A ship-shore linkage type ballast water pathogen instant treatment control method

Abstract

The application discloses a kind of ship-shore linkage type ballast water pathogen real-time processing control method, belong to ship ballast water treatment technical field, it includes obtaining basic configuration information and real-time multi-source sensing data, constructs ship end dynamic digital mirror image;Based on the mirror image state data of ship end dynamic digital mirror image and the ship navigation information obtained, generate collaborative strategy vector;Collaborative strategy vector is injected into ship end dynamic digital mirror image, in combination with real-time multi-source sensing data, drive ship end dynamic digital mirror image to carry out analog deduction, generate ship-shore collaborative mode equipment control instruction set;When the communication state of communication link is changed to abnormal, analog deduction is generated to generate autonomous evolution mode equipment control instruction set;Two kinds of mode equipment control instruction set are used for controlling ship end ballast water treatment equipment under the condition that communication is normal and abnormal respectively.The application adopts ship-shore collaboration and digital mirror image technology, in combination with analog deduction, realizes the decision control of ballast water pathogen treatment.

Description

Technical Field

[0001] This invention relates to the field of ship ballast water treatment technology, and in particular to a control method for the immediate treatment of pathogens in ballast water in a ship-shore integrated manner. Background Technology

[0002] Ballast water is a crucial tool for ships to regulate their draft, attitude, and stability. However, the various aquatic organisms and pathogens carried in ballast water can be transported to new waters during ocean voyages, potentially posing a serious threat to global marine ecosystems, fisheries resources, and human health. To control this risk, the International Maritime Organization has established a strict ballast water management convention, requiring ships to be equipped with ballast water treatment systems and to effectively treat ballast water before discharge to kill pathogens and ensure compliance with discharge standards.

[0003] Among related technologies, Chinese invention patent with announcement number CN120353124B discloses a collaborative control method for filtration and ultraviolet sterilization in a purely physical ballast water management system. The method includes: based on real-time monitoring of water flow characteristics, filter status, water quality parameters and ultraviolet sterilization effect, intelligent feedback regulation, machine learning optimization, adaptive flow rate adjustment and dynamic energy consumption optimization are used to achieve efficient operation of the ballast water treatment system.

[0004] However, the aforementioned existing technical solutions suffer from the following technical shortcomings: Lack of ship-shore collaborative decision-making capabilities: Existing technologies may rely solely on local data and computing power of the shipboard system for decision-making, failing to leverage the more powerful data processing centers and richer global historical data and experience at the shore level. This results in insufficient global optimization and foresight of the control strategy. Weak communication interruption response capabilities: In scenarios with unstable communication, such as ocean voyages, existing systems may experience processing interruptions or performance degradation due to the inability to receive external commands, lacking the resilience to maintain efficient and compliant operation autonomously during communication anomalies. Dynamic deviation between model and physical entity: The control model of existing technologies may lack a closed-loop mechanism for self-calibration based on actual processing effects, making it difficult to cope with dynamic changes such as equipment aging, pipeline scaling, or sudden changes in water quality, leading to a decrease in control accuracy over time. Inability to continuously evolve system knowledge: Valuable operating condition and optimization data accumulated on the shipboard during operation may not be effectively recycled and used to optimize the decision-making model of the entire system, resulting in a lack of the system's ability to learn and evolve from practice. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a control method for real-time treatment of ballast water pathogens through ship-shore collaboration. By employing ship-shore coordination and digital mirroring technology, combined with simulation and deduction, decision control for ballast water pathogen treatment is achieved.

[0006] The above objectives can be achieved through the following approach:

[0007] A control method for real-time treatment of pathogens in ballast water via ship-shore linkage includes: acquiring basic configuration information and real-time multi-source sensor data to construct a ship-side dynamic digital mirror; generating a cooperative strategy vector based on the mirror state data of the ship-side dynamic digital mirror and the acquired ship navigation information; injecting the cooperative strategy vector into the ship-side dynamic digital mirror, and combining it with real-time multi-source sensor data to drive the ship-side dynamic digital mirror to perform simulation and generate a ship-shore cooperative mode equipment control instruction set; when the communication link status changes to abnormal, performing simulation and generating an autonomous evolution mode equipment control instruction set; the two mode equipment control instruction sets are used to control the ship-side ballast water treatment equipment under normal and abnormal communication conditions, respectively.

