A port cluster channel guide and assist navigation decision method and system based on rule reasoning

Abstract

The application discloses a port group navigation channel guide and aid navigation decision method and system based on rule reasoning, and belongs to the technical field of intelligent port channel, which comprises the following steps: obtaining and standardizing three types of input data of environment, management and business, and ship state; calculating four types of risk states of wind, wave, tide and navigation density, and weighting and fusing into a comprehensive risk value; executing constraint judgment of tide level, visibility, navigation density, wind speed and the like, and triggering forced restriction when not satisfied; performing rule reasoning based on state, risk value and constraint set, generating candidate actions, and outputting optimal actions through utility function conflict resolution; mapping color state according to risk level to synchronously construct explainability output; and synchronously transmitting the optimal actions to a digital twin platform to realize dynamic linkage, and supporting artificial feedback writing back. The application realizes multi-source data fusion, continuous risk quantization, rule conflict resolution, decision explainability and twin closed-loop feedback, and improves the intelligent level of port group channel scheduling.

Description

Technical Field

[0001] This invention belongs to the field of intelligent port and waterway technology, and in particular relates to a method and system for navigation-aided decision-making in port clusters based on rule-based reasoning. Background Technology

[0002] In the context of large coastal port clusters and waterways, port dispatchers need to comprehensively assess multiple environmental factors, including wind speed, gusts, wave height, wave period, tide height, visibility, and traffic density, to determine navigational aid operations such as vessel entry, departure, waiting, speed-limited passage, or enhanced monitoring. Because port cluster waterways typically involve complex navigation conditions across multiple stations, sections, and vessel types, and because various environmental factors are interconnected and threshold conditions dynamically shift with seasonal and tidal changes, navigational aid decisions become a typical multi-factor comprehensive evaluation problem, demanding extremely high levels of professional experience and comprehensive judgment from on-duty personnel.

[0003] Currently, coastal ports have widely deployed marine meteorological monitoring platforms, port business management platforms, AIS vessel dynamic monitoring systems, and digital twin visualization scenarios. Relevant environmental data can be presented to on-duty personnel in the form of numerical values, curves, and cloud maps. However, the completeness of these systems in data collection, transmission, and display has not effectively translated into decision support capabilities, and the contradiction between abundant data and insufficient decision-making is becoming increasingly prominent. Specifically, existing systems have revealed the following systemic defects in practical applications.

[0004] First, there is a lack of automated bridge between the display of environmental information and decision-making. While existing systems can display real-time values ​​and historical trend charts of elements such as wind, waves, and tides, all navigation-related decisions, such as whether entry into port is permitted under current conditions or whether departure needs to be delayed, still rely on subjective interpretation by on-duty personnel based on their personal experience. The system itself lacks the ability to transform multi-source data into business recommendations. This "human-views-data-, human-makes-judgments" model creates a significant disconnect between the data display layer and the decision-making output layer. Navigation efficiency is limited by personnel skill and fluctuating conditions, failing to meet the rapid response requirements of high-frequency, high-concurrency vessel scheduling scenarios in port clusters.

[0005] Secondly, the lack of a standardized rule mapping mechanism makes it difficult to ensure consistency in decision-making. Even when faced with the same set of wind, wave, and tide parameters, different duty personnel may arrive at drastically different navigational aid judgments due to differences in experience, risk preferences, and cognitive habits, leading to frequent instances of "different judgments under the same conditions." This not only makes decision-making results vary from person to person, lacking repeatability and traceability, but also makes it difficult to standardize and reuse excellent duty experience, hindering the accumulation and inheritance of knowledge in port scheduling operations. At a deeper level, this decision-making model, which relies on individual experience rather than standardized rules, actually constitutes the core bottleneck for the intelligent upgrading of port cluster channel scheduling. It cannot support unified scheduling management across large-scale, multi-site operations, nor can it provide usable historical decision samples for subsequent data-driven optimization.

