Port attachment probability prediction method and system

Abstract

The invention discloses a port attachment probability prediction method and system. The method comprises the following steps: acquiring ship data of a plurality of candidate ships; according to the ship data, screening out a target ship berthing at a target port from the plurality of candidate ships; determining a ship type corresponding to the target ship according to the ship data corresponding to the target ship; and based on the port docking prediction model corresponding to the target port, according to the ship data and the ship type corresponding to the target ship, predicting the docking berth and the waiting time length of the target ship. Therefore, accurate berth allocation and time length prediction based on dynamic data can be realized, the port scheduling efficiency and the accuracy of ship berth management are improved, and the risk of port congestion and scheduling delay is reduced.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method and system for predicting the probability of port berthing. Background Technology

[0002] With the rapid development of the global shipping industry and the increasing demands for port management, port operators are placing greater emphasis on improving scheduling efficiency and vessel berthing management by optimizing berth allocation. Existing technologies typically determine the berth for target vessels by collecting basic data on candidate vessels, using fixed berth allocation rules or manual scheduling methods, and estimating waiting times based on historical average data to support port operations. However, existing solutions lack comprehensive analysis of vessel types and dynamic data, as well as accurate modeling of berthing prediction models. This makes it difficult to accurately predict berth demand and waiting times, and the commonly used static allocation strategies are ill-suited to complex port environments, leading to low scheduling efficiency, port congestion, or delays, and limiting the accuracy of vessel berthing management and port operational efficiency. Therefore, existing technologies have shortcomings that urgently need to be addressed. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a port berthing probability prediction method and system, which can realize accurate berth allocation and duration prediction based on dynamic data, improve port scheduling efficiency and ship berthing management accuracy, and reduce port congestion and scheduling delay risks.

[0004] To address the aforementioned technical problems, the first aspect of this invention discloses a method for predicting port berthing probabilities, the method comprising:

[0005] Obtain ship data from multiple candidate ships;

[0006] Based on the vessel data, select the target vessel to dock at the target port from the plurality of candidate vessels;

[0007] Based on the vessel data corresponding to the target vessel, determine the vessel type corresponding to the target vessel;

[0008] Based on the port berthing prediction model corresponding to the target port, the berthing berth and waiting time of the target vessel are predicted according to the vessel data and vessel type.

[0009] As an optional implementation, in the first aspect of the present invention, the ship data includes ship sensing data, ship communication data, and ship AIS data; the ship sensing data includes at least one of image data, temperature data, humidity data, sound data, and light reflection data.

[0010] As an optional implementation, in the first aspect of the invention, the step of selecting the target vessel to dock at the target port from the plurality of candidate vessels based on the vessel data includes:

[0011] For each candidate vessel, predict its future route based on its vessel data;

[0012] Based on the future vessel route, determine at least one associated vessel corresponding to the candidate vessel from the other candidate vessels;

[0013] Calculate the average of the minimum distances between the candidate vessel and the port location of the target port for the future vessel routes of all the associated vessels to obtain the berthing parameters corresponding to the candidate vessel;

[0014] Candidate vessels whose docking parameters are less than a preset threshold are selected to obtain the target vessel.

[0015] As an optional implementation, in the first aspect of the invention, predicting the future route of the candidate vessel based on its vessel data includes:

[0016] The candidate vessel's data is input into a trained route prediction model to obtain the predicted route corresponding to the candidate vessel; the route prediction model is trained using a training dataset that includes data from multiple training vessels and corresponding route labels.

[0017] Determine whether the candidate vessel's vessel data contains preset route plan information to obtain the first judgment result;

[0018] If the first judgment result is negative, the portion of the predicted route after the current position of the candidate vessel is determined as the future route of the candidate vessel.

[0019] If the first judgment result is yes, the planned vessel route of the candidate vessel is determined according to the route plan information;

[0020] Calculate the difference between the planned vessel route and the predicted route, and determine whether the difference is greater than a preset difference threshold to obtain a second determination result;

[0021] If the second judgment result is negative, the portion of the planned vessel route after the current position of the candidate vessel shall be determined as the future vessel route of the candidate vessel.

[0022] If the second judgment result is yes, calculate the weighted average result of the planned vessel route and the predicted route to obtain the calculated route, and determine the portion of the route after the current position of the candidate vessel in the calculated route as the future vessel route of the candidate vessel.

[0023] As an optional implementation, in the first aspect of the invention, determining at least one associated vessel corresponding to a candidate vessel from among the other candidate vessels based on the future vessel route includes:

[0024] A mathematical model is used to calculate the relationship between the distance of the candidate vessel from each waypoint on the future vessel route and the target port as a function of time.

[0025] The coefficient parameters in the mathematical relationship model are determined as the route coefficient parameters of the candidate vessel;

[0026] Calculate the parameter similarity between the route coefficient parameters of each of the other candidate vessels and the route coefficient parameters of the candidate vessel;

[0027] From all other candidate ships, ships whose parameter similarity is greater than a preset similarity threshold are selected to obtain at least one associated ship corresponding to the candidate ship.

[0028] As an optional implementation, in the first aspect of the invention, determining the vessel type corresponding to the target vessel based on the vessel data corresponding to the target vessel includes:

[0029] The ship data corresponding to the target ship is input into the trained ship type recognition model to obtain the ship type corresponding to the target ship; the ship type is a conventional ship, a special ship, or a special pulp ship; the ship type recognition model is trained by a training dataset that includes multiple training ship data and corresponding ship type labels.

[0030] As an optional implementation, in the first aspect of the present invention, the step of predicting the berth and waiting time of the target vessel based on the port berthing prediction model corresponding to the target port, according to the vessel data and vessel type corresponding to the target vessel, includes:

[0031] Based on the vessel data of the target vessel, determine the predicted arrival time of the target vessel;

[0032] The vessel data and vessel type corresponding to the target vessel are input into the port berthing prediction model corresponding to the target port to obtain the time interval berthing prediction distribution corresponding to the target vessel; the time interval berthing prediction distribution includes the predicted berthing berths and predicted waiting time for multiple time intervals;

[0033] Based on the predicted berthing distribution over the time interval, the predicted berthing berth and predicted waiting time of the target vessel in the time interval of the predicted arrival time are determined, thus obtaining the berthing berth and waiting time of the target vessel.