[0008] Optionally, the construction and maintenance of the ship's dynamic digital mirror representing the real-time state includes: establishing a topology model of equipment and pipelines based on the basic configuration information; receiving real-time multi-source sensor data, including water quality parameters, equipment operating parameters, and ship attitude parameters; inputting the real-time multi-source sensor data into the topology model of equipment and pipelines for calculation, and outputting a current system state vector composed of multiple state dimension indicators in real time; and constructing and updating the ship's dynamic digital mirror using the current system state vector based on a preset update frequency to ensure synchronization.

[0009] Optionally, generating a collaborative strategy vector for the ship-end ballast water treatment process includes: extracting key operating condition features from the mirrored state data and associating them with the navigation area and planned route in the ship's navigation information to form a real-time operating condition context at the ship end; retrieving and generating a set of historical similar operating condition cases based on the real-time operating condition context at the ship end; inputting the real-time operating condition context at the ship end and the set of historical similar operating condition cases into the shore-based strategy generation model; and the shore-based strategy generation model outputting a collaborative strategy vector that includes the priority of the processing target, the boundary of the control parameter adjustment, the coupling relationship between parameters, and the principle of anomaly response.

[0010] Optionally, the step of driving the ship-side dynamic digital mirror to perform simulation and generate a ship-shore cooperative mode equipment control instruction set includes: parsing the cooperative strategy vector into an objective function and constraints for the ship-side dynamic digital mirror to perform simulation and optimization; using the latest acquired real-time multi-source sensor data as the boundary input conditions of the ship-side dynamic digital mirror; performing multiple rounds of forward simulation calculations on the ship-side dynamic digital mirror within the framework of the objective function and constraints, evaluating and generating a virtual control parameter combination; and converting the virtual control parameter combination into a ship-shore cooperative mode equipment control instruction set.

[0011] Optionally, the method further includes: acquiring effect verification sensing data; feeding back the effect verification sensing data to the ship-side dynamic digital mirror; the ship-side dynamic digital mirror comparing the preset expected processing effect with the actual processing effect obtained based on the effect verification sensing data to generate effect deviation data; based on the effect deviation data, the ship-side dynamic digital mirror adaptively adjusts the local parameters of its internal inference algorithm and re-performs simulation to generate a corrected ship-shore collaborative mode equipment control instruction set.

[0012] Optionally, monitoring the communication status of the communication link includes: periodically acquiring the signal strength, data transmission delay, and packet loss rate parameters of the communication link; performing a weighted evaluation on the signal strength, data transmission delay, and packet loss rate parameters to calculate a communication quality score; and comparing the communication quality score with preset normal thresholds for determining normal status and abnormal thresholds for determining abnormal status to determine the communication status.

[0013] Optionally, controlling the ship-side dynamic digital mirror to switch to autonomous evolution mode includes: continuously calculating and recording the communication quality score; when the communication quality score is lower than the normal threshold but higher than the abnormal threshold within a monitoring period, determining the communication state as a warning state and sending a warning notification to both the ship and shore ends; when the communication quality score is lower than the abnormal threshold in any monitoring period, immediately determining the communication state as an abnormal state and triggering a switching command; based on the switching command, locking the currently running cooperative strategy vector, using it as the fixed constraint in the autonomous evolution mode, and starting an autonomous control loop centered on local simulation.

[0014] Optionally, controlling the ship-side ballast water treatment equipment to perform operations based on the ship-shore collaborative mode equipment control instruction set includes: parsing the ship-shore collaborative mode equipment control instruction set and converting it into control signals for the corresponding specific execution equipment; sending the control signals to the corresponding ballast water treatment equipment; monitoring the execution feedback signals of the ballast water treatment equipment in real time to confirm whether the equipment operates correctly according to the instructions; and feeding the execution feedback signals as part of the real-time multi-source sensor data back to the ship-side dynamic digital mirror.