[0006] Furthermore, even though some ports have attempted to incorporate certain environmental thresholds (such as issuing warnings when wind speed exceeds a certain value) into their business systems, these implementations are mostly static alarm logic based on single factors and single thresholds, far from forming a complete rule-based reasoning capability. Specifically, this manifests in: a lack of an architecture that integrates multiple factors such as wind speed, wave height, tide level, navigation density, and vessel draft into a unified rule system for comprehensive reasoning; a lack of conflict resolution mechanisms when multiple rules are triggered simultaneously, meaning the system cannot provide the optimal coordinated suggestion when "high wind speed suggests delaying departure" and "high-priority vessel emergency entry request" coexist; and a lack of continuous mathematical modeling of environmental states and risk levels, causing existing threshold judgments to easily jump at boundaries (e.g., a judgment of safety at 9.9 m / s suddenly jumps to danger at 10.1 m / s), which does not conform to the gradual physical characteristics of actual navigation safety. These deficiencies make existing technology unsuitable for the complex and ever-changing actual business scenarios of port clusters and waterways with multiple coupled rules.

[0007] Building upon the above, there is also a dual dilemma regarding decision interpretability and human-machine collaboration. On the one hand, even if a few systems can provide some advice or warnings, their internal judgment logic is usually invisible to on-duty personnel. It's impossible to clearly explain which environmental factors, reasoning paths, and constraints led to a particular advice. This "black box" tendency reduces on-duty personnel's trust in system advice and increases the difficulty of business verification and manual intervention. In practice, this often results in system advice being shelved or used only as a reference rather than a true basis for decision-making. On the other hand, with the widespread application of digital twin technology in port management, new integration requirements have been placed on navigation aid decision-making systems: they not only need to provide decision results but also integrate real-time input parameters, trend curves, detailed data previews, and decision suggestions into the same interactive interface, and be able to link and display with the 3D waterway scene and provide closed-loop feedback. Existing systems still have significant shortcomings in interface integration and the depth of twin linkage, and lack the ability to automatically write back the results of manual intervention to the rule base and form a continuous optimization loop.

[0008] In summary, existing technologies suffer from at least the following core deficiencies: a lack of joint reasoning capabilities based on multi-source environmental factors; a lack of a continuous mathematical modeling mechanism for environmental states and risk levels; a lack of multi-rule conflict resolution mechanisms; a lack of interpretable output of decision results; and a lack of closed-loop linkage between human intervention and digital twins. These deficiencies collectively point to a more fundamental problem: existing systems remain at the "data display" level and have not yet established a complete intelligent link from "data perception" to "decision cognition" and then to "action feedback." Therefore, there is an urgent need for an intelligent decision-making method and system for port cluster navigation aids that can achieve multi-source data fusion, dynamic environmental state determination, comprehensive risk quantification calculation, rule-based reasoning and conflict resolution, interpretable output, and digital twin linkage. Summary of the Invention

[0009] The purpose of this invention is to provide a port cluster navigation aid decision-making method and system based on rule reasoning, in order to solve the technical problems in the prior art, such as the difficulty in joint reasoning of multi-source environmental factors, the lack of continuous risk modeling and conflict resolution mechanisms, the lack of transparency in the decision-making process and the lack of digital twin closed-loop feedback, and to realize a complete intelligent link from data perception to decision cognition and then to action feedback.

[0010] To achieve the above objectives, this invention provides a port cluster channel navigation aid decision-making method based on rule-based reasoning, comprising the following steps: Step S1: Obtain the input data required for navigation aid decision-making in the port cluster. The input data includes a set of environmental variables. Set of management and business variables and the set of ship state variables And perform standardization processing on each environmental variable; Step S2: Calculate the state determination result set S based on the input data. The state determination results include the risk status. Wave risk status Tide level risk status and traffic density status ; Step S3: Calculate the comprehensive risk value R based on the state determination result set S and the input data; Step S4: Perform constraint condition determination based on the input data and establish a constraint condition set. When any constraint condition is not met, a mandatory restriction rule is triggered, where, Indicates the first One constraint condition; Step S5: Based on the set of state determination results Comprehensive risk value and constraint set Perform rule-based reasoning to generate a set of candidate navigation actions. When multiple rules are triggered simultaneously, conflict resolution is performed, and the optimal navigation aid action is output. ; Step S6: Map the risk level based on the comprehensive risk value R. And based on the optimal navigation aid actions And the risk level L output navigation aid color status; Step S7: Optimal navigation aid actions Synchronize with the digital twin platform to build a dynamic linkage mechanism, and perform manual intervention and result writing based on the manual feedback function.

[0011] Preferably, in step S1: The set of environmental variables includes wind speed ,gust , wave height Wave period , tide level ,wind direction Flow direction air pressure and temperature , represented as ; The set of management and operational variables includes waterway grades. Real-time air traffic density Time period type and vessel traffic control levels , represented as ; The set of ship state variables includes the ship's draft. Ship length Tug configuration level , represented as ; Standardize each environmental variable: ; in, Represents the original environment variables. Represents standardized environmental variables. This represents the minimum value. This represents the maximum value.

