Monitoring and prediction system for dynamically monitoring port ship pressure

Abstract

The invention provides a monitoring and predicting system for dynamically monitoring harbor group ship harbor pressure, which can provide harbor logistics business busy degree quantitative information in real time, quickly and accurately, and solves the problem of harbor use efficiency for each harbor by combining ship dynamic information, so that accurate monitoring and predicting of harbor pressure of ships can be realized. According to the technical scheme, the system is characterized by comprising a data source module (1), a data storage module (2), an engine service module (3), an application service module (4) and an interaction layer module (5).

Description

Technical Field

[0001] The present invention belongs to the technical field of port ship congestion monitoring, and particularly relates to a monitoring and prediction system for dynamically monitoring port ship congestion. Background Art

[0002] Traditional ship shipping visualization tracking systems take ship dynamics as the core object and mainly focus on ships to monitor information such as ship position, route, and speed. Regarding ports, they mainly rely on the current situation of ships in the port and cannot reflect the port usage efficiency from the perspective of ship berthing and operation time.

[0003] During the shipping business process, a port is a hub for centralized shipment of goods. Judging ship congestion based on anchoring time directly affects the port operation efficiency, which is directly related to the efficiency and service quality of shipping cargo transportation. Therefore, among user groups such as freight service providers and customers centered around shipping, there is an urgent need to understand the ship congestion status of ports in real time to understand the port operation busyness and the operation efficiency of ships.

[0004] Currently, monitoring systems in the shipping field mainly monitor single ports, mainly from the number of ships and the volume of operations in the current port. The problems are that single ports have low cargo dispatching capabilities, weak capabilities in handling port cargo backlogs, low shipping logistics operation efficiency, and increased port logistics pressure.

[0005] With the development of the ability to collect and analyze ship dynamic big data, by using real-time ship dynamic data and through big data analysis and algorithm models, it is possible to monitor the overall situation of selected port groups, monitor and predict the real-time ship congestion situation of each port, provide a prediction basis for shipping companies, shipping lines, freight forwarders, and customers in the cargo transportation cycle and efficiency, facilitate decision-making for customer groups such as arranging and allocating goods in advance, choosing reasonable transportation methods, and allocating the volume of goods transported, and can also provide a reference factor for transportation service providers as a reference for freight rate changes. Summary of the Invention

[0006] In order to solve the problem that the current monitoring systems in the shipping field mainly focus on the dynamics of ships, and the ports mainly monitor the number of ships in the port and the volume of operations, and it is impossible to reflect the port utilization efficiency from more dimensions such as ship berthing and operation time. Therefore, there are further problems such as low cargo dispatching ability, weak ability to handle port cargo backlogs, low shipping logistics operation efficiency, and increased port logistics pressure, and it is impossible to accurately monitor and predict ship congestion. The present invention provides a monitoring and prediction system for dynamically monitoring ship congestion in port groups, which can provide quantitative information on the busyness of port logistics operations in real time, quickly, and accurately. By combining ship dynamic information, it solves the problem of port utilization efficiency for each port, and thus can achieve accurate monitoring and prediction of ship congestion.

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

[0008] A monitoring and prediction system for dynamically monitoring ship congestion in ports, characterized in that it is composed of a data source module (1), a data storage module (2), an engine service module (3), an application service module (4), and an interaction layer module (5); among them,

[0009] The data source module (1) provides data sources such as ship AIS, GIS, ship archive data, and dynamic data of each port for data calculation and business analysis. The AIS data of the ship includes the running speed, status, direction of ships in each port, and the real-time number of ships approaching the port. The GIS data includes basic geographical information data of China and various countries. The ship archive data includes archive data provided by HiFleet such as ship basic information, ship dimensions, and MMSI change records. The dynamic data of each port includes historical and real-time ship berthing, docking, and anchoring operation time data;

[0010] The data storage module (2) is used to store the relevant data received by the data source module (1), establish an ODS data storage center, an analysis database and an early warning database connected to the ODS, and output the stored data to the engine service module (3);

[0011] The engine service module (3) performs cleaning, processing, and thematization on the data stored in the data storage module (2) through big data analysis. The thematization refers to the thematization processing of the data source data, so that the data source data is based on the real-time dynamic data of ships in each port; using the processed data, real-time calculations are performed with the changes in the operation time of ship pre-arrival, berthing, docking, and anchoring in the port as indicators, and multi-dimensional quantitative data analysis is carried out to calculate the time of ship congestion in each port and predict the operation and berthing time of ships arriving at the port. The engine service module (3) includes a quantitative analysis unit, an analysis model unit, and a prediction model unit;