[0015] Optionally, the method further includes: when the communication status of the communication link is detected to have recovered from abnormal to normal, extracting the autonomous evolution log recorded in the autonomous evolution mode, wherein the autonomous evolution log includes an environmental data sequence, a locally generated autonomous evolution mode device control instruction set sequence, an effect verification sensor data sequence, and the local optimization strategy variant information; uploading the autonomous evolution log to the shore-based data center; the shore-based data center using the autonomous evolution log to update a preset regional historical database, and retraining and optimizing the shore-based strategy generation model based on the updated regional historical database.

[0016] Based on the same inventive concept, this invention also provides a control system for real-time treatment of ballast water pathogens in a ship-shore integrated manner. The system includes: a data acquisition and digital mirror construction module, used to acquire basic configuration information and real-time multi-source sensor data, and construct and maintain a ship-side dynamic digital mirror representing the real-time state; a shore-based strategy generation and communication module, used to calculate, based on the mirror state data of the ship-side dynamic digital mirror and acquired ship navigation information, using a preset shore-based strategy generation model to generate a collaborative strategy vector for the ship-side ballast water treatment process; and a ship-side strategy deduction and control command generation module, used to inject the collaborative strategy vector into the ship-side dynamic digital mirror, and, combined with the latest acquired real-time multi-source sensor data, drive the ship-side dynamic digital mirror to perform simulation deduction to generate... The system includes: a ship-shore collaborative mode equipment control instruction set; a communication status monitoring and autonomous evolution control module for monitoring the communication status of the communication link; when the communication status changes from normal to abnormal, controlling the ship-side dynamic digital mirror to switch to autonomous evolution mode; in the autonomous evolution mode, the ship-side dynamic digital mirror continuously performs simulation and deduction based on the latest acquired real-time multi-source sensor data, using the most recently received collaborative strategy vector as a fixed constraint, and generates an autonomous evolution mode equipment control instruction set; and a ballast water treatment equipment execution control module for controlling the ship-side ballast water treatment equipment to perform operations based on the ship-shore collaborative mode equipment control instruction set under normal communication conditions, and under abnormal communication conditions.

[0017] Compared with the prior art, the present invention has the following advantages:

[0018] This invention constructs a dynamic digital mirror of the ship's end and links it with a shore-based strategy generation model. This allows for the integration of the ship's real-time operating conditions and navigation information with the powerful data processing capabilities and historical experience of shore-based systems, enabling precise and predictive control of the ballast water treatment process. This ensures that the control strategy is not only based on the current state but also considers future navigation plans and regional environmental characteristics, thereby improving the reliability of the treatment effect and the economy of energy consumption.

[0019] The communication interruption autonomous evolution mode designed in this invention enhances the resilience and all-weather operation capability of the ballast water treatment system. In common scenarios such as ocean voyages where communication is unstable, the system can seamlessly switch to autonomous control, continuously optimizing locally based on the most recent shore-based strategy. This avoids processing interruptions or degradation caused by communication interruptions, ensuring that the vessel can continuously meet environmental regulations under any communication conditions.

[0020] This invention introduces an adaptive adjustment closed loop based on effect feedback at the ship's end, enabling the ship's dynamic digital mirror model to adjust the parameters of its internal inference algorithm in real time according to the deviation between the actual and expected processing results. This self-calibration mechanism allows the digital model to continuously closely reflect the dynamic changes of the physical entity, such as equipment aging and pipeline scaling, thereby ensuring the continuous effectiveness of control commands and improving the long-term accuracy and robustness of the control system.

[0021] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a schematic flowchart of a control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system, according to an embodiment of the present invention.

[0024] Figure 2 This is a heatmap showing the correlation between operating condition characteristics and strategy parameters in an embodiment of the present invention.

[0025] Figure 3 This is the collaborative strategy vector optimization derivation space diagram of an embodiment of the present invention.

[0026] Figure 4 This is a schematic diagram of the structure of a ship-shore linked ballast water pathogen real-time treatment control system according to an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0028] Reference Figure 1 One embodiment of the present invention proposes a control method for real-time treatment of ballast water pathogens in a ship-shore coordinated manner. By employing ship-shore collaboration and digital mirroring technology, combined with simulation and deduction, decision control for ballast water pathogen treatment is realized.