[0012] Preferably, in step S2, the state determination function is defined as follows: ,in Represents the j-th type of state. This represents the corresponding state mapping function, specifically: Wind risk status: ; Wave risk status: ; Tide level risk status: ; Traffic density status: ; in, These represent the weighting coefficients for wind risk status, wave risk status, tide level risk status, and navigation density status, respectively. These represent the threshold values ​​for wind risk status, wave risk status, and navigation density status, respectively. Indicates the minimum safe tide height. The Sigmoid function is defined as follows: .

[0013] Preferably, in step S3, the formula for calculating the comprehensive risk value R is: ; in, These represent the risk weight coefficients for wind risk status, wave risk status, tide risk status, navigation density status, ship draft, ship length, and tugboat configuration level, respectively.

[0014] Preferably, in step S4, the constraints include: Tide level constraints ; Visibility constraints ; Traffic density constraints ; Wind constraint ; in, Indicates the minimum safe visibility. Indicates the maximum permissible air traffic density. Indicates the maximum permissible wind speed. This indicates real-time visibility.

[0015] Preferably, the optimal navigation aid action is output in step S5. The specific process is as follows: S51. Construct a rule-based reasoning model. ,in, Represents a rule-based reasoning function; S52, Construct a set of candidate navigation actions These include delayed departures, delayed arrivals, speed limits, enhanced monitoring, cautious navigation, and normal navigation. S53. Define the action utility function as follows: ,in, This represents the utility mapping function for the i-th action; S54. When multiple rules are triggered simultaneously, conflict resolution is performed, and the optimal navigation aid action is output. : .

[0016] Preferably, in step S6, the output process of the navigation aid color status is as follows: S61. Calculate the risk level based on the comprehensive risk value: ; Where L represents the risk level, including: High risk: ; Medium risk: ; Low risk: ; S62. Establish a color mapping function: ; Among them, Color represents the navigation aid color status; the color correspondences are as follows: red corresponds to delayed departure, orange corresponds to enhanced monitoring, gray corresponds to cautious navigation, green corresponds to recommended entry, blue corresponds to speed limit passage, and black corresponds to normal navigation. S63. Construct an interpretable output model : ; in, For the first The contribution of each input variable to the risk outcome. The set of rules that are triggered. For the first Class state determination result, The state is where the constraint conditions are satisfied; where the first... The contribution of each input variable to the risk outcome is defined as follows: , Let i be the risk weight corresponding to the i-th variable. Let be the value of the i-th input variable.

[0017] Preferably, the feature is that: in step S7, the artificial feedback function is: ; in, This indicates a manual correction action. Indicates the feedback fusion function; The dynamic linkage mechanism is as follows: ; in, The digital twin channel state at time t, The updated digital twin state for the next moment. For the digital twin state update function, Input data for the real-time environment at time t. Update the step size for the digital twin.

[0018] This invention also provides a port cluster navigation aid decision-making system based on rule-based reasoning, comprising: The data acquisition module is used to collect meteorological data, wave data, tidal data, AIS data, and port business data, and uniformly map them into structured data of "time-site-environmental variables". The state calculation module is used to calculate the set of state determination results based on the input data. ; The comprehensive risk calculation module is used to calculate the risk based on the set of status determination results. Calculate the overall risk value using the input data. ; The constraint determination module is used to perform constraint determination based on the input data. The rule-based reasoning module is used to determine the result set based on the state. Comprehensive risk value Perform rule-based reasoning with the set of constraints C to generate the optimal navigation aid actions. ; The color mapping module is used to determine the overall risk value. Mapping risk level And based on the optimal navigation aid actions And the risk level L output navigation aid color status; The digital twin linkage module is used to realize decision display, 3D scene linkage, and human feedback writing.

[0019] Therefore, the present invention employs the above-mentioned rule-based reasoning-based port cluster navigation aid decision-making method and system, which has the following beneficial effects: (1) More than ten types of variables, such as wind speed, wave height, tide level, navigation density, and ship draft, are uniformly incorporated into the rule system. Through state judgment and comprehensive risk calculation, the joint quantitative assessment of multi-source environmental elements is realized, overcoming the limitations of single-element static alarm and adapting to the complex navigation conditions of port clusters.