[0034] As an optional implementation, in the first aspect of the invention, the port berthing prediction model is trained through the following steps:

[0035] Obtain historical vessel arrival data for the target port from the port terminal database;

[0036] Based on the historical ship arrival data, a multidimensional dataset is constructed;

[0037] The data in the multidimensional dataset is divided into multiple time interval datasets by using a time-segmented statistical method to divide the data into parking duration intervals.

[0038] For each time interval dataset, berth statistics and waiting time statistics are performed to obtain the data annotation corresponding to the multidimensional dataset;

[0039] The multidimensional dataset and corresponding data labels are input into a preset prediction model for iterative training to obtain the port berthing prediction model.

[0040] A second aspect of this invention discloses a port call probability prediction system, the system comprising:

[0041] The acquisition module is used to acquire ship data from multiple candidate ships;

[0042] The filtering module is used to filter out the target vessel to dock at the target port from the plurality of candidate vessels based on the vessel data;

[0043] The determination module is used to determine the type of the target vessel based on the vessel data corresponding to the target vessel.

[0044] The prediction module is used to predict the berthing berth and waiting time of the target vessel based on the port berthing prediction model corresponding to the target port and according to the vessel data and vessel type of the target vessel.

[0045] As an optional implementation, in a second aspect of the invention, the ship data includes ship sensor data, ship communication data, and ship AIS data; the ship sensor data includes at least one of image data, temperature data, humidity data, sound data, and light reflection data.

[0046] As an optional implementation, in a second aspect of the invention, the specific method by which the filtering module selects the target vessel to dock at the target port from the plurality of candidate vessels based on the vessel data includes:

[0047] For each candidate vessel, predict its future route based on its vessel data;

[0048] Based on the future vessel route, determine at least one associated vessel corresponding to the candidate vessel from the other candidate vessels;

[0049] Calculate the average of the minimum distances between the candidate vessel and the port location of the target port for the future vessel routes of all the associated vessels to obtain the berthing parameters corresponding to the candidate vessel;

[0050] Candidate vessels whose docking parameters are less than a preset threshold are selected to obtain the target vessel.

[0051] As an optional implementation, in a second aspect of the invention, the specific method by which the screening module predicts the future route of a candidate vessel based on the vessel data of the candidate vessel includes:

[0052] The candidate vessel's data is input into a trained route prediction model to obtain the predicted route corresponding to the candidate vessel; the route prediction model is trained using a training dataset that includes data from multiple training vessels and corresponding route labels.

[0053] Determine whether the candidate vessel's vessel data contains preset route plan information to obtain the first judgment result;

[0054] If the first judgment result is negative, the portion of the predicted route after the current position of the candidate vessel is determined as the future route of the candidate vessel.

[0055] If the first judgment result is yes, the planned vessel route of the candidate vessel is determined according to the route plan information;

[0056] Calculate the difference between the planned vessel route and the predicted route, and determine whether the difference is greater than a preset difference threshold to obtain a second determination result;

[0057] If the second judgment result is negative, the portion of the planned vessel route after the current position of the candidate vessel shall be determined as the future vessel route of the candidate vessel.

[0058] If the second judgment result is yes, calculate the weighted average result of the planned vessel route and the predicted route to obtain the calculated route, and determine the portion of the route after the current position of the candidate vessel in the calculated route as the future vessel route of the candidate vessel.

[0059] As an optional implementation, in a second aspect of the invention, the specific method by which the screening module determines at least one associated vessel corresponding to a candidate vessel from among the other candidate vessels based on the future vessel route includes:

[0060] A mathematical model is used to calculate the relationship between the distance of the candidate vessel from each waypoint on the future vessel route and the target port as a function of time.

[0061] The coefficient parameters in the mathematical relationship model are determined as the route coefficient parameters of the candidate vessel;

[0062] Calculate the parameter similarity between the route coefficient parameters of each of the other candidate vessels and the route coefficient parameters of the candidate vessel;

[0063] From all other candidate ships, ships whose parameter similarity is greater than a preset similarity threshold are selected to obtain at least one associated ship corresponding to the candidate ship.

[0064] As an optional implementation, in a second aspect of the invention, the specific method by which the determining module determines the vessel type corresponding to the target vessel based on the vessel data corresponding to the target vessel includes:

[0065] The ship data corresponding to the target ship is input into the trained ship type recognition model to obtain the ship type corresponding to the target ship; the ship type is a conventional ship, a special ship, or a special pulp ship; the ship type recognition model is trained by a training dataset that includes multiple training ship data and corresponding ship type labels.

[0066] As an optional implementation, in a second aspect of the invention, the prediction module, based on a port berthing prediction model corresponding to the target port, predicts the specific method by which it predicts the berthing berth and waiting time of the target vessel according to the vessel data and vessel type corresponding to the target vessel, including:

[0067] Based on the vessel data of the target vessel, determine the predicted arrival time of the target vessel;

[0068] The vessel data and vessel type corresponding to the target vessel are input into the port berthing prediction model corresponding to the target port to obtain the time interval berthing prediction distribution corresponding to the target vessel; the time interval berthing prediction distribution includes the predicted berthing berths and predicted waiting time for multiple time intervals;

[0069] Based on the predicted berthing distribution over the time interval, the predicted berthing berth and predicted waiting time of the target vessel in the time interval of the predicted arrival time are determined, thus obtaining the berthing berth and waiting time of the target vessel.