[0029] The method described in this embodiment specifically includes:

[0030] S1. Obtain basic configuration information and real-time multi-source sensor data, and construct and maintain a dynamic digital image of the ship's end representing the real-time state.

[0031] Optionally, the construction and maintenance of the ship's end dynamic digital mirror representing the real-time state includes:

[0032] Based on the aforementioned basic configuration information, establish a topology model of the equipment and pipelines;

[0033] The system receives real-time multi-source sensor data, which includes water quality parameters, equipment operating parameters, and ship attitude parameters.

[0034] The real-time multi-source sensor data is input into the device and pipeline topology model for calculation, and the current system state vector composed of multiple state dimension indicators is output in real time.

[0035] Based on a preset update frequency to ensure synchronization, the ship's dynamic digital image is constructed and updated using the current system state vector.

[0036] Specifically, based on the acquired basic configuration information of the ship's ballast water system, a topology model of the equipment and piping is constructed. This basic configuration information includes static engineering data such as the rated power and performance curves of the ballast pumps, the filter specifications and backwash cycles of the filtration units, the lamp power and layout of the ultraviolet disinfection units, the diameter and length of each pipeline, and valve types. The equipment and piping topology model is a digital structure stored in the form of a graphical database, where equipment is the node and pipelines are the edge, accurately mapping the connection relationships and fluid paths of the physical system.

[0037] To drive the model, real-time multi-source sensor data is received via an onboard data bus at a frequency of 1-5 Hz. This real-time multi-source sensor data is categorized into three types in engineering: the first type reflects water quality parameters of the operating environment, such as suspended solids concentration collected by a turbidity sensor, salinity measured by a conductivity meter, and water temperature measured by a temperature probe; the second type represents equipment operating parameters characterizing system load and efficiency, such as pipeline flow rate feedback from a flow meter, pressure difference across the filter monitored by a differential pressure sensor, dose intensity measured by an ultraviolet irradiation probe, and motor current and voltage; the third type comprises ship attitude parameters affecting fluid stability, such as the ship's roll and pitch angles provided by the inertial measurement unit. These ship attitude parameters are primarily used to correct for the impact of local pressure fluctuations caused by hull rolling and uneven fluid distribution within the pipeline on the model calculations.

[0038] Real-time multi-source sensor data is dynamically input into the device and pipeline topology model for calculation, generating a current system state vector in real time. This current system state vector is a multi-dimensional array, with each dimension's indices calculated by the model rather than directly measured by sensors. For example, the model's fluid dynamics solver calculates the instantaneous pressure and flow rate at any critical node in the system based on pump operating parameters, valve openings, and ship attitude parameters, combined with pipeline topology and friction coefficients. The calculation process follows fluid network analysis methods, estimating pipe pressure drop using the following formula. :

[0039] ,

[0040] Here, The pipe section resistance coefficient is a constant that integrates pipe length, diameter, roughness, and fluid viscosity. It is pre-calculated and calibrated by the equipment and pipe topology model based on the basic configuration information. The real-time flow rate of this pipe segment is calculated from the flow meter's sensor data or a model. The complete pressure distribution can be obtained by solving the nodal pressure method over the entire pipeline network. The current system state vector not only includes these physical states but also comprehensive indicators such as system processing efficiency, energy consumption levels, and equipment health assessments, providing a comprehensive understanding of the system's state for higher-level decision-making.

[0041] To ensure close synchronization between the virtual and real worlds, a preset update frequency, typically set at 0.5-2 Hz, is used to refresh the ship's dynamic digital mirror using the latest generated system state vector. The ship's dynamic digital mirror is a comprehensive object encompassing the equipment and piping topology model and all its internal state variables. The update process is not a simple data overwrite; rather, it uses the current system state vector as input to drive the state evolution of each component within the mirror, thus completing a state mapping from the physical world to the digital world. This step ensures that the ship's dynamic digital mirror always reflects the instantaneous state of the ship's physical system with high fidelity, forming the foundation for the entire control method's simulation, deduction, and autonomous decision-making.

