison and analysis. The flowchart of this matrix mapping-based shipping indicator dashboard analysis method is shown below. Figure 1 As shown, the steps are as follows:

[0056] I. Steps for constructing multidimensional port attribution information: Obtain vessel information, port information, and berth information; the vessel information includes the vessel's MMSI code and deadweight tonnage; the port information includes the port's latitude and longitude coordinates and the country to which the port belongs; based on the port's latitude and longitude coordinates and the country to which the port belongs, classify each port to the corresponding country, trade zone, or continent node dimension, thereby constructing multidimensional port attribution information.

[0057] This step aims to construct a multi-dimensional attribution system for ports, providing a foundation for subsequently resolving macro-level nodes (such as countries, trade zones, and continents) into specific port sets. Specifically, it first retrieves ship information, port information, and berth information from the database.

[0058] a. Vessel information includes: the vessel's MMSI code (e.g., 123456789), serving as the vessel's unique identifier; deadweight tonnage (e.g., 146,000 tons), used as a freight index value for subsequent statistics; vessel type (e.g., passenger ship, container ship, liquid bulk carrier, dry bulk carrier, general cargo ship, and special-purpose vessel, etc.); and the standard container capacity (TEU) of the container ship (e.g., 13,800). The collected vessel information is shown in Table 1.

[0059] Table 1

[0060]

[0061] b. Port information includes: port latitude and longitude coordinates (used to accurately locate the port), the country to which the port belongs (used for classification by country, such as "China"), and the port name (such as "Port A").

[0062] c. Berth information includes: berth name (e.g., "berth A1"), port to which the berth belongs (e.g., "port A"), and types of cargo that can be loaded and unloaded at each berth (e.g., containers, coal, iron ore, steel, cement, sand and gravel, crude oil, LNG, LPG, grain, raw sugar, bauxite, automobiles, etc.). The collected port berth information is shown in Table 2.

[0063] Table 2

[0064]

[0065] Then, the ports are processed in multiple dimensions. Based on the port's latitude and longitude or its country of origin, each port is categorized into the corresponding country, trade zone, or continent node dimension. The specific categorization rules are as follows:

[0066] 1) Country Dimension: Mapping is based on the country to which each port belongs. For example, if the country to which port A belongs is China, then port A belongs to the country node "China".

[0067] 2) Trade Zone Dimension: Ports are grouped into predefined trade zones based on their latitude and longitude. These trade zones include: Caribbean, East South America, West North America, Bay of Bengal (India), Southeast Asia, West Africa, Black Sea, East North America, Middle East, Mediterranean, East Africa, Far East, Northwest Europe, South Africa, West South America, Australia, and New Zealand. For example, if port A's latitude and longitude fall within the Far East trade zone, then port A belongs to the trade zone node "Far East".

[0068] 3) Continental dimension: Based on the country to which each port belongs, and then determined according to the continent to which the country belongs (such as Asia, Africa, Europe, North America, South America, Central America, Antarctica, Oceania). For example, China belongs to Asia, so port A belongs to the continental node "Asia"; Australia belongs to Oceania, so port C belongs to the continental node "Oceania".

[0069] Through the above processing, the attribution relationship between each port and nodes such as country, trade zone, and continent is established, forming a multi-dimensional port attribution information table, as shown in Table 3.

[0070] Table 3

[0071]

[0072] Simultaneously, the types of cargo handled at each port's berths are collected. The types of cargo that can be loaded and unloaded at each berth include: containers, coal, iron ore, steel, cement, sand and gravel, crude oil, LNG, LPG, grain, raw sugar, bauxite, and automobiles. As shown in the "Cargo Types" column of the port berth information table above, berth A1 can handle containers, and berth A2 can handle iron ore, etc.

[0073] II. Steps for generating inter-port navigation records: Real-time acquisition of vessel AIS dynamic data; based on the port latitude and longitude coordinates in the port multidimensional attribution information, voyage identification and association are performed on the AIS dynamic data of each vessel; complete voyage information from departure from the port of origin to berthing at the destination port is extracted to generate an inter-port navigation record; the inter-port navigation record includes the vessel's MMSI code, port of origin identifier, departure timestamp, destination port identifier, arrival timestamp, and deadweight tonnage, etc.