[0070] As an optional implementation, in a second aspect of the invention, the port berthing prediction model is trained through the following steps:

[0071] Obtain historical vessel arrival data for the target port from the port terminal database;

[0072] Based on the historical ship arrival data, a multidimensional dataset is constructed;

[0073] The data in the multidimensional dataset is divided into multiple time interval datasets by using a time-segmented statistical method to divide the data into parking duration intervals.

[0074] For each time interval dataset, berth statistics and waiting time statistics are performed to obtain the data annotation corresponding to the multidimensional dataset;

[0075] The multidimensional dataset and corresponding data labels are input into a preset prediction model for iterative training to obtain the port berthing prediction model.

[0076] A third aspect of this invention discloses another port call probability prediction system, the system comprising:

[0077] Memory containing executable program code;

[0078] A processor coupled to the memory;

[0079] The processor calls the executable program code stored in the memory to execute some or all of the steps in the port berthing probability prediction method disclosed in the first aspect of the present invention.

[0080] The fourth aspect of the present invention discloses a computer storage medium storing computer instructions, which, when invoked, are used to execute some or all of the steps in the port berthing probability prediction method disclosed in the first aspect of the present invention.

[0081] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0082] This invention acquires vessel data from multiple candidate vessels and filters target vessels for berthing at a target port. By combining vessel type identification and a port berthing prediction model, it predicts berth locations and waiting times, thereby enabling accurate berth allocation and waiting time prediction based on dynamic data. This improves port scheduling efficiency and the accuracy of vessel berthing management, and reduces the risk of port congestion and scheduling delays. Attached Figure Description

[0083] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0084] Figure 1 This is a flowchart illustrating a port berthing probability prediction method disclosed in an embodiment of the present invention.

[0085] Figure 2 This is a schematic diagram of the structure of a port berthing probability prediction system disclosed in an embodiment of the present invention.

[0086] Figure 3 This is a schematic diagram of another port berthing probability prediction system disclosed in an embodiment of the present invention.

[0087] Figure 4 This is a schematic diagram of the visual analysis data output of a port berthing probability prediction system disclosed in an embodiment of the present invention.

[0088] Figure 5 This is a schematic diagram of the analysis and training data of a port berthing probability prediction model disclosed in an embodiment of the present invention. Detailed Implementation

[0089] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0090] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0091] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0092] This invention discloses a port berthing probability prediction method and system. By acquiring vessel data from multiple candidate vessels and filtering for target vessels berthing at the desired port, and combining vessel type identification with a port berthing prediction model, it predicts berth locations and waiting times. This enables accurate berth allocation and waiting time prediction based on dynamic data, improving port scheduling efficiency and vessel berthing management accuracy, and reducing port congestion and scheduling delay risks. Detailed explanations follow.

[0093] Example 1

[0094] Please see Figure 1 , Figure 1 This is a flowchart illustrating a port berthing probability prediction method disclosed in an embodiment of the present invention. Figure 1 The described port call probability prediction method can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). For example... Figure 1 As shown, the port call probability prediction method may include the following operations:

[0095] 101. Obtain ship data for multiple candidate ships.

[0096] 102. Based on the vessel data, select the target vessel to dock at the target port from multiple candidate vessels.

[0097] 103. Determine the ship type corresponding to the target ship based on the ship data corresponding to the target ship.

[0098] 104. Based on the port berthing prediction model corresponding to the target port, predict the berthing berth and waiting time of the target vessel according to the vessel data and vessel type.

[0099] Optionally, the port berthing prediction model can be a deep learning model, a statistical model, or a time series prediction model; this invention does not impose any limitations.

[0100] Optionally, the berth can be a specific berth number, berth area, or berth coordinates; this invention does not impose any limitations.

[0101] Optionally, the berthing time can be in the range of minutes, hours, or days, and this invention does not limit it.

[0102] As can be seen, the above-described embodiments of the invention acquire ship data from multiple candidate ships and filter target ships that will berth at the target port. By combining ship type identification and port berthing prediction models to predict berthing positions and waiting times, the invention can achieve accurate berth allocation and waiting time prediction based on dynamic data, thereby improving port scheduling efficiency and the accuracy of ship berthing management, and reducing the risk of port congestion and scheduling delays.

[0103] As an optional embodiment, in the above steps, the ship data includes ship sensor data, ship communication data, and ship AIS data; the ship sensor data includes at least one of image data, temperature data, humidity data, sound data, and light reflection data.

[0104] Optionally, the vessel data may include real-time data, historical data, or predicted data, and the present invention does not limit it.

[0105] Optionally, the vessel's sensor data can be acquired by a camera, temperature sensor, humidity sensor, acoustic sensor, or photoelectric sensor; this invention does not limit the scope of the data acquisition.

[0106] Optionally, the ship's communication data may include radio communication data, satellite communication data, or network transmission data.

[0107] As can be seen, the above optional embodiments limit the details of the vessel data to comprehensively characterize the vessel's route status, assist in achieving accurate berth allocation and duration prediction based on dynamic data, improve port scheduling efficiency and the accuracy of vessel berthing management, and reduce the risk of port congestion and scheduling delays.

[0108] As an optional embodiment, the step described above, selecting the target vessel to dock at the target port from multiple candidate vessels based on vessel data, includes:

[0109] For each candidate vessel, predict its future route based on its vessel data.

[0110] Based on the future vessel route, identify at least one associated vessel corresponding to the candidate vessel from among other candidate vessels;

[0111] Calculate the average of the minimum distances between the future vessel routes and the port locations of the target port for the candidate vessel and all associated vessels to obtain the berthing parameters for the candidate vessel.

[0112] Candidate vessels whose docking parameters are less than the preset parameter threshold are selected to obtain the target vessel.

[0113] Optionally, the future ship route can be a predicted path, a sequence of route points, or a probability distribution path; this invention does not impose any limitations.

[0114] Optionally, the prediction process for future ship routes can be based on time series analysis, machine learning models, or route planning algorithms, and this invention does not limit the scope of the prediction.