[0042] For example, the system first obtains the basic configuration information of the ship's ballast water system to establish a topology model of the equipment and pipelines. In this model, for a specific pipe section connecting the ballast pump and the filter unit, the system pre-calculates and calibrates its overall resistance coefficient. The value is 0.0005. Subsequently, the system receives multi-source sensor data in real time, including real-time flow rate monitored by the flow meter. The flow rate is 200 cubic meters per hour, along with the ship's roll angle fed back by the inertial measurement unit. These real-time data are input into the equipment and the pipeline topology model, and fluid network analysis methods are used to calculate the instantaneous pressure drop of this pipe section. According to the formula Perform calculations, that is The system calculates a real-time pressure drop of 20 kPa for this pipe section. This calculated pressure, along with other state-level indicators such as system processing efficiency and energy consumption calculated from the model, forms the current system state vector. Finally, based on a preset update frequency of 1 Hz, the system injects this current system state vector as input into the ship's dynamic digital mirror, driving the evolution of the internal component states within the mirror, thereby completing a high-fidelity state mapping and synchronous update from the physical world to the digital world.

[0043] S2. Based on the mirror state data of the ship's end dynamic digital mirror and the acquired ship navigation information, a collaborative strategy vector for the ship's end ballast water treatment process is generated by calculating through a preset shore-based strategy generation model.

[0044] Optionally, the generation of the cooperative strategy vector for the ship's end ballast water treatment process includes:

[0045] Key operating condition features are extracted from the mirror state data and associated with the navigation area and planned route in the ship navigation information to form a real-time operating condition context at the ship's end.

[0046] Based on the real-time operating condition context at the ship's end, a set of historical similar operating condition cases is retrieved and generated;

[0047] The shipboard real-time operating condition context and the set of historical similar operating condition cases are input into the shore-based strategy generation model.

[0048] The output of the shore-based strategy generation model includes a collaborative strategy vector that includes the priority of processing targets, the boundary of control parameter adjustment, the coupling relationship between parameters, and the principle of anomaly response.

[0049] Specifically, key operational features are extracted from the image status data uploaded by the ship's dynamic digital mirror. These features include at least 10 core operational indicators selected from the current system state vector, such as fluid turbidity, salinity, water temperature, total system flow rate, and filter differential pressure. Simultaneously, the system receives ship navigation information provided by the integrated ship navigation system, primarily referring to the ship's current geographical location, its specific navigation area (e.g., emission control zones or high-risk areas for biological invasion), and its planned route for the next 24 hours. By structurally fusing these internal state features with external environmental and task characteristics, the system forms a feature vector called the ship's real-time operational context, which comprehensively describes all the conditions faced by the current task.

[0050] Based on the ship's real-time operating condition context, a search is performed in the regional historical database stored in the shore-based data center to generate a set of historical similar operating condition cases. The core of the search is a similarity matching algorithm; the system uses the ship's real-time operating condition context as a query vector and compares it with tens of thousands of historical operating condition records in the database. Similarity scoring is then performed. It can be calculated using the following weighted distance function:

[0051] ,

[0052] in The calculation formula is:

[0053] ,

[0054] In this formula, This represents the weighted Euclidean distance between the current operating condition and a certain historical operating condition. The numbers represent different feature dimensions in the working condition vector, as shown in the following indexes. Indicates the current value, subscript Represents historical values. For example, It can be a geolocation code. It could be the turbidity of the water. It can be the planned navigation area type. These are the weighting coefficients for each feature dimension, with values ​​ranging from 0.1 to 1.0. These coefficients are dynamically adjusted by the shore-based expert system based on different navigation area regulations and biological risk levels. For example, in high-risk areas for biological invasion, the weights of geographical location and water parameters are increased. The system selects the top 5 to 20 historical records with the smallest D-values, along with the strategies, control parameters, and final treatment effect evaluation data used at that time, to form a set of historical similar case studies.