[0020] (2) The Sigmoid function is used to continuously map the wind, wave, tide and density states. The comprehensive risk value is the weighted sum of each state, avoiding the jump problem at the threshold boundary, so that the risk assessment conforms to the gradual physical characteristics of navigation safety and improves the rationality of decision-making under boundary conditions.

[0021] (3) Define the action utility function. When multiple rules are triggered at the same time, the conflict is resolved by maximizing the utility function and the unique optimal action is output, which effectively solves the coordination decision problem in contradictory scenarios such as "high wind speed suggests delay" and "urgent port entry demand".

[0022] (4) Construct an interpretable model that includes the contribution of each variable, the set of triggering rules, the state judgment results and the constraint satisfaction state, so that the duty personnel can clearly understand the basis of the recommendations, enhance the trust in the system, and facilitate business verification and experience verification.

[0023] (5) The decision results are synchronized to the three-dimensional waterway twin scene in real time to realize color identification and dynamic updates; a manual feedback function is introduced to support the automatic writing back of manual correction actions to the rule base, forming a continuous optimization closed loop of "perception-decision-execution-feedback".

[0024] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0025] Figure 1 This is a block diagram of the overall architecture of a port cluster navigation aid decision-making method based on rule reasoning, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of the navigation-aid decision system module structure according to an embodiment of the present invention; Figure 3 This is a flowchart illustrating the rule-based reasoning process in an embodiment of the present invention. Figure 4 This is a schematic diagram of navigation aid action suggestions and risk color mapping according to an embodiment of the present invention; Figure 5 This is a schematic diagram of an interactive navigation aid interface according to an embodiment of the present invention; Figure 6 This is a flowchart illustrating the manual intervention and result write-back process in an embodiment of the present invention. Figure 7 This is a schematic diagram of a digital twin linkage application according to an embodiment of the present invention; Figure 8 This is a schematic diagram of a navigation aid decision-making system based on the Tianjin Port Channel Group - Xingang Station, according to an embodiment of the present invention. Detailed Implementation

[0026] The following detailed description of embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0027] like Figure 1This invention provides a port cluster navigation aid decision-making method based on rule-based reasoning, comprising the following steps: Step S1: Obtain the input data required for navigation aid decision-making in the port cluster. The input data includes a set of environmental variables. Set of management and business variables and the set of ship state variables And perform standardization on each environmental variable; where: The set of environmental variables includes wind speed ,gust , wave height Wave period , tide level ,wind direction Flow direction air pressure Temperature and visibility , represented as ; The set of management and operational variables includes waterway grades. Real-time air traffic density Time period type and vessel traffic control levels , represented as ; The set of ship state variables includes the ship's draft. Ship length Tug configuration level , represented as ; Standardize each environmental variable: ; in, Represents the original environment variables. Represents standardized environmental variables. This represents the minimum value. This represents the maximum value.

[0028] Step S2: Calculate the state determination result set S based on the input data. The state determination results include the risk status. Wave risk status Tide level risk status and traffic density status ; The state determination function is defined as follows: ,in Represents the j-th type of state. This represents the corresponding state mapping function, specifically: Wind risk status: ; Wave risk status: ; Tide level risk status: ; Traffic density status: ; in, These represent the weighting coefficients for wind risk status, wave risk status, tide level risk status, and navigation density status, respectively. These represent the threshold values ​​for wind risk status, wave risk status, and navigation density status, respectively. Indicates the minimum safe tide height. The Sigmoid function is defined as follows: .

[0029] Step S3: Calculate the comprehensive risk value R based on the state determination result set S and the input data; wherein, the expression for calculating the comprehensive risk value R is: ; in, These represent the risk weight coefficients for wind risk status, wave risk status, tide risk status, navigation density status, ship draft, ship length, and tugboat configuration level, respectively.

[0030] Step S4: Perform constraint condition determination based on the input data and establish a constraint condition set. When any constraint condition is not met, a mandatory restriction rule is triggered, where, Indicates the first There are several constraints; among which, the constraints include: Tide level constraints ; Visibility constraints ; Traffic density constraints ; Wind constraint ; in, Indicates the minimum safe visibility. Indicates the maximum permissible air traffic density. Indicates the maximum permissible wind speed. This indicates real-time visibility.