[0115] Optionally, the prediction of future ship routes can be optimized by combining weather data or sea state data, and this invention does not limit it.

[0116] Optionally, the associated vessel can be a vessel with a similar route, a vessel sailing in coordination, or a vessel with the same goal; this invention does not impose any limitations.

[0117] Optionally, the process of determining the associated vessels can be based on route similarity, collaborative analysis, or clustering algorithms, and this invention does not limit this.

[0118] Optionally, the threshold parameter can be a fixed threshold, a dynamic threshold, or a threshold based on port capacity adjustment; this invention does not impose any limitations.

[0119] As can be seen, through the above optional embodiments, by predicting future routes and identifying associated vessels based on the vessel data of each candidate vessel, calculating the average of the minimum distances between the future routes of the candidate vessels and associated vessels and the target port location as berthing parameters, and screening candidate vessels with berthing parameters less than a threshold as target vessels, accurate target vessel screening based on route prediction and association analysis can be achieved, thereby improving the accuracy and scheduling efficiency of port berthing management and reducing the risk of resource waste caused by misselection of vessels.

[0120] As an optional embodiment, the step of predicting the future route of the candidate vessel based on its vessel data, as described above, includes:

[0121] The vessel data of the candidate vessel is input into the trained route prediction model to obtain the predicted route corresponding to the candidate vessel; optionally, the route prediction model is trained using a training dataset that includes data of multiple training vessels and corresponding route labels.

[0122] Determine whether the candidate vessel's vessel data contains preset route plan information to obtain the first judgment result;

[0123] If the first judgment result is negative, the portion of the route after the current position of the candidate vessel in the predicted route will be determined as the future route of the candidate vessel.

[0124] If the first judgment result is yes, the planned vessel route of the candidate vessel is determined based on the route plan information;

[0125] Calculate the difference between the planned vessel route and the predicted route, and determine whether the difference is greater than a preset difference threshold to obtain a second judgment result;

[0126] If the second judgment result is negative, the portion of the planned vessel route after the current position of the candidate vessel will be determined as the future vessel route of the candidate vessel.

[0127] If the second judgment result is yes, calculate the weighted average result of the planned ship route and the predicted route to obtain the calculated route, and determine the part of the route after the current position of the candidate ship in the calculated route as the future ship route of the candidate ship.

[0128] Optionally, the route prediction model can be an LSTM neural network, a Transformer model, or a regression model; this invention does not impose any limitations.

[0129] Optionally, the predicted route can be a continuous path, a discrete point sequence, or a probability distribution path; this invention does not impose any limitations.

[0130] Optionally, the training dataset may include historical flight route data, simulated flight route data, or real-time collected data; this invention does not impose any limitations.

[0131] Optionally, the route plan information can be a preset route plan, a vessel scheduling table, or port instructions; this invention does not impose any limitations.

[0132] Optionally, the current location can be GPS coordinates, AIS positioning point, or reference point; this invention does not impose any limitations on this.

[0133] Optionally, the determination of the future ship route can be based on time truncation, path segmentation, or predictive pruning, and this invention does not limit this.

[0134] Optionally, the difference can be a path distance difference, angle deviation, or distribution difference, and the present invention does not limit it.

[0135] Optionally, the difference threshold can be a fixed threshold, a dynamic threshold, or a threshold adjusted based on route complexity; this invention does not impose any limitations.

[0136] As can be seen, through the above optional embodiments, by inputting the ship data of candidate ships into the trained route prediction model to predict the route, and combining the existence of route plan information and the degree of difference between the route and the predicted route, the future ship route is determined by using the predicted route, the planned route, or a weighted average of the two, thereby achieving accurate route prediction based on model prediction and plan correction, improving the accuracy of target ship selection and port scheduling efficiency, and reducing the risk of berthing misjudgment caused by route deviation.

[0137] As an optional embodiment, the step described above, determining at least one associated vessel corresponding to the candidate vessel from other candidate vessels based on the future vessel route, includes:

[0138] A mathematical model for calculating the relationship between the distance to the target port from each waypoint on the future route of the candidate vessel and the time of the waypoint.

[0139] The coefficient parameters in the mathematical relationship model are determined as the route coefficient parameters of the candidate ship.

[0140] Calculate the parameter similarity between the route coefficient parameters of each other candidate vessel and the route coefficient parameters of the candidate vessel;

[0141] From all other candidate ships, select ships whose parameter similarity is greater than a preset similarity threshold, and obtain at least one associated ship corresponding to that candidate ship.

[0142] Optionally, the mathematical relationship model can be a linear model, a quadratic model, or an exponential model; this invention does not specify otherwise.

[0143] Optionally, the coefficient parameter can be a slope coefficient, a quadratic term coefficient, or a higher-order term coefficient; this invention does not impose any limitations.

[0144] Optionally, the similarity parameter can be cosine similarity, Euclidean distance, or Pearson correlation coefficient; this invention does not impose any limitation on it.

[0145] Optionally, the calculation of the similarity of this parameter can be based on vector comparison, statistical analysis or feature matching, and this invention does not limit it.

[0146] Optionally, the similarity threshold can be a fixed threshold, a dynamic threshold, or a threshold adjusted based on route characteristics; this invention does not impose any limitations.

[0147] As can be seen, through the above optional embodiments, by calculating the mathematical relationship model of the distance between each route point in the future route of the candidate vessel and the target port as a function of time, the model coefficients are extracted as route coefficient parameters, the similarity between the route coefficient parameters of other candidate vessels and these parameters is calculated, and vessels with similarity exceeding the threshold are selected as associated vessels. This achieves accurate identification of associated vessels based on the similarity of route features, improves the accuracy of target vessel selection and port scheduling efficiency, and reduces the risk of resource waste caused by misselected vessels.