[0074] This step aims to process the raw AIS dynamic data into structured port navigation records, providing a data foundation for subsequent statistical analysis. Specifically, the AIS dynamic data of the vessels is first acquired in real time and saved into the database. The AIS dynamic data includes: the vessel's MMSI code, the vessel's draft, real-time position (latitude and longitude coordinates), timestamp (data reporting time), and navigation status (such as "underway", "berthed", etc.).

[0075] The raw AIS dynamic data is a record sent by each ship at regular intervals (usually in seconds or minutes). Taking a ship with MMSI 123456789 as an example, some of its AIS data is shown in Table 4 below.

[0076] Table 4

[0077]

[0078] Then, based on the port latitude and longitude coordinates in the port multidimensional attribution information, voyage identification and association processing are performed on the AIS dynamic data of each vessel to extract the complete voyage information of the vessel from departure from the port of origin to berthing at the port of destination. The specific processing logic is as follows:

[0079] Departure detection: When the system detects that the ship's status changes from "berthed" to "sailing" and its position gradually moves away from the latitude and longitude coordinates of a port, it determines that the ship has departed from that port and records the origin port identifier and departure timestamp.

[0080] Arrival Detection: When the system detects that the ship's status changes from "Sailing" to "Boating", and its position enters the latitude and longitude coordinate range of a port and stays there for more than a preset time, it determines that the ship has arrived at the port and records the destination port identifier and arrival timestamp.

[0081] Voyage Linking: This involves linking the departure and subsequent arrival events of the same vessel to form a complete voyage record, and linking it to cargo indicators in the vessel's file (such as deadweight tonnage, vessel number, TEU, ton-nautical miles, etc.). TEU is the international standard unit of measurement for a 20-foot standard container (one TEU represents one 20-foot standard container), and ton-nautical miles is the unit of measurement for shipping cargo turnover, equal to deadweight tonnage multiplied by the distance traveled (nautical miles).

[0082] Through the above processing, standardized inter-port navigation records are generated. Each inter-port navigation record includes: vessel MMSI code, port of origin identifier, departure timestamp, port of destination identifier, arrival timestamp, cargo index value (such as deadweight tonnage), and optional fields such as: port of origin identifier, port of destination identifier, empty / full load status, etc.

[0083] Taking the shipping data for September as an example, the generated port-to-port navigation records are shown in Table 5 below:

[0084] Table 5

[0085]

[0086] Preferably, for both empty and full load states, the determination can be based on the comparison between the ship's current draft and the empty and full load drafts: if the current draft is greater than or equal to the full load draft, the empty / full load state is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load state is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load state is determined to be half load.

[0087] III. Node Pair Matrix Definition Steps: Receive the node pair matrix defined by the user on the visualization dashboard, along with the analysis time range, statistical period (e.g., day, week, month, quarter, year), and product type input by the user; the matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the end point of the analysis; each cell of the node pair matrix corresponds to a node pair to be analyzed, consisting of a starting node in a matrix column and a destination node in a matrix row, thereby establishing a mapping relationship between starting nodes, destination nodes, and matrix rows and columns.

[0088] This step aims to receive the node pair matrix defined by the user on the visualization dashboard, along with the analysis time range and statistical period, enabling "one-stop" configuration for multiple flow directions. The user defines the node pair matrix on the visualization dashboard interface. The node pair matrix is ​​equivalently mapped to a Cartesian coordinate system, with the top-left corner of the matrix as the origin. Horizontal matrix columns correspond to the x-coordinate of the Cartesian coordinate system (also called the row axis), used to arrange multiple originating nodes. Vertical matrix rows correspond to the y-coordinate of the Cartesian coordinate system (also called the column axis), used to arrange multiple destination nodes. Each cell in the matrix is ​​uniquely determined by the originating node corresponding to its matrix column and the destination node corresponding to its matrix row, corresponding to a set of originating node-destination node combinations to be analyzed. Each originating node and destination node is selected from the node dimensions in the port multidimensional attribution information constructed in the above port multidimensional attribution information construction steps. These can be specific ports (e.g., "Port A") or macro-level nodes (e.g., "Asia," "Far East," etc.).

[0089] Users specify the node pairs to be analyzed by defining the rows and columns of a matrix. Each cell in the matrix corresponds to the starting node in the column and the destination node in the row. For example, a user-defined node pair matrix is ​​shown in Table 6 below.