[0031] Step S5: Based on the set of state determination results Comprehensive risk value and constraint set Perform rule-based reasoning to generate a set of candidate navigation actions. When multiple rules are triggered simultaneously, conflict resolution is performed, and the optimal navigation aid action is output. Among them, the optimal navigation aid maneuver is output. The specific process is as follows: S51. Construct a rule-based reasoning model. ,in, Represents a rule-based reasoning function; S52, Construct a set of candidate navigation actions , Indicates the first One candidate navigation aid action; the set of navigation aid candidate actions in this embodiment At least including: in: Indicates a delay in departure; This indicates a delay in arrival at the port; It is suggested that the vessel enter port; This indicates a speed limit. This indicates that monitoring will be strengthened; Indicates a cautious approach to resuming navigation; This indicates that air traffic is operating normally.

[0032] Candidate navigation aid set It can be further expanded according to the needs of port cluster and waterway operations; the expansion action is recorded as follows: ; S53. Define the action utility function as follows: ,in, This represents the utility mapping function for the i-th action; S54. When multiple rules are triggered simultaneously, conflict resolution is performed, and the optimal navigation aid action is output. : .

[0033] Step S6: Map the risk level based on the comprehensive risk value R. And based on the optimal navigation aid actions The risk level L is used to output the navigation aid color status; the output process of the navigation aid color status is as follows: S61. Calculate the risk level based on the comprehensive risk value: ; Where L represents the risk level, including: High risk: ; Medium risk: ; Low risk: ; S62. Establish a color mapping function: ; Among them, Color represents the navigation aid color status; the color correspondence is as follows: red corresponds to delayed departure / delayed arrival, orange corresponds to enhanced monitoring, gray corresponds to cautious navigation, green corresponds to recommended arrival, blue corresponds to speed limit passage, and black corresponds to normal navigation. S63. Construct an interpretable output model : ; in, This represents the set of information explaining navigation aid decisions; Indicates the first Each risk contribution factor affects the overall risk value. The degree of contribution; Indicates the set of rules that were triggered; This represents the overall risk value; Indicates the first Class state determination result; This indicates the state where the constraint condition is satisfied.

[0034] To characterize the impact of each risk factor on the overall risk value The degree of contribution is used to define the risk contribution factor vector. : ; in, Represents the risk contribution factor vector. Indicates the first Each risk contribution factor This indicates the total number of risk contribution factors.

[0035] In this embodiment, the aforementioned risk contribution factor vector Elements and overall risk value The calculation items correspond one-to-one, specifically: ; ; ; ; ; ; ; Therefore, in this embodiment: ; The corresponding risk weight vector is: ; in, This represents the risk contribution factor weight vector. Indicates the first The weights of each risk contribution factor.

[0036] In this embodiment, the risk weight vector The specific elements are as follows: ; ; ; ; ; ; ; Therefore, in this embodiment: ; No. The contribution of each risk contribution factor is defined as follows: ; in, Indicates the first The risk contribution of each risk contributing factor Indicates the first The weights of each risk contribution factor Indicates the first The values ​​of each risk contribution factor.

[0037] Therefore, the overall risk value This can be expressed as the sum of the contributions of each risk: ; In this embodiment, due to ,therefore: ; Right now: ; Using the above method, the system can output the overall risk value based on the contribution of each risk factor. The degree of contribution is determined to form an interpretable output result.

[0038] For example: In this embodiment, each risk contribution degree and the overall risk value The calculation items correspond one-to-one, specifically: ; ; ; ; ; ; ; in, Indicates the contribution of wind risk status. Indicates the contribution of wave risk status. Indicates the contribution of tide level risk status. Indicates the contribution of air traffic density status. Indicates the contribution of the ship's draft. Indicates the contribution of ship length. This indicates the contribution level of the tugboat configuration.

[0039] For example, when wind speed When it increases, because: ; middle The increase in the number of items makes the risk status more stable. Increase, and thus: ; An increase indicates that the risk level affects the overall risk value. The contribution has increased.

[0040] when the wave is high or wave period When it increases, because: ; in, The increase in the term leads to a higher risk status for waves. Increase, and thus: ; An increase indicates that wave risk affects the overall risk value. The contribution has increased.

[0041] The constraint state is represented as: ; in, This indicates that all constraints are satisfied. This indicates that at least one constraint condition is not satisfied. Specifically: ; For example, when At that time, due to the tidal level constraint: ; at this time: ; This indicates that the tide level constraint is not met, and the system outputs... This indicates that a constraint has been violated, and the corresponding mandatory restriction rule has been triggered.