[0148] As an optional embodiment, the step of determining the vessel type corresponding to the target vessel based on the vessel data corresponding to the target vessel in the above steps includes:

[0149] The ship data corresponding to the target ship is input into the trained ship type recognition model to obtain the ship type corresponding to the target ship; optionally, the ship type is a conventional ship, a special ship, or a special pulp ship; the ship type recognition model is trained by a training dataset that includes multiple training ship data and corresponding ship type labels.

[0150] Optionally, the vessel type identification model can be a convolutional neural network, a random forest model, or a support vector machine model; this invention does not impose any limitations.

[0151] Optionally, the vessel type may include additional classifications, such as cargo ships, passenger ships, or military ships, which are not limited by this invention.

[0152] Optionally, the training dataset may include historical ship data, simulated data, or labeled data, and this invention does not impose any limitations.

[0153] As can be seen, through the above optional embodiments, by inputting the target vessel's vessel data into the trained vessel type recognition model to determine the vessel type, data-driven accurate vessel type classification is achieved, thereby improving the accuracy of port berthing prediction and berth allocation efficiency, and reducing scheduling risks caused by type misjudgment.

[0154] As an optional embodiment, the above steps, based on the port berthing prediction model corresponding to the target port, predict the berthing berth and waiting time of the target vessel according to the vessel data and vessel type, include:

[0155] Based on the target vessel's vessel data, determine the predicted arrival time of the target vessel.

[0156] Input the vessel data and vessel type corresponding to the target vessel into the port berthing prediction model corresponding to the target port to obtain the time interval berthing prediction distribution corresponding to the target vessel; optionally, the time interval berthing prediction distribution includes the predicted berthing berths and predicted waiting time for multiple time intervals.

[0157] Based on the predicted berthing distribution over the time interval, the predicted berthing berth and predicted waiting time of the target vessel in the time interval of the predicted arrival time are determined, thus obtaining the berthing berth and waiting time of the target vessel.

[0158] Optionally, the predicted arrival time can be a specific time point, a time range, or a probability distribution time; this invention does not impose any limitations.

[0159] Optionally, the port berthing prediction model can be a deep learning model, a regression model, or a probability model; this invention does not impose any limitations.

[0160] Optionally, the stop prediction distribution for the time interval can be a discrete distribution, a continuous distribution, or a mixed distribution; this invention does not impose any limitations.

[0161] As can be seen, through the above optional embodiments, the predicted arrival time is determined based on the vessel data of the target vessel, and the vessel data and vessel type are input into the port berthing prediction model of the target port to obtain the berthing prediction distribution for the time interval. The berthing berth and waiting time are determined based on the prediction results of the time interval in which the predicted arrival time is located. This achieves accurate berth allocation and time prediction based on arrival time and prediction distribution, improves port scheduling efficiency and the accuracy of vessel berthing management, and reduces the risk of berth conflicts and waiting delays.

[0162] As an optional embodiment, the port berthing prediction model in the above steps is trained through the following steps:

[0163] Obtain historical vessel arrival data for the target port from the port terminal database;

[0164] A multidimensional dataset was constructed based on historical ship arrival data.

[0165] The data in the multidimensional dataset was divided into multiple time interval datasets by using a time-segmented statistical method to divide the data into parking duration intervals.

[0166] For each time interval dataset, berth statistics and waiting time statistics are performed to obtain the data annotations corresponding to the multidimensional dataset;

[0167] The multidimensional dataset and corresponding data labels are input into the preset prediction model for iterative training to obtain the port berthing prediction model.

[0168] Optionally, the port terminal database can be a local database, a cloud database, or a distributed database; this invention does not impose any limitations.

[0169] Optionally, the historical vessel arrival data may include arrival time, berth allocation, or waiting time records, which are not limited in this invention.

[0170] Optionally, the multidimensional dataset may include a time dimension, a vessel dimension, or a berth dimension; this invention does not impose any limitations.

[0171] Optionally, the time-segmented statistical method can be divided into equal intervals, dynamic intervals, or based on traffic distribution; this invention does not impose any limitations.

[0172] Optionally, this iterative training can be implemented based on gradient descent, stochastic optimization, or adaptive optimization algorithms, and this invention does not limit it.

[0173] Optionally, the training of the port berthing prediction model can be optimized by combining real-time data feedback or simulation data, and this invention does not limit it.

[0174] In a specific implementation plan, AIS data on ship arrivals can be obtained in real time from a port terminal database to establish a multidimensional dataset containing the distribution characteristics of ship berthing durations. A dynamic probability calculation model based on time windows can be constructed, and berthing durations can be divided into intervals using a time-segmented statistical method. Ship identity information can be integrated through multi-source data fusion technology to establish a specific ship type identification mechanism. A dual-channel probability calculation engine can be designed to handle the berthing pattern analysis of regular ships and special ships respectively. A dynamic weighted algorithm can be used to generate berthing duration probability distribution maps for different ship categories, and the obtained data can be used as the training dataset for the port berthing prediction model to train the model.

[0175] More specifically, the aforementioned data processing is implemented through a single system, including a data acquisition module, a feature processing module, a probability calculation engine, and an API service gateway. The data acquisition module interfaces with the port database and the vessel AIS system; the feature processing module standardizes the time format and classifies vessel types; the probability calculation engine performs time-segmented statistics and probability distribution calculations; and the API service gateway provides encrypted data transmission and a results visualization interface. It innovatively employs dynamic query technology based on SQL templates to achieve adaptive data acquisition for different port codes and introduces a real-time vessel information verification mechanism to ensure data validity.

[0176] Specifically, based on the historical vessel berthing records of each port, the system will display the number of berthings and corresponding probabilities of all historical vessels, all special transport vessels, and special transport pulp ships within different time ranges under different waiting times.

[0177] Specifically, users can select the statistical time range and view the corresponding statistical tables and charts for intuitive data viewing.

[0178] Optionally, the statistical tables and charts output by this system can be referenced. Figure 4It enables the output of historical berth statistics for port congestion prediction, allowing users to select the statistical time range and generate corresponding statistical tables and charts.