[0090] Table 6

[0091] Oceania █ █ █ Far East █ █ █ Port B █ █ █

[0092] Of these, 9 cells (represented by █) are selected, corresponding to the following 9 pairs of nodes to be analyzed:

[0093] 1) Asia → Oceania

[0094] 2) Australia and New Zealand → Oceania

[0095] 3) Port A → Oceania

[0096] 4) Asia → Far East

[0097] 5) Australia and New Zealand → Far East

[0098] 6) Port A → Far East

[0099] 7) Asia → Port B

[0100] 8) Australia and New Zealand → Port B

[0101] 9) Port A → Port B

[0102] Meanwhile, the statistical parameters input by the user include the analysis time range and the statistical period (the statistical period can be daily, weekly, monthly, quarterly, or yearly). For example, the user inputs the analysis time range as "2023-09-01 to 2023-09-30", the statistical period as "week", and the shipping indicator type to be analyzed as "deadweight tons". In addition, users can also choose to input optional filter conditions, such as cargo type (e.g., "iron ore"), empty / full load status (e.g., "fully loaded"), and moving average window parameters (e.g., 3, i.e., "MA3"), etc.

[0103] IV. Port Set Construction Steps: For each cell in the node pair matrix, based on the multidimensional port affiliation information, query all ports belonging to the originating node of the row coordinate axis of that cell to form the originating port set; and query all ports belonging to the destination node of the column coordinate axis of that cell to form the destination port set.

[0104] This step aims to resolve the abstract node flow dimension into a specific set of originating ports and a set of destination ports for each cell in the node pair matrix. For each cell in the node pair matrix, based on the multi-dimensional port attribution information, the following queries are executed:

[0105] 1) Query the set of originating ports: Find all ports that belong to the originating node of the row coordinate axis of this cell, and form the set of originating ports.

[0106] 2) Query the destination port set: Find all ports that belong to the destination node of the column coordinate axis of this cell, forming the destination port set.

[0107] Taking the cell “Asia → Oceania” in Table 6 as an example: the starting node of the row axis is “Asia”. Based on the multi-dimensional port attribution information shown in Table 3, all ports belonging to Asia are queried to obtain the set of originating ports = {Port A, Port B}; the destination node of the column axis is “Oceania”. Based on the multi-dimensional port attribution information shown in Table 3, all ports belonging to Oceania are queried to obtain the set of destination ports = {Port C}.

[0108] Similarly, for the cell “Australia and New Zealand → Far East”: the starting node of the row axis is “Australia and New Zealand”, and the query returns the set of originating ports = {Port C}; the destination node of the column axis is “Far East”, and the query returns the set of destination ports = {Port A, Port B}.

[0109] For the cell “Port A → Port B”: the starting node of the row axis is “Port A” (the specific port), and the query returns the set of starting ports = {Port A}; the destination node of the column axis is “Port B” (the specific port), and the query returns the set of destination ports = {Port B}.

[0110] The above analysis transforms abstract node pairs into specific port set pairs, laying the foundation for subsequent data filtering.

[0111] V. Steps for Generating the Time Series of Indicators: Based on the analysis time range and statistical period, perform the following operations on each cell in the node pair matrix:

[0112] The analysis time range is divided into a continuous periodic sequence according to the statistical period;

[0113] From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell;

[0114] The basic record set is periodically summarized according to the periodic sequence. The deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value for that period is set to zero, thereby generating the index time series of that cell.

[0115] This step includes the following sub-steps:

[0116] 1) Sub-step for generating periodic sequences:

[0117] The user-input analysis time range (e.g., from September 1, 2023 to September 30, 2023) is divided into continuous periodic sequences according to the statistical period (week). The compiled date range data for September is shown in Table 7 below.

[0118] Table 7

[0119]

[0120] 2) Steps for filtering the basic record set:

[0121] For each cell in the node pair matrix, select records from the port navigation records generated above that simultaneously meet the following conditions, and use these records as the base record set for that cell:

[0122] a. The port of origin belongs to the set of ports of origin for this cell;

[0123] b. The destination port belongs to the set of destination ports for this cell;

[0124] c. Departure time falls within the analysis period (September 1, 2023 to September 30, 2023).