[0012] The quantization analysis unit is based on the ship's navigation position, navigation speed, navigation direction, and the port area of the destination port for each port, and takes the analysis of the time changes of the ship's pre-arrival at the port, berthing at the port, docking at the port, and anchoring at the port as indicators for real-time calculation and analysis summary, forming a multi-dimensional data analysis model of port, ship, status, time, etc., which is stored in the analysis model unit. By establishing a multi-dimensional data analysis model of the port, calling various algorithms constructed by the analysis model unit, and combining data thematic design for full integration of algorithms and data, the berthing and operation time of current and historical port ships are analyzed to obtain the current and historical ship quantization port congestion situation information. The quantization port congestion situation information includes the port congestion situation obtained according to the key port congestion data index items and the port real-time port congestion status classification definition;

[0013] The analysis model unit stores the data analysis model constructed by the quantization analysis unit, and constructs an algorithm for predicting the arrival time of ship types, an algorithm for analyzing the ship's berthing time, and an algorithm for analyzing the operation time of port ships;

[0014] The algorithm for predicting the arrival time is based on the current ship segment dynamic L curr , the current time T curr and the destination port P. Through the average ship speed V avg and the remaining voyage D left of the ship segment, calculate the predicted arrival time T cta of the current destination port P of the ship:

[0015] T cta = T curr +(D left / V avg )

[0016] The algorithm for analyzing the berthing time is based on the current ship dynamic L curr , the start anchoring time T moor when arriving at the destination port and the start berthing time T berth and the end berthing time T leave of the ship, calculate the anchoring duration Dur moor and the berthing duration Dur berth of the ship at the destination port; Based on all ship types V vessel_type in the port and the anchoring duration Dur moor and the berthing duration Dur berth calculate the average anchoring duration PDur moor {vessel_type = V vessel_type} and the average berthing duration PDur berth {vessel_type = V vessel_type}

[0017] Durmoor = T berth -T moor

[0018] Dur berth = T leave -T berth

[0019] PDur moor {vessel_type = V vessel_type}} = Avg(Dur moor ,

[0020] vessel_type = V vessel_type )

[0021] PDur berth {vessel_type = V vessel_type}} = Avg(Dur berth ,

[0022] vessel_type = V vessel_type )

[0023] The port ship operation time analysis algorithm is based on the current dynamics L of the ship curr , the anchoring duration Dur moor at the destination port and the berthing duration Dur berth to calculate the operation duration Dur port of the ship at the destination port; based on the ship type V vessel_type of all ships in the port and the operation duration Dur port to calculate the average operation duration PDur port {vessel_type = V vessel_type}

[0024] Dur port = Dur moor + Dur berth

[0025] PDur port {vessel_type = V vessel_type}} = Avg(Dur port ,

[0026] vessel_type = V vessel_type )

[0027] The prediction model unit is based on information such as the dynamic data of ship AIS, destination port, navigation, current location, and navigation speed. According to the information on the congestion situation of each port obtained by the quantitative analysis unit, that is, combined with the actual congestion situation of the current port and the expected arrival volume situation of the port, a multi-dimensional calculation model with time serialization, operation status, and space prefabrication is constructed. That is, for the ships about to arrive at the port, according to information such as ship type, load capacity, route, and speed, a prediction analysis model is constructed to realize the prediction of the arrival operation and berthing operation time of the ships about to arrive at the port, the expected quantitative calculation of the arrival and port operation time, the quantitative measurement of the stay time and operation time of the ships after arriving at the port, the monitoring of the in-port situation of ships in each port, the monitoring of the expected port stay of ships in each port, the monitoring of the expected port stay trend of ships in each port, the monitoring of the port stay indicators of ships in each port, and the hierarchical warning of the in-port time of ships.

[0028] The application service module (4) uses the calculation results of the engine service module (3). After data screening, analysis, judgment, calculation, and processing, it functionally displays relevant data. The display content includes: ship operation time, port operation time, ship berthing time, ship arrival information, ship expected arrival and departure time, warning information, and statistical analysis.

[0029] The interaction layer module (5) realizes the data display of the backend system, including GIS map display and mobile application support display.

[0030] The engine service module (3) also includes a data processing unit, a data cleaning unit, a data quality unit, a data thematic unit, a job scheduling unit, and an algorithm engine unit. The data processing unit, data cleaning unit, data quality unit, and data thematic unit clean, process, and theme the data stored in the data storage module (2) to obtain high-quality data with usability and processing efficiency. The job scheduling unit transfers the high-quality data and the ship quantitative congestion situation information obtained by the quantitative analysis unit to the prediction model unit for quantitative measurement of the stay time and operation time of the ships after arriving at the port. The algorithm engine unit provides algorithm support for the prediction model unit and the analysis model unit.

[0031] The key port congestion data index items include:

[0032] Anchored ships (number): The number of ships waiting for berthing in the port that are safely berthed by using the anchor mooring method.

[0033] Pre-arriving ships (number): The number of ships expected to arrive at the port within the agreed time range.

[0034] Berthing ships (number): The number of ships berthing at the port.

[0035] Average anchoring time (hours): The waiting time for a ship to anchor before waiting to berth at a port.