[0042] Step S7: Optimal navigation aid actions A dynamic linkage mechanism is established by synchronizing with the digital twin platform, enabling manual intervention and result feedback based on a human feedback function; the human feedback function is as follows: ; in, This indicates a manual correction action. Indicates the feedback fusion function; The dynamic linkage mechanism is as follows: ; in, The digital twin channel state at time t, The updated digital twin state for the next moment. For the digital twin state update function, Input data for the real-time environment at time t. The system updates the step size for the digital twin. It synchronously writes human feedback results back to the log database, rule database, and digital twin for subsequent rule optimization, shift review, and scenario linkage, achieving a closed-loop update for navigation-aided decision-making. This step enables the system to simultaneously possess automated decision-making assistance capabilities and human controllability.

[0043] like Figure 2 As shown, the present invention also provides a port cluster navigation aid decision-making system based on rule-based reasoning, comprising: The data acquisition module is used to collect meteorological data, wave data, tidal data, AIS data, and port business data, and uniformly map them into structured data of "time-site-environmental variables". The state calculation module is used to calculate the set of state determination results based on the input data. ; The comprehensive risk calculation module is used to calculate the risk based on the set of status determination results. Calculate the overall risk value using the input data. ; The constraint determination module is used to perform constraint determination based on the input data. The rule-based reasoning module is used to determine the result set based on the state. Comprehensive risk value Perform rule-based reasoning with the set of constraints C to generate the optimal navigation aid actions. ; The color mapping module is used to determine the overall risk value. Mapping risk level And based on the optimal navigation aid actions And the risk level L output navigation aid color status; The digital twin linkage module is used to realize decision display, 3D scene linkage, and human feedback writing.

[0044] Example This embodiment further illustrates the invention based on the Tianjin Port Channel Group—Xingang Station, and the specific implementation process includes: (1) Input data preparation Read a fused navigation aid environment data table (Tianjin Port Xingang Station), as shown in Table 1. In the current prototype, this data table is formed by reconstructing wide table environment data into a long table, performing numerical transformation, interpolating missing values, standardizing, and reconstructing from a perspective, and then saving it as a fused data file. The data includes at least the following elements: wind speed, wave height, wave period, tide level, and navigation density.

[0045] Table 1: Input Data of Tianjin Port Xingang Station on April 10

[0046] (2) Results of state calculation and risk calculation Based on the formulas in steps S2 and S3, each state value and the overall risk value are calculated sequentially. To illustrate the calculation process of the state values ​​and the overall risk value in this embodiment, the model parameters are set as shown in Table 2 below.

[0047] Table 2: Model Parameter Values

[0048] in, These are the ship's state parameters after normalization. In this embodiment, the ship's state risk item is: Taking the 06:00 time slot as an example, the following will be: ; Substitute into the state determination function; The risk status is: ; The wave risk status is: ; The tide level risk status is as follows: ; The air traffic density status is as follows: ; Therefore, the overall risk value under the 06:00 time slot is: ; Right now: ; because: ; Therefore, the 06:00 time slot corresponds to a high-risk level.

[0049] The risk calculation results for each time slot are shown in Table 3 below: Table 3: Risk Calculation Results

[0050] (3) Results of rule-based reasoning System basis Output navigation aid actions, rule reasoning process as follows Figure 4 As shown, the mapping relationship between action suggestions and risk colors is as follows: Figure 4 As shown in Table 4, the output results for this time slot are as follows.

[0051] Table 4: Action Output Results

[0052] (4) Digital twin linkage The system synchronizes navigation aid actions to the digital twin platform, enabling digital twin-based collaborative applications such as... Figure 7 As shown. The system supports manual modification of navigation aid actions by on-duty personnel and automatically writes back to the rule base. The manual intervention and result write-back process is as follows: Figure 6 As shown.

[0053] (5) Interactive input and rule-based reasoning Set the environment input items in the navigation aid interface; the interactive navigation aid interface is as follows: Figure 5 As shown, this is used to receive environmental parameters at the current or simulated moment. In the current prototype, at least the following input items should be set: wind speed, wave height, tide height, and air temperature. These input items can be implemented through sliding controls, input boxes, API push, or system synchronous updates. The input parameters are then fed into a rule function for inference. In the current prototype, at least the following rules should be set: when the wind speed is greater than 15 or the wave height is greater than 5, the system outputs "Delay departure"; when the tide height is less than 1.5 and the wave height is less than 5, the system outputs "Recommend entry"; in other cases, the system outputs other actions according to the specific rules. Different action suggestions are mapped to different colors: red for delayed departure, green for recommended entry, and orange for enhanced monitoring.