[0179] Specifically, the probabilistic statistical model of this scheme statistically analyzes port call data for a specific month of a specific year, and can obtain similar results. Figure 5 The model table is used as input data for the prediction model. Specifically, each row of the table is explained as follows:

[0180] First row (number of vessels on the day of the previous 7 days): Total number of vessels in port and vessels expected to arrive on the day of the previous 7 days (sample density divided into intervals);

[0181] The second line (direct berthing ratio of specified berths): Statistics on the two berths with the most historical special transport pulp ships berthing, where all berthing vessels fall within the range, and the percentage of berthing times with a waiting time of 0 is out of all berthing times;

[0182] The third line (average anchorage time at the specified berth (h)): Select the two berths with the most historical special transport pulp ships to calculate the average waiting time for the ship in different intervals;

[0183] The fourth row (80% confidence interval): Calculates the 80% confidence interval within different intervals.

[0184] Specifically, after each port compiles the above model table, it retrieves the total number of vessels in port and expected to arrive on the previous seven days. Based on the direct berthing ratio in the table, it calculates the direct berthing probability. Based on the historical average waiting time in the table, it calculates the confidence interval range based on the 80% confidence interval within the interval, in order to obtain the analysis data for the historical time period.

[0185] As can be seen, through the above optional embodiments, by obtaining historical ship arrival data of the target port from the port terminal database and constructing a multidimensional dataset, using a time-segmented statistical method to divide the berthing duration intervals to obtain multiple time interval datasets, statistically analyzing the berthing berths and waiting times of each dataset to generate data labels, and iteratively training a preset prediction model to obtain a port berthing prediction model, a precise prediction model based on historical data and time segmentation can be trained, thereby improving the accuracy of the prediction of the target ship's berthing berths and waiting times, the efficiency of port scheduling, and reducing the risk of errors in berth allocation and duration prediction.

[0186] Example 2

[0187] Please see Figure 2 , Figure 2 This is a schematic diagram of a port berthing probability prediction system disclosed in an embodiment of the present invention. Figure 2The described port call probability prediction system can be applied to data processing systems / data processing equipment / data processing servers (wherein, the server includes local processing servers or cloud processing servers). For example... Figure 2 As shown, the port call probability prediction system may include:

[0188] The acquisition module 201 is used to acquire ship data from multiple candidate ships.

[0189] The filtering module 202 is used to filter out the target vessel that will dock at the target port from multiple candidate vessels based on the vessel data.

[0190] The determination module 203 is used to determine the ship type corresponding to the target ship based on the ship data corresponding to the target ship.

[0191] The prediction module 204 is used to predict the berthing berth and waiting time of the target vessel based on the port berthing prediction model corresponding to the target port, according to the vessel data and vessel type of the target vessel.

[0192] As can be seen, the above-described embodiments of the invention acquire ship data from multiple candidate ships and filter target ships that will berth at the target port. By combining ship type identification and port berthing prediction models to predict berthing positions and waiting times, the invention can achieve accurate berth allocation and waiting time prediction based on dynamic data, thereby improving port scheduling efficiency and the accuracy of ship berthing management, and reducing the risk of port congestion and scheduling delays.

[0193] As an optional embodiment, the ship data includes ship sensor data, ship communication data, and ship AIS data; the ship sensor data includes at least one of image data, temperature data, humidity data, sound data, and light reflection data.

[0194] As can be seen, the above optional embodiments limit the details of the vessel data to comprehensively characterize the vessel's route status, assist in achieving accurate berth allocation and duration prediction based on dynamic data, improve port scheduling efficiency and the accuracy of vessel berthing management, and reduce the risk of port congestion and scheduling delays.

[0195] As an optional embodiment, the filtering module selects the target vessel to dock at the target port from multiple candidate vessels based on vessel data, including the following specific methods:

[0196] For each candidate vessel, predict its future route based on its vessel data.

[0197] Based on the future vessel route, identify at least one associated vessel corresponding to the candidate vessel from among other candidate vessels;

[0198] Calculate the average of the minimum distances between the future vessel routes and the port locations of the target port for the candidate vessel and all associated vessels to obtain the berthing parameters for the candidate vessel.

[0199] Candidate vessels whose docking parameters are less than the preset parameter threshold are selected to obtain the target vessel.

[0200] As can be seen, through the above optional embodiments, by predicting future routes and identifying associated vessels based on the vessel data of each candidate vessel, calculating the average of the minimum distances between the future routes of the candidate vessels and associated vessels and the target port location as berthing parameters, and screening candidate vessels with berthing parameters less than a threshold as target vessels, accurate target vessel screening based on route prediction and association analysis can be achieved, thereby improving the accuracy and scheduling efficiency of port berthing management and reducing the risk of resource waste caused by misselection of vessels.

[0201] As an optional embodiment, the filtering module predicts the future route of a candidate vessel based on its vessel data in the following ways:

[0202] The vessel data of the candidate vessel is input into the trained route prediction model to obtain the predicted route corresponding to the candidate vessel; optionally, the route prediction model is trained using a training dataset that includes data of multiple training vessels and corresponding route labels.

[0203] Determine whether the candidate vessel's vessel data contains preset route plan information to obtain the first judgment result;

[0204] If the first judgment result is negative, the portion of the route after the current position of the candidate vessel in the predicted route will be determined as the future route of the candidate vessel.

[0205] If the first judgment result is yes, the planned vessel route of the candidate vessel is determined based on the route plan information;

[0206] Calculate the difference between the planned vessel route and the predicted route, and determine whether the difference is greater than a preset difference threshold to obtain a second judgment result;

[0207] If the second judgment result is negative, the portion of the planned vessel route after the current position of the candidate vessel will be determined as the future vessel route of the candidate vessel.

[0208] If the second judgment result is yes, calculate the weighted average result of the planned ship route and the predicted route to obtain the calculated route, and determine the part of the route after the current position of the candidate ship in the calculated route as the future ship route of the candidate ship.