[0036] Average berthing time (hours): The time from when a ship berths at a port to when it leaves.

[0037] Average stay time in port (hours): The total stay time within the port area, including operation time, anchoring time, and berthing time.

[0038] The classification definition of the real-time congestion status of the port includes: different colored dots representing different anchoring times, where the different anchoring times include: anchoring time ≤ 48 hours; 48h < anchoring time ≤ 72h; 72h < anchoring time ≤ 96h; anchoring time > 96h; the classification definition of the real-time congestion status of the port also includes the decrease or increase in anchoring time compared to yesterday represented by different colored arrows.

[0039] The monitoring and prediction system for dynamically monitoring the congestion of ships in a port group is characterized in that in the application service module (4),

[0040] The ship operation time includes the average ship operation time, which uses big data algorithms to determine the average port operation time of different types of ships.

[0041] The port operation time includes the average port operation time, which analyzes the current ship operation time and historical operation time of the port based on current and historical AIS+GIS data using big data algorithms to obtain information on changes in the average port operation time.

[0042] The ship berthing time refers to the ship's berthing time at the port, which is obtained by analyzing historical AIS+GIS data to obtain the ship's historical port docking data.

[0043] The ship arrival information uses AIS+GIS data to determine the navigation, berthing, or anchoring status of the ship.

[0044] The estimated arrival and departure time of the ship is a predictive calculation for ships arriving at the destination port in transit by comprehensively considering the current port operation time of the port, the expected number of berthing ships, and the load information, providing predictions of the estimated arrival and departure times for the ships.

[0045] The data display of the backend system includes two methods: nautical map display and nautical map + list display.

[0046] The hierarchical early warning of the prediction model unit refers to different early warning levels for ships of different deadweights.

[0047] The specific warning classification is as follows: for ships of different deadweights, an early warning mechanism is implemented based on the length of stay in port, such as the number of days. Different colors represent different warning levels and the range of the ship's length of stay in port.

[0048] The effects of the present invention are as follows:

[0049] The monitoring and prediction system for ship congestion in port clusters of the present invention, its data source module (1) provides data sources such as ship AIS, GIS, ship archive data, and dynamic data of each port, which are used for data calculation and business analysis. That is, the data sources included in the monitoring system in the shipping field involved in the present invention are extensive and comprehensive, not only involving ship dynamic data, but also including dynamic data of different ports, as well as ship archive data and dynamic data of each port, providing comprehensive data support for multi-dimensional analysis, monitoring, and early warning.

[0050] The quantitative analysis unit is based on the ship's navigation position, navigation speed, navigation direction, and the port area of the navigation destination for each port, and takes the analysis of the time changes of the ship's expected arrival at the port, berthing at the port, mooring at the port, and anchoring at the port as indicators for real-time calculation and analysis summary, forming a multi-dimensional data analysis model with ports, ships, status, time, etc. Thus, it has changed the situation that the previous monitoring systems mainly focused on ship dynamics as the core object, and the port aspect mainly focused on the current situation of ships in port, and all were for single-port monitoring. The present invention can take the port as the core analysis target, and through the comprehensive summary and real-time calculation and analysis of multi-parameters of ships coming and going between this port and other ports, a multi-dimensional data analysis model can be obtained. That is, the port congestion situation monitoring platform is built from the perspective of the port congestion situation of this port and the dynamic relationship of ships from other ports to this port. It completely takes the port as the core object, uses the dynamic data of ships to analyze the time information such as the entry and departure of ships from the port, berthing, mooring, and port operations, and analyzes and monitors the ship operation efficiency status of the port from a business perspective, providing highly available scheduling basic information for shipping freight.

[0051] Based on information such as the dynamic data of ship AIS, the destination port, navigation, current location, and navigation speed, the prediction model unit constructs a multi-dimensional calculation model with time serialization, operation status, and spatial prefabrication according to the information on the congestion situation of each port obtained by the quantization analysis unit, that is, combining the actual congestion situation of the current port and the expected arrival volume situation of the port. For the ships about to arrive at the port, according to information such as ship type, load capacity, shipping route, and speed, a prediction analysis model is constructed to predict the arrival operation and berthing operation time of the ships about to arrive at the port. In particular, for ships, a quantitative calculation of the expected arrival and port operation time within 72 hours is carried out, and the residence time and operation time of the ships after arriving at the port are quantitatively measured, so as to realize the monitoring of the in-port situation of ships in each port, the monitoring of the expected port stay of ships in each port, the monitoring of the expected port stay trend of ships in each port, the monitoring of the port stay indicators of ships in each port, and the classification warning of the in-port time of ships. That is, through the comprehensive analysis of new information such as the entry and departure of ships from the port, berthing, docking, and operation, the analysis and prediction of the port congestion time situation of ships are realized, and the dynamic ability of the port congestion situation of ships at all times and the prediction ability of the expected arrival situation of a single ship are realized, providing timely, efficient, and accurate port operation efficiency information for ship management and port freight transportation enterprises. Without the need for users to enter data and intervene, it has the quantitative service capabilities of automation, intelligence, and visualization, and provides the real-time ship operation efficiency status information of the port for shipping enterprises, ship dispatching, and logistics enterprise dispatching, realizing the quantitative information of the real-time, fast, and accurate port logistics business busyness, helping business-related parties such as shippers, freight forwarders, and shipowners to master the port busyness, improving the cargo dispatching ability, and providing efficient support for reducing the backlog of port goods, overall improving the operation efficiency of shipping logistics, and providing guarantee for the port to relieve the logistics pressure.