[0054] (6) Interface display and application method In the current embodiment, the system outputs at least the current navigation aid suggestion, corresponding color indicators, tide height trend chart, and fused data preview on the same interface, as shown in the interface display method. Figure 5 As shown in the figure. The tide height trend chart is generated by reading the tide height field from the fused data file, and is used to help duty personnel observe environmental change trends. Duty personnel decide whether to proceed with port entry, departure, wait, or continue monitoring based on the navigation aid suggestions, color indicators, and trend chart information displayed on the interface. In this embodiment, the system first reads the environmental elements associated with the current time node from the data organization structure of the rule base, then calls the rule base to determine the action suggestion, and outputs the results on the interface. The overall application of the navigation aid decision system at the Tianjin Port New Port Station corresponding to this embodiment is as follows: Figure 8 As shown.

[0055] Therefore, this invention employs a rule-based reasoning-based decision-making method and system for port cluster navigation aids. Addressing the challenge of jointly applying multi-source heterogeneous data, the method first acquires three types of input data: environmental, management and operational, and vessel status. Environmental variables are standardized to lay the foundation for subsequent unified reasoning. To address the shortcomings of traditional threshold judgments, which are prone to abrupt changes at boundaries and do not conform to the gradual physical characteristics of safety, a sigmoid function is used to construct continuous mapping models for wind, wave, tide, and navigation density risks. A comprehensive risk value is obtained through weighted fusion, while simultaneously performing constraint judgments on tide level, visibility, density, and wind speed to ensure that the safety baseline is not breached. To address the lack of coordination mechanisms when multiple rules are triggered simultaneously, an action utility function is defined to resolve conflicts, outputting a unique optimal navigation aid action based on utility maximization. To address the issues of opaque decision-making processes and low trust among watch personnel, an interpretable output model is constructed, including variable contribution, trigger rule sets, state results, and constraint satisfaction, making the decision-making basis clear and traceable. Finally, the optimal actions are synchronized in real time to the digital twin platform to achieve 3D scene linkage and support manual intervention and correction. The correction results are automatically written back to the rule base, forming a continuous optimization closed loop from perception to feedback. This invention effectively improves the intelligence level, consistency, and human-machine collaboration efficiency of port cluster channel scheduling decisions.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A port cluster navigation aid decision-making method based on rule-based reasoning, characterized in that, Includes the following steps: Step S1: Obtain the input data required for navigation aid decision-making in the port cluster. The input data includes a set of environmental variables. Set of management and business variables and the set of ship state variables And perform standardization processing on each environmental variable; Step S2: Calculate the state determination result set S based on the input data. The state determination results include the risk status. Wave risk status Tide level risk status and traffic density status ; Step S3: Calculate the comprehensive risk value R based on the state determination result set S and the input data; Step S4: Perform constraint condition determination based on the input data and establish a constraint condition set. When any constraint condition is not met, a mandatory restriction rule is triggered, where, Indicates the first One constraint condition; Step S5: Based on the set of state determination results Comprehensive risk value and set of constraints Perform rule-based reasoning to generate a set of candidate navigation actions. When multiple rules are triggered simultaneously, conflict resolution is performed, and the optimal navigation aid action is output. ; Step S6: Map the risk level based on the comprehensive risk value R. And based on the optimal navigation aid actions And the risk level L output navigation aid color status; Step S7: Optimal navigation aid actions Synchronize with the digital twin platform to build a dynamic linkage mechanism, and perform manual intervention and result writing based on the manual feedback function.

2. The port cluster navigation aid decision-making method based on rule-based reasoning according to claim 1, characterized in that, In step S1: The set of environmental variables includes wind speed ,gust , wave height Wave period , tide level ,wind direction Flow direction air pressure and temperature and visibility , represented as ; The set of management and operational variables includes waterway grades. Real-time air traffic density Time period type and vessel traffic control levels , represented as ; The set of ship state variables includes the ship's draft. Ship length Tug configuration level , represented as ; Standardize each environmental variable: ； in, Represents the original environment variables. Represents standardized environmental variables. This represents the minimum value. This represents the maximum value.