[0125] Taking the "Asia → Oceania" cell as an example: the set of originating ports = {Port A, Port B}, the set of destination ports = {Port C}, and records that meet the criteria are filtered from the navigation records between ports. For example, based on the weekly range, the data from Asia to Oceania is statistically summarized one by one. In the first period from September 1, 2023 to September 3, 2023, no data was matched, so it is recorded as 0; in the second period from September 4, 2023 to September 10, 2023, the data from Port A to Port B in the previous period is matched, but Port A and Port B are both in Asia, while the destination port B is not in Oceania, so the destination node does not match, and it is recorded as 0; in the third period from September 11, 2023 to September 2... At 023-09-17, two data points were found: container ship A sailed from port B to port C, lasting from 2023-09-11 to 2023-09-21; and dry bulk ship B sailed from port B to port C, lasting from 2023-09-14 to 2023-09-19. Port B is in Asia, and port C is in Oceania, so both data points match. The deadweight tonnage is the sum of the deadweight tonnage of the two ships on a single voyage, which is 146000 + 49000 = 195000. These two records are used as the base record set for this cell.

[0126] 3) Sub-step for generating indicator time series through periodic aggregation:

[0127] The filtered basic record set is periodically summarized according to the periodic sequence. The specific steps are as follows:

[0128] a. For each period, sum the freight index values ​​(deadweight tons) of all records departing within that period;

[0129] b. If there are no records in a certain period, the freight index value for that period will be set to zero.

[0130] Optionally, a full-week conversion process is also performed: If the user selects a week as the statistical period, and the current time is within an incomplete period (e.g., the current time is Wednesday, and the week only has 3 days of data), then the cumulative deadweight tonnage value for that period is proportionally converted to a full 7-day estimated value. The conversion method is: the cumulative value for that period divided by the number of days that have passed in that period, then multiplied by 7, to ensure that the data for each period are comparable on the same time base. For example, assuming the current time is August 12, 2026 (Wednesday), and the user's current week is August 10, 2026 (Monday) to August 16, 2026 (Sunday), but there is only data for 3 days from August 10 to August 12, the cumulative deadweight tonnage value for that week is 150,000. Then, the converted estimated value for the full week is: 150,000 ÷ 3 × 7 = 350,000. In the time series index sequence, the converted estimated value of 350,000 is used for this period. By converting, indicators from different periods are unified to the same time benchmark, enhancing data comparability.

[0131] For example, in the deadweight tonnage statistics for the "Asia → Oceania" cell, the values ​​are 0 in week 36, 0 in week 37, 195,000 in week 38 (two records of departures → 146,000 + 49,000 = 195,000), 0 in week 39, and 0 in week 40. The results are shown in Table 8 below:

[0132] Table 8

[0133]

[0134] This generates the time series for the indicator in that cell: [0, 0, 0, 195000, 0]. Similarly, the time series for the indicator in other cells can be calculated. For example, the statistical results for the cell "Australia and New Zealand → Far East" may differ.

[0135] 4) Optional filtering condition processing sub-steps:

[0136] If the user specifies cargo type filtering criteria (such as "iron ore"), then before the periodic summary, based on the berth information, only navigation records where the cargo type can be loaded and unloaded at both the originating berth and the destination berth will be retained.

[0137] If the user specifies the empty / full load status filter criteria (such as "full load"), then only the flight records with the empty / full load status of "full load" will be retained before the periodic summary.

[0138] 5) Optional moving average smoothing sub-step:

[0139] If the user specifies a moving average window parameter (e.g., MA3), then after generating the index time series, a moving average smoothing process is performed on the index time series of each cell. The smoothing process includes: obtaining the number of moving average window periods N input by the user; for each cell's index time series, traversing from the starting period of the index time series, calculating the arithmetic mean based on the cumulative deadweight tonnage value of the current period and the N-1 consecutive periods preceding it, and updating the cumulative deadweight tonnage value of the current period with this arithmetic mean, thereby generating the smoothed index time series.

[0140] Taking the time series of the indicator [0,0,195000,0,0] in the "Asia → Oceania" cell as an example, the MA3 moving average is calculated. The rule for MA3 is: starting from the Nth period (i.e., the 3rd period), the arithmetic mean is calculated based on the values ​​of the current period and the N-1 consecutive periods before it (i.e., the previous 2 periods).

[0141] Week 36: No preceding data → 0 (starting period)

[0142] Week 37: (0+0) / 2 = 0

[0143] Week 38: (0 + 0 + 195000) / 3 = 65000

[0144] Week 39: (0 + 195000 + 0) / 3 = 65000

[0145] Week 40: (195000 + 0 + 0) / 3 = 65000

[0146] The smoothed index time series is generated as follows: [0,0,65000,65000,65000], and the results are shown in Table 9 below.