[0052] At the same time as the quantitative analysis and monitoring of port congestion, the quantitative operation time prediction ability for relevant shipping ships expected to arrive at the port is also realized, effectively predicting the arrival time, expected anchoring, docking time, and operation time of the ships, and providing a reference-value shipping business prediction service for the ships.

[0053] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. Description of the Drawings

[0054] Figure 1 is the block diagram of the architecture of the present invention;

[0055] Figure 2 is the structural schematic diagram of the present invention;

[0056] Figure 3 is the service flow chart of the present invention;

[0057] Figure 4It is a monitoring chart of the berthing situation of port ships;

[0058] Figure 5 It is a monitoring chart of the in-port situation of port ships;

[0059] Figure 6 It is a monitoring chart of the expected berthing trend of port ships;

[0060] Figure 7 It is a monitoring chart of the expected in-port trend of port ships;

[0061] Figure 8 It is a monitoring chart of port ship berthing indicators;

[0062] Figure 9 It is a classification definition table of the real-time port congestion status of ports;

[0063] Figure 10 It is a classification definition table of the early warning for the in-port time of ships. Detailed implementation mode

[0064] Such as Figure 1 As shown, the monitoring and prediction system for dynamically monitoring port ship congestion consists of five main parts: a data source module 1, a data storage module 2, an engine service module 3, an application service module 4, and an interaction layer module 5.

[0065] Figure 2 Among them, the data source module 1 is an important data source of the present invention. It is a data source data formed by taking the public ship AIS, GIS, and ship archive data as the core and introducing the dynamic data of each port. The AIS data of the ship includes the running speed, status, direction of ships in each port, and the real-time number of ships approaching the port. The GIS data includes the basic geographic information data of China and various countries. The ship archive data includes the basic information of the ship, ship dimensions, MMSI change records, and other archive data provided by HiFleet. The dynamic data of each port includes the operation time data of historical and real-time ship berthing, docking, and anchoring in the port; it is used to support the data calculation and business analysis of the present invention.

[0066] The data storage module 2 is used to store the relevant original data received by the accessed data source module 1. It constructs two main storage structures with ODS as the data storage center and two data entities, an analysis database and an early warning database, connected to ODS, for improving the execution efficiency and analysis effect of the data analysis objectives of the present invention. The data storage module 2 outputs the stored data to the engine service module 3.

[0067] The engine service module 3 is the core service module for data analysis and business prediction. Using big data analysis technology, it cleans, processes, and thematizes the data stored in the data storage module 2. Thematization refers to the thematization process of the data source data, making the data source data the real-time dynamic data of ships based on the scope of each port area. Using the processed data, real-time calculations are carried out with the changes in the operation times of ships' pre-arrival, berthing, mooring, and anchoring at ports as indicators, and multi-dimensional quantitative data analysis is performed to calculate the time of ship congestion at each port and predict the berthing operation time of ships arriving at the port. The engine service module 3 includes a quantitative analysis unit, an analysis model unit, and a prediction model unit.

[0068] The quantitative analysis unit, based on the ship's navigation position, speed, direction, and the port area of the navigation destination for each port, takes the analysis of the time changes of ships' pre-arrival at the port, berthing at the port, mooring at the port, and anchoring at the port as indicators to perform real-time calculations and analysis summaries, forming a multi-dimensional data analysis model based on ports, ships, status, time, etc., which is stored in the analysis model unit. By establishing a multi-dimensional data analysis model for ports, calling various algorithms constructed by the analysis model unit, and combining with data thematization design for full integration of algorithms and data, the berthing and operation times of current and historical port ships are analyzed to obtain the current and historical ship quantitative congestion situation information. The quantitative congestion situation information includes the congestion situation obtained according to the key congestion data index items and the port real-time congestion status classification definition.

[0069] The analysis model unit stores the data analysis model constructed by the quantitative analysis unit, and constructs the pre-arrival time algorithm for ship types, the ship berthing time analysis algorithm, and the port ship operation time analysis algorithm, that is, different ship type congestion time calculation methods are constructed respectively based on the pre-arrival time, ship berthing time, and port ship operation time, so as to accurately reflect the congestion change status.