[0209] As can be seen, through the above optional embodiments, by inputting the ship data of candidate ships into the trained route prediction model to predict the route, and combining the existence of route plan information and the degree of difference between the route and the predicted route, the future ship route is determined by using the predicted route, the planned route, or a weighted average of the two, thereby achieving accurate route prediction based on model prediction and plan correction, improving the accuracy of target ship selection and port scheduling efficiency, and reducing the risk of berthing misjudgment caused by route deviation.

[0210] As an optional embodiment, the filtering module determines the specific method by which it identifies at least one associated vessel corresponding to a candidate vessel from other candidate vessels based on the future vessel route, including:

[0211] A mathematical model for calculating the relationship between the distance to the target port from each waypoint on the future route of the candidate vessel and the time of the waypoint.

[0212] The coefficient parameters in the mathematical relationship model are determined as the route coefficient parameters of the candidate ship.

[0213] Calculate the parameter similarity between the route coefficient parameters of each other candidate vessel and the route coefficient parameters of the candidate vessel;

[0214] From all other candidate ships, select ships whose parameter similarity is greater than a preset similarity threshold, and obtain at least one associated ship corresponding to that candidate ship.

[0215] As can be seen, through the above optional embodiments, by calculating the mathematical relationship model of the distance between each route point in the future route of the candidate vessel and the target port as a function of time, the model coefficients are extracted as route coefficient parameters, the similarity between the route coefficient parameters of other candidate vessels and these parameters is calculated, and vessels with similarity exceeding the threshold are selected as associated vessels. This achieves accurate identification of associated vessels based on the similarity of route features, improves the accuracy of target vessel selection and port scheduling efficiency, and reduces the risk of resource waste caused by misselected vessels.

[0216] As an optional embodiment, the determining module determines the specific method by which it determines the ship type corresponding to the target ship based on the ship data corresponding to the target ship, including:

[0217] The ship data corresponding to the target ship is input into the trained ship type recognition model to obtain the ship type corresponding to the target ship; optionally, the ship type is a conventional ship, a special ship, or a special pulp ship; the ship type recognition model is trained by a training dataset that includes multiple training ship data and corresponding ship type labels.

[0218] As can be seen, through the above optional embodiments, by inputting the target vessel's vessel data into the trained vessel type recognition model to determine the vessel type, data-driven accurate vessel type classification is achieved, thereby improving the accuracy of port berthing prediction and berth allocation efficiency, and reducing scheduling risks caused by type misjudgment.

[0219] As an optional embodiment, the prediction module, based on the port berthing prediction model corresponding to the target port, predicts the specific method of the target vessel's berthing location and waiting time according to the vessel data and vessel type, including:

[0220] Based on the target vessel's vessel data, determine the predicted arrival time of the target vessel.

[0221] Input the vessel data and vessel type corresponding to the target vessel into the port berthing prediction model corresponding to the target port to obtain the time interval berthing prediction distribution corresponding to the target vessel; optionally, the time interval berthing prediction distribution includes the predicted berthing berths and predicted waiting time for multiple time intervals.

[0222] Based on the predicted berthing distribution over the time interval, the predicted berthing berth and predicted waiting time of the target vessel in the time interval of the predicted arrival time are determined, thus obtaining the berthing berth and waiting time of the target vessel.

[0223] As can be seen, through the above optional embodiments, the predicted arrival time is determined based on the vessel data of the target vessel, and the vessel data and vessel type are input into the port berthing prediction model of the target port to obtain the berthing prediction distribution for the time interval. The berthing berth and waiting time are determined based on the prediction results of the time interval in which the predicted arrival time is located. This achieves accurate berth allocation and time prediction based on arrival time and prediction distribution, improves port scheduling efficiency and the accuracy of vessel berthing management, and reduces the risk of berth conflicts and waiting delays.

[0224] As an optional implementation, the port berthing prediction model is trained through the following steps:

[0225] Obtain historical vessel arrival data for the target port from the port terminal database;

[0226] A multidimensional dataset was constructed based on historical ship arrival data.

[0227] The data in the multidimensional dataset was divided into multiple time interval datasets by using a time-segmented statistical method to divide the data into parking duration intervals.

[0228] For each time interval dataset, berth statistics and waiting time statistics are performed to obtain the data annotations corresponding to the multidimensional dataset;

[0229] The multidimensional dataset and corresponding data labels are input into the preset prediction model for iterative training to obtain the port berthing prediction model.

[0230] As can be seen, through the above optional embodiments, by obtaining historical ship arrival data of the target port from the port terminal database and constructing a multidimensional dataset, using a time-segmented statistical method to divide the berthing duration intervals to obtain multiple time interval datasets, statistically analyzing the berthing berths and waiting times of each dataset to generate data labels, and iteratively training a preset prediction model to obtain a port berthing prediction model, a precise prediction model based on historical data and time segmentation can be trained, thereby improving the accuracy of the prediction of the target ship's berthing berths and waiting times, the efficiency of port scheduling, and reducing the risk of errors in berth allocation and duration prediction.

[0231] Example 3

[0232] Please see Figure 3 , Figure 3 This is another port berthing probability prediction system disclosed in the embodiments of the present invention. Figure 3 The described port call probability prediction system is applied in a data processing system / data processing equipment / data processing server (wherein, the server includes a local processing server or a cloud processing server). For example... Figure 3 As shown, the port call probability prediction system may include:

[0233] Memory 301 storing executable program code;

[0234] Processor 302 coupled to memory 301;

[0235] The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the port berthing probability prediction method described in Embodiment 1.

[0236] Example 4

[0237] This invention discloses a computer read storage medium that stores a computer program for electronic data interchange, wherein the computer program causes a computer to execute the steps of the port call probability prediction method described in Embodiment 1.

[0238] Example 5

[0239] This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps of the port call probability prediction method described in Embodiment 1.