[0147] Table 9

[0148]

[0149] In the above-mentioned moving average smoothing process, a pre-data expansion optimization method can also be selected: when performing MA3 moving average smoothing, two weeks of data are additionally queried before the start time of the analysis time range as pre-data. This pre-data is only used for MA3 smoothing calculation and is not included in the final time series indicator curve display. After using pre-data expansion, starting from the first period of the analysis time range, the smoothed value of each period is calculated based on the complete three weeks of data, thus ensuring that each data point displayed on the dashboard is the result of complete smoothing. For example, when a user views the MA3 smoothed curve from week 36 to week 40, the data from week 34 to week 40 is actually queried (i.e., week 34 and week 35 are additionally queried as pre-data before week 36). Among them, the pre-data from week 34 and week 35 is only used to calculate the MA3 value of week 36 (requiring data from three periods: week 34, week 35, and week 36). The smoothed values ​​from week 36 onwards are displayed normally, while week 34 and week 35 themselves are not displayed on the dashboard. This way, every data point that a user sees starting from week 36 is a smoothed value calculated based on the complete 3 weeks of data.

[0150] VI. Steps for Generating and Displaying Time Series Indicator Curves: Organize the time series indicators of all cells according to their row and column positions in the node pair matrix to form the time series indicator curves of the corresponding cells; on the visualization dashboard interface, display the time series indicator curves corresponding to multiple cells in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node but different origin nodes for easy horizontal comparison; cells in the same column of the matrix correspond to the same origin node but different destination nodes for easy vertical comparison, so as to realize the comparative analysis of multi-direction shipping indicators corresponding to multiple cells in the matrix.

[0151] This step aims to visually represent the time series indicators of all cells in a unified matrix layout, forming a shipping indicator dashboard that allows for simultaneous comparative analysis across multiple flow directions. This facilitates the analysis of data trends from multiple originating nodes to destination nodes, for the same cargo type, empty / full load status, and within the same time period. Data from originating nodes to destination nodes can be displayed using horizontal and vertical axes. Within the same row, one can compare cargo volumes destined for a specific destination node, while within the same column, one can compare cargo volumes originating from a specific originating node. For example... Figure 2 As shown, on the visualization dashboard (indicator dashboard) interface, the top of the grid is horizontally set with the originating node (From) as the matrix column dimension, and the left side of the grid is vertically set with the destination node (To) as the matrix row dimension. Each cell within the grid corresponds to a set of shipping flows from "originating node → destination node". Each cell displays the time-series indicator curve for the corresponding flow direction using a line chart. Figure 2The example shows the cumulative deadweight tonnage as the indicator, with a weekly time series. The horizontal axis of the line chart represents the weekly time period (week 36 to week 40), and the vertical axis represents the cumulative deadweight tonnage.

[0152] With this matrix-style visualization layout, users can complete a comprehensive comparative analysis of shipping data from multiple directions on a single interface, without having to switch repeatedly between multiple independent charts, which greatly improves the efficiency and intuitiveness of shipping data analysis.

[0153] This invention also relates to a shipping indicator dashboard analysis system based on matrix mapping. This system corresponds to the aforementioned shipping indicator dashboard analysis method based on matrix mapping and can be understood as a system that implements the above method. The system includes, in sequence, a port multidimensional attribution information construction module, a port navigation record generation module, a node pair matrix definition module, a port set construction module, an indicator time series generation module, and a time series indicator curve generation and display module. Specifically,

[0154] The port multidimensional attribution information construction module acquires ship information and port information; the ship information includes the ship's MMSI code and deadweight tonnage; the port information includes the port's latitude and longitude coordinates and the country to which the port belongs; based on the port's latitude and longitude coordinates and the country to which the port belongs, each port is classified into the corresponding country, trade zone or continent node dimension, thereby constructing the port multidimensional attribution information;

[0155] The inter-port navigation record generation module acquires the AIS dynamic data of ships in real time. Based on the port latitude and longitude coordinates in the port multidimensional attribution information, it identifies and associates the voyage data of each ship's AIS dynamic data, extracts the complete voyage information of the ship from departure from the port of origin to berthing at the port of destination, and generates an inter-port navigation record. The inter-port navigation record includes the ship's MMSI code, port of origin identifier, departure timestamp, port of destination identifier, arrival timestamp, and deadweight tonnage.

[0156] The node pair matrix definition module obtains the node pair matrix defined by the user on the visualization dashboard and the statistical parameters input by the user. The statistical parameters include the analysis time range and the statistical period. The matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the ending point of the analysis. Each cell of the node pair matrix corresponds to the starting node in the column and the destination node in the row, thereby establishing a mapping relationship between the starting nodes, destination nodes and the matrix rows and columns.