[0070] The pre-arrival time algorithm is based on the current ship segment dynamics L curr , the current time T curr , and the destination port P. Through the average ship speed V avg and the remaining voyage distance D left of the ship segment, the pre-arrival time T cta of the ship at the current destination port P is calculated:

[0071] T cta = T curr + (D left / V avg )

[0072] The berthing time analysis algorithm is based on the current ship dynamics L curr , the start anchoring time T moor when arriving at the destination port and the start berthing time Tberth and the end berthing time T leave , calculate the anchoring duration Dur of the ship at the destination port moor and the berthing duration Dur berth ; based on the vessel types V of all ships in the port vessel_type and the anchoring duration Dur moor and the berthing duration Dur berth calculate the average anchoring duration PDur of the port P moor {vessel_type = V vessel_type} and the average berthing duration PDur berth {vessel_type = V vessel_type}

[0073] Dur moor = T berth - T moor

[0074] Dur berth = T leave - T berth

[0075] PDur moor {vessel_type = V vessel_type}= Avg(Dur moor ,

[0076] vessel_type = V vessel_type )

[0077] PDur berth {vessel_type = V vessel_type}= Avg(Dur berth ,

[0078] vessel_type = V vessel_type )

[0079] The port ship operation time analysis algorithm is based on the current dynamics L of the ship curr , the anchoring duration Dur at the destination port moor and the berthing duration Dur berth , calculate the operation duration Dur of the ship at the destination port port ; based on the vessel types V of all ships in the port vessel_type and the operation duration Dur port calculate the average operation duration PDur of the port P port {vessel_type = V vessel_type}

[0080] Dur port = Dur moor+Dur berth

[0081] PDur port {vessel_type = V vessel_type} = Avg(Dur port ,

[0082] vessel_type = V vessel_type )

[0083] Based on information such as the dynamic data of ship AIS, destination port, navigation, current location, and navigation speed, the prediction model unit constructs a multi-dimensional calculation model with time serialization, operation status, and space prefabrication according to the port congestion situation information of each port obtained by the quantitative analysis unit, that is, combined with the actual congestion status of the current port and the expected volume status of the incoming port. For ships approaching the port, according to information such as ship type, load capacity, route, and speed, a prediction analysis model is constructed to predict the arrival operation and berthing operation time of the ships approaching the port, and to perform quantitative calculation of the expected arrival and port operation time, so as to quantitatively measure the stay time and operation time of the ships after arriving at the port, and to monitor the situation of ships in each port, the expected stay of ships in each port, the expected stay trend of ships in each port, the stay indicators of ships in each port, and to issue hierarchical warnings for the stay time of ships in the port.

[0084] The application service module 4 uses the calculation results of the engine service module 3, and after data screening, analysis, judgment, calculation, and processing, it functionally displays the relevant data. The display content includes: ship operation time, port operation time, ship berthing time, ship arrival information, ship expected arrival and departure time, warning information, and statistical analysis;

[0085] The interaction layer module 5 realizes the data display of the backend system, including GIS map display and mobile application support display.

[0086] The engine service module 3 also includes a data processing unit, a data cleaning unit, a data quality unit, a data thematization unit, a job scheduling unit, and an algorithm engine unit; the data processing unit, the data cleaning unit, the data quality unit, and the data thematization unit clean, process, and thematize the data stored in the data storage module 2 to obtain high-quality data with usability and processing efficiency; the job scheduling unit transfers the high-quality data and the ship quantitative congestion situation information obtained by the quantitative analysis unit to the prediction model unit for quantitatively measuring the stay time and operation time of the ship after arriving at the port; the algorithm engine unit provides algorithm support for the prediction model unit and the analysis model unit.

[0087] The key port congestion data indicator items include:

[0088] Anchored vessels (number): The number of vessels that are safely anchored in the port by using the anchoring method while waiting to berth.

[0089] Vessels expected to arrive (number): The number of vessels expected to arrive at the port within the agreed time range.

[0090] Berthing vessels (number): The number of vessels berthing at the port.

[0091] Average anchoring time (hours): The waiting time for anchoring before the vessels in the port wait to berth.

[0092] Average berthing time (hours): The time from when a vessel berths at the port to when it leaves.

[0093] Average stay time in port (hours): The total stay time within the port area, including operation time, anchoring time, and berthing time.

[0094] In the data analysis stage, by establishing a data analysis model at the port dimension, analyze the current and historical port vessel berthing and operation times to obtain the current and historical port congestion information of vessels. In the vessel prediction stage, for the vessels about to arrive at the port, based on information such as vessel type, load capacity, route, and speed, construct a prediction analysis model to predict the berthing and operation times of the vessels about to arrive at the port.

[0095] The application service module 4, through the results calculated by the engine service, after data screening, analysis, judgment, calculation, and processing, functionally displays the relevant data. It includes:

[0096] The vessel operation time includes the average vessel operation time, which scientifically determines the average port operation time of different types of vessels by using big data algorithms.