3. The port cluster navigation aid decision-making method based on rule-based reasoning according to claim 2, characterized in that, In step S2, the state determination function is defined as follows: ,in Represents the j-th type of state. This represents the corresponding state mapping function, specifically: Wind risk status: ； Wave risk status: ； Tide level risk status: ； Traffic density status: ； in, These represent the weighting coefficients for wind risk status, wave risk status, tide level risk status, and navigation density status, respectively. These represent the threshold values ​​for wind risk status, wave risk status, and navigation density status, respectively. Indicates the minimum safe tide height. The Sigmoid function is defined as follows: 。 4. The port cluster navigation aid decision-making method based on rule-based reasoning according to claim 3, characterized in that, In step S3, the formula for calculating the comprehensive risk value R is: ； in, These represent the risk weight coefficients for wind risk status, wave risk status, tide risk status, navigation density status, ship draft, ship length, and tugboat configuration level, respectively.

5. A port cluster navigation aid decision-making method based on rule-based reasoning according to claim 4, characterized in that, In step S4, the constraints include: Tide level constraints ; Visibility constraints ; Traffic density constraints ; Wind constraint ; in, Indicates the minimum safe visibility. Indicates the maximum permissible air traffic density. Indicates the maximum permissible wind speed. This indicates real-time visibility.

6. The port cluster navigation aid decision-making method based on rule-based reasoning according to claim 5, characterized in that, In step S5, the optimal navigation aid action is output. The specific process is as follows: S51. Construct a rule-based reasoning model. ,in, Represents a rule-based reasoning function; S52, Construct a set of candidate navigation actions , Indicates the first One candidate navigation aid maneuver; S53. Define the action utility function as follows: ,in, This represents the utility mapping function for the i-th action; S54. When multiple rules are triggered simultaneously, perform conflict resolution and output the optimal navigation aid action. : 。 7. A port cluster navigation aid decision-making method based on rule-based reasoning according to claim 6, characterized in that, In step S6, the output process of the navigation aid color status is as follows: S61. Calculate the risk level based on the comprehensive risk value: ； Where L represents the risk level, including: High risk: ; Medium risk: ; Low risk: ; S62. Establish a color mapping function: ； Among them, Color represents the navigation aid color status; the color correspondence is as follows: red corresponds to delayed departure / delayed arrival, orange corresponds to enhanced monitoring, gray corresponds to cautious navigation, green corresponds to recommended arrival, blue corresponds to speed limit passage, and black corresponds to normal navigation. S63. Construct an interpretable output model: ； in, This represents the set of information explaining navigation aid decisions; Indicates the first Each risk contribution factor affects the overall risk value. The degree of contribution; Indicates the set of rules that were triggered; This represents the overall risk value; Indicates the first Class state determination result; This represents the state where the constraint condition is satisfied; where, the first... Each risk contribution factor affects the overall risk value. The contribution is defined as , Indicates the first The weights of each risk contribution factor Indicates the first The values ​​of each risk contribution factor.

8. The port cluster navigation aid decision-making method based on rule reasoning according to claim 7, characterized in that: In step S7, the manual feedback function is: ； in, This indicates a manual correction action. Indicates the feedback fusion function; The dynamic linkage mechanism is as follows: ； in, The digital twin channel state at time t, The updated digital twin state for the next moment. For digital twin state update functions, Input data for the real-time environment at time t. Update the step size for the digital twin.

9. A port cluster channel navigation aid decision system based on rule-based reasoning, applied to the port cluster channel navigation aid decision method based on rule-based reasoning as described in any one of claims 1-8, characterized in that, include: The data acquisition module is used to collect meteorological data, wave data, tidal data, AIS data, and port business data, and uniformly map them into structured data of "time-site-environmental variables"; The state calculation module is used to calculate the set of state determination results based on the input data. ; The comprehensive risk calculation module is used to calculate the risk based on the set of status determination results. Calculate the overall risk value using the input data. ; The constraint determination module is used to perform constraint determination based on the input data. The rule-based reasoning module is used to determine the result set based on the state. Comprehensive risk value Perform rule-based reasoning with the set of constraints C to generate the optimal navigation aid actions. ; The color mapping module is used to determine the overall risk value. Mapping risk level And based on the optimal navigation aid actions And the risk level L output navigation aid color status; The digital twin linkage module is used to realize decision display, 3D scene linkage, and human feedback writing.

Images