[0157] The port set construction module, for each cell in the node pair matrix, queries all ports belonging to the originating node of the matrix column where the cell is located, based on the multidimensional port affiliation information, to form an originating port set; and queries all ports belonging to the destination node of the matrix row where the cell is located, to form a destination port set.

[0158] The indicator time series generation module performs the following operations on each cell in the node pair matrix based on the analysis time range and statistical period:

[0159] The analysis time range is divided into a continuous periodic sequence according to the statistical period;

[0160] From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell;

[0161] The basic record set is periodically summarized according to the periodic sequence, and the deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value of that period is set to zero, thereby generating the index time series of that cell.

[0162] The time-series indicator curve generation and display module organizes the time series indicators of all cells according to their row and column positions in the node pair matrix, forming the time-series indicator curves of the corresponding cells. On the visualization dashboard interface, the time-series indicator curves corresponding to multiple cells are displayed in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node but different origin nodes, which facilitates horizontal comparison. Cells in the same column of the matrix correspond to the same origin node but different destination nodes, which facilitates vertical comparison, so as to realize the comparative analysis of multi-direction shipping indicators corresponding to multiple cells in the matrix.

[0163] Preferably, the port multidimensional attribution information construction module also acquires berth information, which includes berth name, port to which the berth belongs, and types of cargo that can be loaded and unloaded at each berth; in the port navigation record generation module, the port navigation record also includes the originating berth identifier, the destination berth identifier, and the empty / full load status; in the node pair matrix definition module, the statistical parameters also include cargo types.

[0164] The indicator time series generation module, after filtering out the basic record set, further includes: determining whether the statistical parameters input by the user contain cargo types; if so, filtering out and retaining records from the basic record set where both the originating berth and the destination berth can load and unload the cargo type based on the berth information; determining whether the statistical parameters input by the user contain empty / full load status; if so, filtering out and retaining records from the basic record set whose empty / full load status matches the empty / full load status specified by the user; and performing the periodic summary based on the finally filtered navigation records.

[0165] Preferably, the indicator time series generation module further performs smoothing processing on the generated indicator time series. The smoothing processing includes: obtaining the moving average window period data N input by the user; for the indicator time series of each cell, traversing from the starting period of the indicator time series, calculating the arithmetic mean based on the cumulative deadweight tonnage value of the current period and the previous N-1 consecutive periods, and updating the cumulative deadweight tonnage value of the current period with the arithmetic mean, thereby generating the smoothed indicator time series.

[0166] Preferably, in the port navigation record generation module, the AIS dynamic data includes the ship's draft, real-time location, timestamp, and navigation status; the empty / full load status is determined based on the comparison between the ship's current draft and the empty / full load draft: if the current draft is greater than or equal to the full load draft, the empty / full load status is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load status is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load status is determined to be half load.

[0167] Preferably, in the node pair matrix definition module, the statistical period includes day, week, month, quarter, and year.

[0168] This invention provides an objective and scientific method and system for analyzing shipping indicators based on matrix mapping. Through matrix-based data organization, batch statistical processing logic, and visualization of matrix layout, a complete multi-flow shipping indicator analysis system is constructed, which significantly improves the efficiency, accuracy, and comparability of shipping data analysis, and provides a powerful data decision support tool for shipping companies, port operators, traders, and regulatory agencies.

[0169] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail with reference to the accompanying drawings and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention. In short, all technical solutions and improvements that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention patent.