[0097] The port operation time includes the average port operation time, which analyzes the current and historical AIS+GIS data, etc., and uses big data algorithms to analyze the current and historical vessel operation times at the port to obtain information on changes in the average port operation time.

[0098] The vessel berthing time refers to the vessel's berthing time at the port, which is obtained by analyzing the historical AIS+GIS data to obtain the historical port docking data of the vessel.

[0099] The vessel arrival information uses AIS+GIS data to judge the navigation, berthing, or anchoring status of the vessel.

[0100] The expected arrival and departure times of ships are calculated by comprehensively considering the current port operation time, the expected number of ships at anchor in the port, and the deadweight information, and are used to predict the arrival and departure times of ships en route to the destination port, providing ships with predicted arrival and departure times.

[0101] The data display of the back-end system includes two ways: nautical chart display and nautical chart + list display. At the same time, according to the present invention, web GIS display and mobile GIS display capabilities can be provided (see Figures 4 - 8 ).

[0102] Figure 4 It is a monitoring chart of the situation of ships at anchor in the port. Figure 4 In it, the anchorage time ≤ 48 hours is marked green, and the number 1 in the figure represents green; 48 hours < anchorage time ≤ 72 hours is marked yellow, and the number 2 in the figure represents yellow; 72 hours < anchorage time ≤ 96 hours is marked orange, and the number 3 represents orange; the anchorage time > 96 hours is marked red, and the number 4 represents red. The figure also shows that the number of ships at anchor is 35; the number of ships expected to arrive is 24; the number of ships berthed is 12; the average anchorage time is 76 hours, which has decreased compared with before; the average stay time in the port is 89 hours, which has increased compared with before.

[0103] Figure 5 It is a monitoring chart of the situation of ships in the port. Figure 5 In it, the stay time in the port less than 3 days is marked green; 3 days < stay time in the port ≤ 4 days is marked yellow; 4 days < stay time in the port ≤ 5 days is marked orange; the stay time in the port greater than 5 days is marked red.

[0104] Figure 6 It is a monitoring chart of the expected anchorage trend of ships in the port. Figure 6 In it, the expected trend of the anchorage duration is upward, and the current anchorage duration level is marked green, and the number 1 represents green; the expected trend of the anchorage duration is downward, and the current anchorage duration level is marked yellow, and the number 2 represents yellow; the expected trend of the anchorage duration is downward, and the current anchorage duration level is marked orange, and the number 3 represents orange; the expected trend of the anchorage duration is downward, and the current anchorage duration level is marked red, and the number 4 represents red. The figure also shows that the number of ships at anchor is 35; the number of ships expected to arrive is 24; the number of ships berthed is 12; the average anchorage time is 76 hours, which has decreased compared with before; the average stay time in the port is 89 hours, which has increased compared with before.

[0105] Figure 7 It is a monitoring chart of the expected trend of ships in the port. Figure 7Among them, the expected trend of the duration in port is upward, and the current duration-in-port level indicator is green, with the number 1 representing green; when the expected trend of the duration in port is downward, the current duration-in-port level indicator is yellow, with the number 2 representing yellow; when the expected trend of the duration in port is downward, the current duration-in-port level indicator is orange, with the number 3 representing orange; when the expected trend of the duration in port is downward, the current duration-in-port level indicator is red, with the number 4 representing red. The figure also shows that the number of anchored ships is 35; the number of ships expected to arrive is 24; the number of berthed ships is 12; the average anchoring time is 76 hours, which is less than before; the average stay time in port is 89 hours, which is longer than before.

[0106] Figure 8 It is a monitoring chart of port ship stay indicators, Figure 8 showing the K-line chart of the average port berthing time of Caofeidian Port. In the figure, the red curve is the 5-day line, the dark green curve is the 10-day line, and the green curve is the 20-day line. The K-line chart shows the parameters: analysis of arrivals and departures, average port berthing time, average port anchoring time, average stay time in port, berthing situation, and anchoring situation.

[0107] In the small box in the figure, the average port berthing time value of Caofeidian Port on June 14, 2021 is shown as an example. The average value is 38.63 hours, the start (open) berthing time is 38.66 hours, the end (close) berthing time is 40.29 hours, the lowest average time is 38.02 hours, the highest average time is 40.29 hours, the 5-day line is at 37.17 hours, the 10-day line is at 34.60 hours, and the 20-day line is at 35.2 hours.

[0108] The classification definition of the real-time port congestion status of the port includes: different colored dots representing different anchoring times, and different anchoring times include: anchoring time ≤ 48 hours; 48h < anchoring time ≤ 72h; 72h < anchoring time ≤ 96h; anchoring time > 96h; the classification definition of the real-time port congestion status of the port also includes the decrease or increase of the anchoring time compared with yesterday represented by different colored arrows. The classification definition table of the real-time port congestion status of the port is shown in Figure 9 .