[0240] The foregoing has described specific embodiments of this specification; other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims may be performed in a different order than those shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily have to follow the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0241] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0242] For ease of description, the above devices are described in terms of function, divided into various units. Of course, in implementing this specification, the functions of each unit can be implemented in one or more software and / or hardware components.

[0243] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0244] This specification is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this specification. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0245] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0246] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0247] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0248] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0249] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0250] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0251] This specification can be described in the general context of computer-executable instructions that are executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. This specification can also be practiced in distributed computing environments, where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0252] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.

[0253] Finally, it should be noted that the port berthing probability prediction method and system disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention, and are only used to illustrate the technical solutions of the present invention, not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A port calling probability prediction method, characterized in that, The method comprises: obtaining ship data of a plurality of candidate ships; screening a target ship from the plurality of candidate ships according to the ship data, the target ship being docked at a target port; determining a ship type corresponding to the target ship according to the ship data corresponding to the target ship; predicting a docking berth and a waiting time of the target ship according to the ship data and the ship type corresponding to the target ship based on a port docking prediction model corresponding to the target port.

2. The port calling probability prediction method according to claim 1, characterized in that, The ship data comprises ship sensing data, ship communication data and ship AIS data; the ship sensing data comprises at least one of image data, temperature data, humidity data, sound data and light reflection data.

3. The port calling probability prediction method according to claim 1, characterized in that, The screening of the target ship from the plurality of candidate ships according to the ship data comprises: for each of the candidate ships, predicting a future ship route of the candidate ship according to the ship data of the candidate ship; determining at least one associated ship corresponding to the candidate ship from other candidate ships according to the future ship route of the candidate ship; calculating an average of minimum distances between the future ship route of the candidate ship and a port location of the target port of all the associated ships to obtain a docking parameter corresponding to the candidate ship; screening the candidate ship with the docking parameter less than a preset parameter threshold to obtain the target ship.

4. The port calling probability prediction method according to claim 3, characterized in that, The prediction of the future ship route of the candidate ship according to the ship data of the candidate ship comprises: inputting the ship data of the candidate ship into a trained route prediction model to obtain a predicted route corresponding to the candidate ship; the route prediction model is trained by a training data set comprising a plurality of training ship data and corresponding route labels; determining whether there is preset route plan information in the ship data of the candidate ship to obtain a first determination result; if the first determination result is no, determining a part of the route after a current position of the candidate ship in the predicted route as the future ship route of the candidate ship; if the first determination result is yes, determining a planned ship route of the candidate ship according to the route plan information; calculating a difference degree between the planned ship route and the predicted route and determining whether the difference degree is greater than a preset difference degree threshold to obtain a second determination result; if the second determination result is no, determining a part of the route after the current position of the candidate ship in the planned ship route as the future ship route of the candidate ship; if the second determination result is yes, calculating a weighted average result of the planned ship route and the predicted route to obtain a calculation result route, and determining a part of the route after the current position of the candidate ship in the calculation result route as the future ship route of the candidate ship.

5. The port calling probability prediction method according to claim 3, characterized in that, The determination of the at least one associated ship corresponding to the candidate ship from other candidate ships according to the future ship route of the candidate ship comprises: calculating a mathematical relationship model of distance between each route point on the future ship route of the candidate ship and the target port varying with route point time; determine a coefficient parameter in the mathematical relation model as a route coefficient parameter of the candidate ship; calculate a parameter similarity between the route coefficient parameter of each other candidate ship and the route coefficient parameter of the candidate ship; screen out a ship with a parameter similarity greater than a preset similarity threshold from all other candidate ships to obtain at least one associated ship corresponding to the candidate ship.

6. The port calling probability prediction method according to claim 1, characterized in that, The method comprises the following steps: inputting the ship data corresponding to the target ship into a trained ship type identification model to obtain the ship type corresponding to the target ship; the ship type is a conventional ship, a special ship, or a special pulp ship; and the ship type identification model is trained by a training data set comprising a plurality of training ship data and corresponding ship type labels.

7. The port calling probability prediction method according to claim 1, characterized in that, The method comprises the following steps: determining a predicted arrival time of the target ship according to the ship data of the target ship; inputting the ship data and the ship type corresponding to the target ship into the port berthing prediction model corresponding to the target port to obtain a time interval berthing prediction distribution corresponding to the target ship; the time interval berthing prediction distribution comprises predicted berthing positions and predicted berthing durations in a plurality of time intervals; determining the predicted berthing position and the predicted berthing duration of the target ship in a time interval in which the target ship is located at the predicted arrival time according to the time interval berthing prediction distribution to obtain the berthing position and the berthing duration of the target ship.

8. The port calling probability prediction method according to claim 1, characterized in that, The port berthing prediction model is trained by the following steps: obtaining historical ship arrival data of the target port from a port database; constructing a multi-dimensional data set according to the historical ship arrival data; dividing data in the multi-dimensional data set into a plurality of time interval data sets by adopting a time interval statistical method; performing berthing position statistics and berthing duration statistics on each time interval data set to obtain data labels corresponding to the multi-dimensional data set; inputting the multi-dimensional data set and the corresponding data labels into a preset prediction model for iterative training to obtain the port berthing prediction model.

9. A port calling probability prediction system characterized by, The system comprises: an acquisition module configured to acquire ship data of a plurality of candidate ships; a screening module configured to screen a target ship berthing at a target port from the plurality of candidate ships according to the ship data; a determination module configured to determine a ship type corresponding to the target ship according to ship data corresponding to the target ship; a prediction module configured to predict a berthing position and a berthing duration of the target ship according to ship data and a ship type corresponding to the target ship based on a port berthing prediction model corresponding to the target port.

10. A port calling probability prediction system characterized by, The system comprises: a memory storing executable program codes; a processor coupled to the memory; The processor invokes the executable program code stored in the memory to perform the port calling probability prediction method according to any one of claims 1-8.

Images