Claims

1. A shipping indicator dashboard analysis method based on matrix mapping, characterized in that, Includes the following steps: The steps for constructing multidimensional port attribution information are as follows: First, acquire vessel information and port information. The vessel information includes the vessel's MMSI code and deadweight tonnage. The port information includes the port's latitude and longitude coordinates and the country to which the port belongs. Based on the port's latitude and longitude coordinates and the country to which the port belongs, each port is classified into the corresponding country, trade zone, or continent node dimension, thereby constructing multidimensional port attribution information. The steps for generating inter-port navigation records are as follows: Real-time acquisition of vessel AIS dynamic data; based on the port latitude and longitude coordinates in the port multidimensional attribution information, voyage identification and association are performed on the AIS dynamic data of each vessel; complete voyage information from departure from the port of origin to berthing at the destination port is extracted to generate an inter-port navigation record; the inter-port navigation record includes the vessel's MMSI code, port of origin identifier, departure timestamp, destination port identifier, arrival timestamp, and deadweight tonnage; Node pair matrix definition steps: Obtain the node pair matrix defined by the user on the visualization dashboard and the statistical parameters input by the user. The statistical parameters include the analysis time range and the statistical period. The matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the end point of the analysis. Each cell in the matrix corresponds to the starting node in the column and the destination node in the row, thereby establishing a mapping relationship between the starting node, the destination node, and the matrix rows and columns; Port set construction steps: For each cell in the node pair matrix, based on the multidimensional port affiliation information, query all ports belonging to the originating node in the column of that cell to form the originating port set; And query all ports belonging to the destination node of the row containing the cell, forming a set of destination ports; Indicator time series generation steps: Based on the analysis time range and statistical period, perform the following operations on each cell in the node pair matrix: The analysis time range is divided into a continuous periodic sequence according to the statistical period; From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell; The basic record set is periodically summarized according to the periodic sequence, and the deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value of that period is set to zero, thereby generating the index time series of that cell. Steps for generating and displaying time series indicator curves: Organize the time series indicators of all cells according to their row and column positions in the node pair matrix to form the time series indicator curves of the corresponding cells; On the visualization dashboard interface, display the time series indicator curves of multiple cells in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node, and different starting nodes facilitate horizontal comparison. Cells in the same column of the matrix correspond to the same originating node but different destination nodes, which facilitates vertical comparison and enables the node to perform comparative analysis of multi-directional shipping indicators corresponding to multiple cells in the matrix.

2. The shipping indicator dashboard analysis method based on matrix mapping according to claim 1, characterized in that, In the port multidimensional attribution information construction step, berth information is also obtained, including berth name, port to which the berth belongs, and types of cargo that can be loaded and unloaded at each berth; in the port navigation record generation step, the port navigation record also includes the originating berth identifier, the destination berth identifier, and the empty / full load status; in the node pair matrix definition step, the statistical parameters also include cargo types. In the step of generating the time series of indicators, after filtering out the basic record set, the method further includes: determining whether the statistical parameters input by the user contain cargo types; if so, filtering out and retaining records from the basic record set where both the originating berth and the destination berth can load and unload the cargo types based on the berth information; determining whether the statistical parameters input by the user contain empty / full load status; if so, filtering out and retaining records from the basic record set whose empty / full load status matches the empty / full load status specified by the user; and performing the periodic summary based on the navigation records obtained from the final filtering.

3. The shipping indicator dashboard analysis method based on matrix mapping according to claim 1, characterized in that, In the step of generating the indicator time series, the generated indicator time series is also smoothed, and the smoothing process includes: Get the number of moving average window periods N input by the user; for each cell's index time series, start from the starting period of the index time series and traverse it. Calculate the arithmetic mean based on the cumulative deadweight tonnage of the current period and the N-1 consecutive periods before it, and update the cumulative deadweight tonnage of the current period with this arithmetic mean, thereby generating a smoothed index time series.

4. The shipping indicator dashboard analysis method based on matrix mapping according to claim 2, characterized in that, In the port navigation record generation step, the AIS dynamic data includes the ship's draft, real-time position, timestamp, and navigation status. The empty / full load status is determined based on the comparison between the ship's current draft and its empty and full load drafts: if the current draft is greater than or equal to the full load draft, the empty / full load status is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load status is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load status is determined to be half load.

5. The shipping indicator dashboard analysis method based on matrix mapping according to claim 1, characterized in that, In the node pair matrix definition step, the statistical period includes day, week, month, quarter, and year.