[0109] The classification warning of the prediction model unit refers to different warning classifications for ships in port with different deadweights. The warning classification is specifically: for ships with different deadweights, an early warning mechanism is implemented based on the duration of the ship in port (unit: days d). Different colors represent different warning levels and the range of the ship's duration in port. The classification definition table of the classification warning level is shown in Figure 10 .

[0110] The business application process of the system of the present invention is introduced as follows:

[0111] The system of the present invention realizes the dynamic ability of the port ship congestion situation at all times and the prediction ability of the expected arrival situation of a single ship, provides timely, efficient and accurate port operation efficiency information for ship management and port freight transportation enterprises, without the need for users to input data and intervene, and has the quantitative service ability of automation, intelligence and visualization. The specific service process is simple, as shown in Figure 3 .

[0112] The present invention provides monitoring and early warning services through two methods: system docking and platform access, specifically including:

[0113] System docking: The present invention can be integrated and docked with the business of the user through page embedding and data interfaces to directly provide monitoring and early warning services for it.

[0114] Platform access: The present invention can directly apply and use data through the platform visualization interface and mobile application service (H5) provided by the system.

Claims

1. A monitoring and prediction system for dynamically monitoring the port ship congestion, characterized by It is composed of a data source module (1), a data storage module (2), an engine service module (3), an application service module (4) and an interaction layer module (5); wherein, The data source module (1) provides data source data such as ship AIS, GIS, ship archive data and dynamic data of each port for data calculation and business analysis. The ship AIS data includes the running speed, status, direction and real-time number of ships approaching the port at each port. The GIS data includes basic geographic information data of China and other countries. The ship archive data includes archive data provided by HiFleet such as basic ship information, ship dimensions, MMSI change records, etc. The dynamic data of each port includes port history and real-time ship berthing, berthing and anchoring operation time data; The data storage module (2) is used to store the relevant data received by the data source module (1), establish an ODS data storage center, and an analysis database and an early warning database connected to the ODS, and output the stored data to the engine service module (3); The engine service module (3) cleans, processes and thematizes the data stored in the data storage module (2) through big data analysis. The thematization refers to thematizing the data source data so that the data source data is real-time dynamic data of ships at various ports. The processed data is used to perform real-time calculations and multi-dimensional quantitative data analysis based on the changes in the operation time of the port ship's pre-arrival, berthing, mooring and anchoring as indicators, and calculate the ship's port congestion time and the forecast of the ship's arrival and berthing operation time at various ports. The engine service module (3) includes a quantitative analysis unit, an analysis model unit and a prediction model unit. The quantitative analysis unit performs real-time calculation and analysis and summary for each port based on the ship's navigation position, navigation speed, navigation direction, and navigation destination port area, and analyzes the time changes of the ship's expected arrival at the port, berthing port, mooring port, and anchoring port as indicators, to form a multi-dimensional data analysis model with ports, ships, status, time, etc., which is stored in the analysis model unit. By establishing a multi-dimensional data analysis model for the port, calling various algorithms constructed by the analysis model unit, and combining data thematic design to fully integrate algorithms and data, the current and historical port ship berthing and operation time are analyzed to obtain current and historical quantitative port congestion information of ships, and the quantitative port congestion information includes the port congestion situation obtained according to the key port congestion data indicator items and the port real-time port congestion status classification definition; The analysis model unit stores the data analysis model constructed by the quantitative analysis unit, and constructs the estimated arrival time algorithm of the ship type, the ship berthing time analysis algorithm, and the port ship operation time analysis algorithm; The estimated arrival time algorithm is based on the dynamic L of the current voyage of the ship. curr 、Current time T curr and the destination port P, through the average ship speed V avg and the remaining distance of the flight segment D left , calculate the estimated arrival time T of the ship's current destination port P cta : T cta =T curr +(D left / V avg ) The berthing time analysis algorithm is based on the current dynamics of the ship. curr , Arrival at the destination port and start anchoring time T moor and the starting berthing time T berth and the end of berthing time T leave , calculate the anchorage time Dur of the ship at the destination port moor Duration of berthing berth ; Based on all ship types V in the port vessel_type Duration of anchorage moor Duration of berthing berth Calculate the average anchorage time PDur of port P moor {vessel_type = V vessel_type } and average berthing time PDur berth {vessel_type = V vessel_type } Dur moor =T berth -T moor Dur berth =T leave -T berth PDur moor {vessel_type=V vessel_type }=Avg(Dur moor , vessel_type=V vessel_type ) PDur berth {vessel_type=V vessel_type }=Avg(Dur berth , vessel_type=V vessel_type ) The port ship operation time analysis algorithm is based on the current dynamics of the ship L curr Duration of anchorage at the destination port moor Duration of berthing berth , calculate the ship's operating time at the destination port Dur port ; Based on the ship type V of all ships in the port vessel_type Duration of work port Calculate the average operation time PDur of port P port {vessel_type = V vessel_type } Hard port =Hard moor +Hard berth PDur port {vessel_type=V vessel_type }=Avg(Dur port , vessel_type=V vessel_type ) The prediction model unit is based on the information such as the ship's AIS dynamic data, the destination port, navigation, current position, and navigation speed, and according to the information on the port congestion of each port obtained by the quantitative analysis unit, that is, combined with the actual congestion situation of the current port and the expected volume of ports entering the port, a multi-dimensional calculation model with time serialization, operation status, and spatial prefabrication is constructed, that is, for ships that are about to arrive at the port, a prediction analysis model is constructed according to the ship type, load capacity, route, speed and other information to achieve the prediction of the port operation and berthing operation time of the ships that are about to arrive at the port, the expected quantitative calculation of the expected arrival and port operation time, the quantitative measurement of the stay time and operation time of the ship after arriving at the port, the monitoring of the port situation of ships at each port, the monitoring of the expected port stop of ships at each port, the monitoring of the expected port stop trend of ships at each port, the monitoring of the port stop index of ships at each port, and the graded warning of the ship's port time; The application service module (4) uses the calculation results of the engine service module (3) to perform functional display of relevant data after data screening, analysis, judgment, calculation and processing, and the display content includes: ship operation time, port operation time, ship berthing time, ship arrival information, ship expected entry and departure time, warning information and statistical analysis; The interaction layer module (5) realizes data display of the backend system, including GIS map display and mobile application support display.

2. A monitoring and prediction system for dynamically monitoring ship congestion in a port group according to claim 1, characterized in that The engine service module (3) also includes a data processing unit, a data cleaning unit, a data quality unit, a data thematic unit, a job scheduling unit and an algorithm engine unit; the data processing unit, the data cleaning unit, the data quality unit and the data thematic unit clean, process and thematicize the data stored in the data storage module (2) to obtain high-quality data with usability and processing efficiency; the job scheduling unit transmits the high-quality data and the quantitative port congestion information of the ship obtained by the quantitative analysis unit to the prediction model unit for quantitatively calculating the stay time and operation time of the ship after arriving at the port; the algorithm engine unit provides algorithm support for the prediction model unit and the analysis model unit.

3. The monitoring and prediction system for dynamically monitoring the ship congestion in a port group according to claim 1 or 2, characterized in that The key port pressure data indicators include: Anchored ships (ships): the number of ships waiting to berth at the port that are safely anchored by dropping anchor; Expected number of arriving ships: the number of ships expected to arrive at the port within the agreed time range; Berthed ships (ships): the number of ships berthing at the port; Average anchoring time (hours): the waiting time for ships to anchor before berthing in the port; Average berthing time (hours): the time from when a ship berths to when it leaves the port; Average stay time in port (hours): The total stay time in the port, including operation time, anchorage time and berthing time.

4. The monitoring and prediction system for dynamically monitoring the ship congestion in a port group according to claim 3 is characterized by: The real-time port congestion status classification definition includes: different colored dots represent different anchoring times, and different anchoring times include: anchoring time ≦48 hours; 48h<anchoring time ≦72h; 72h<anchoring time ≦96h; anchoring time>96h; the real-time port congestion status classification definition also includes different colored arrows representing anchoring times that are lower or higher than yesterday.

5. The monitoring and prediction system for dynamically monitoring the ship congestion in a port group according to claim 4 is characterized by: In the application service module (4), Ship operation time includes the average ship operation time, which is determined by using big data algorithms to determine the average port operation time of different types of ships; Port operation time includes the average port operation time, which is obtained by analyzing the current and historical ship operation time and historical operation time of the port based on the current and historical AIS+GIS data and using big data algorithms to obtain the change information of the average port operation time; The ship berthing time refers to the time the ship stays at the port, which is obtained by analyzing the historical AIS+GIS data and obtaining the historical port berthing data of the ship; Ship arrival information uses AIS+GIS data to determine the ship's navigation, berthing or anchoring status; The estimated arrival and departure time of a ship is a comprehensive calculation of the current port operation time, the expected number of ships berthed at the port and the deadweight information, which is used to predict the arrival and departure time of ships in transit to the destination port.

6. The system for monitoring and predicting the dynamic monitoring of ship congestion in a port group according to claim 5 is characterized by: The data display of the back-end system includes two modes: nautical map display and nautical map + list display.

7. The monitoring and prediction system for dynamically monitoring the ship congestion in a port group according to claim 6 is characterized by: The graded warning of the prediction model unit refers to different warning grades for ships of different deadweight tonnages.

8. The monitoring and prediction system for dynamically monitoring the ship congestion in a port group according to claim 7 is characterized by: The warning classification is specifically: for ships of different deadweight tonnage, an early warning mechanism is implemented based on the length of time the ship is in port, such as the number of days, and different colors represent different warning levels and the range of the length of time the ship is in port.

Images