6. A shipping indicator dashboard analysis system based on matrix mapping, characterized in that, This includes, in sequence, a module for constructing multidimensional port attribution information, a module for generating navigation records between ports, a module for defining node pair matrices, a module for constructing port sets, a module for generating time series indicators, and a module for generating and displaying time series indicator curves. The port multidimensional attribution information construction module acquires ship information and port information; the ship information includes the ship's MMSI code and deadweight tonnage; the port information includes the port's latitude and longitude coordinates and the country to which the port belongs; based on the port's latitude and longitude coordinates and the country to which the port belongs, each port is classified into the corresponding country, trade zone or continent node dimension, thereby constructing the port multidimensional attribution information; The inter-port navigation record generation module acquires the AIS dynamic data of ships in real time. Based on the port latitude and longitude coordinates in the port multidimensional attribution information, it identifies and associates the voyage data of each ship's AIS dynamic data, extracts the complete voyage information of the ship from departure from the port of origin to berthing at the port of destination, and generates an inter-port navigation record. The inter-port navigation record includes the ship's MMSI code, port of origin identifier, departure timestamp, port of destination identifier, arrival timestamp, and deadweight tonnage. The node pair matrix definition module obtains the node pair matrix defined by the user on the visualization dashboard and the statistical parameters input by the user. The statistical parameters include the analysis time range and the statistical period. The matrix columns of the node pair matrix are used to arrange multiple starting nodes as the starting point of the analysis, and the matrix rows are used to arrange multiple destination nodes as the end point of the analysis. Each cell in the matrix corresponds to the starting node in the column and the destination node in the row, thereby establishing a mapping relationship between the starting node, the destination node, and the matrix rows and columns; The port set construction module, for each cell in the node pair matrix, queries all ports belonging to the originating node in the column of the cell according to the multidimensional port affiliation information, thus forming an originating port set. And query all ports belonging to the destination node of the row containing the cell, forming a set of destination ports; The indicator time series generation module performs the following operations on each cell in the node pair matrix based on the analysis time range and statistical period: The analysis time range is divided into a continuous periodic sequence according to the statistical period; From the port navigation records, select records whose port of origin belongs to the set of ports of origin of this cell, whose port of destination belongs to the set of ports of destination of this cell, and whose departure time is within the analysis time range, and use these records as the basic record set for this cell; The basic record set is periodically summarized according to the periodic sequence, and the deadweight tonnage in each period is accumulated. If there are no navigation records in a certain period, the accumulated deadweight tonnage value of that period is set to zero, thereby generating the index time series of that cell. The time series indicator curve generation and display module organizes the indicator time series of all cells according to their row and column positions in the node pair matrix to form the time series indicator curve of the corresponding cell. On the visualization dashboard interface, the time series indicator curves corresponding to multiple cells are displayed in parallel according to the row and column layout of the node pair matrix. Cells in the same row of the matrix correspond to the same destination node, and different starting nodes facilitate horizontal comparison. Cells in the same column of the matrix correspond to the same originating node but different destination nodes, which facilitates vertical comparison and enables the node to perform comparative analysis of multi-directional shipping indicators corresponding to multiple cells in the matrix.

7. The shipping indicator dashboard analysis system based on matrix mapping according to claim 6, characterized in that, The port multidimensional attribution information construction module also acquires berth information, which includes berth name, port to which the berth belongs, and types of cargo that can be loaded and unloaded at each berth; the inter-port navigation record generation module also includes the originating berth identifier, destination berth identifier, and empty / full load status; the node pair matrix definition module also includes cargo types as statistical parameters. The indicator time series generation module, after filtering out the basic record set, further includes: determining whether the statistical parameters input by the user contain cargo types; if so, filtering out and retaining records from the basic record set where both the originating berth and the destination berth can load and unload the cargo type based on the berth information; determining whether the statistical parameters input by the user contain empty / full load status; if so, filtering out and retaining records from the basic record set whose empty / full load status matches the empty / full load status specified by the user; and performing the periodic summary based on the finally filtered navigation records.

8. The shipping indicator dashboard analysis system based on matrix mapping according to claim 6, characterized in that, The indicator time series generation module further performs smoothing processing on the generated indicator time series, the smoothing processing including: Get the number of moving average window periods N input by the user; for each cell's index time series, start from the starting period of the index time series and traverse it. Calculate the arithmetic mean based on the cumulative deadweight tonnage of the current period and the N-1 consecutive periods before it, and update the cumulative deadweight tonnage of the current period with this arithmetic mean, thereby generating a smoothed index time series.

9. The shipping indicator dashboard analysis system based on matrix mapping according to claim 7, characterized in that, In the port navigation record generation module, the AIS dynamic data includes the ship's draft, real-time location, timestamp, and navigation status. The empty / full load status is determined based on the comparison between the ship's current draft and its empty and full load drafts: if the current draft is greater than or equal to the full load draft, the empty / full load status is determined to be full load; if the current draft is less than or equal to the empty draft, the empty / full load status is determined to be empty; if the current draft is between the empty and full load drafts, the empty / full load status is determined to be half load.

10. The shipping indicator dashboard analysis system based on matrix mapping according to claim 6, characterized in that, In the node pair matrix definition module, the statistical period includes daily, weekly, monthly, quarterly, and annual periods.

Images