[0055] The shore-based strategy generation model processes the newly generated shipboard operational context and a set of historical similar operational case studies containing both successful and failed experiences as dual inputs. This shore-based strategy generation model is a hybrid intelligent system based on deep learning and rule-based reasoning. The historical case set provides an empirical foundation, avoiding strategy search from scratch, while the current operational context defines the specific problem that needs to be solved. By analyzing the differences between the current operational situation and historical cases, and combining this with a built-in International Maritime Organization (IMO) regulatory library and equipment performance degradation model, the model infers the optimal macro-strategy.

[0056] The shore-based strategy generation model outputs a structured collaborative strategy vector. This vector is not a direct equipment control command, but a high-level set of guidelines, explicitly containing four core components: first, the priority of processing objectives, such as defining the primary objective as ensuring a biological inactivation rate of 99.99%, followed by minimizing energy consumption; second, the adjustment boundaries of key control parameters, such as specifying that the adjustment range of UV lamp power is 70% to 110% of the rated power; third, the coupling relationships between parameters, given in the form of functions or lookup tables, such as lowering the filter backwash trigger differential pressure threshold by 5% for every 100 cubic meters per hour increase in flow rate; and fourth, the anomaly response principles, providing high-level contingency plans, such as prioritizing the activation of backup filter units when the instantaneous turbidity value of the water exceeds the preset upper limit, rather than simply increasing the treatment intensity of the main unit. This collaborative strategy vector is then sent to the ship as a top-level constraint for detailed simulation and optimization of the ship's dynamic digital mirror.

[0057] For example, a heatmap showing the correlation between operating condition characteristics and strategy parameters is shown below. Figure 2 As shown, the system first extracts key operating condition features, including fluid turbidity, salinity, water temperature, and total system flow, from the mirror state data of the ship's dynamic digital mirror. These features are then correlated with the planned route for the next 24 hours from the ship's navigation information to form a real-time operating condition context at the ship's end. Within this operating condition context, the current geolocation code is defined. The real-time water turbidity is 120. The value is 35, and the planned navigation area type is... The value is 1. To generate a set of historical similar working condition cases, the system retrieves a specific set of historical records from the regional historical database, whose feature value is a geographic location code. The turbidity of the water body is 115. 30, region type The value is 1. At this point, the shore-based expert system adjusts the weighting coefficient based on the biological risk level of the navigation area. Set to 0.5 Set to 0.4. Set to 0.1. Calculate according to the weighted distance function formula. Then, using the formula Calculate the similarity score, i.e. The system selects a set of historical similar case cases, including this highly similar case, and inputs them, along with the ship's real-time operational context, into the shore-based strategy generation model. Finally, after inference and computation, the model outputs a structured collaborative strategy vector that includes the priority of processing objectives, the boundaries of control parameter adjustment, the coupling relationships between parameters, and the principles for handling anomalies.

[0058] S3. Inject the collaborative strategy vector into the ship-side dynamic digital image, and combine it with the latest acquired real-time multi-source sensor data to drive the ship-side dynamic digital image to perform simulation and generate a set of ship-shore collaborative mode equipment control instructions.

[0059] Optionally, the step of driving the ship-side dynamic digital mirror to perform simulation and generate a set of ship-shore collaborative mode equipment control instructions includes:

[0060] The collaborative strategy vector is parsed into the objective function and constraints of the ship's dynamic digital image for deduction and optimization.

[0061] The newly acquired real-time multi-source sensor data is used as the boundary input condition for the ship's end dynamic digital image.

[0062] Within the framework of the objective function and constraints, multiple rounds of forward simulation calculations are performed on the ship's end dynamic digital image to evaluate and generate virtual control parameter combinations;

[0063] The virtual control parameter combination is converted into a set of control instructions for ship-shore cooperative mode equipment.

[0064] Specifically, the received cooperative strategy vector is parsed into an objective function and constraints for optimization based on the ship's dynamic digital image. This parsing process is a structured transformation; for example, the processing objective priority "biological inactivation rate first, with energy consumption considered" in the cooperative strategy vector is transformed into a multi-objective optimization function. :

[0065] ,

[0066] in, This represents the biological inactivation rate predicted by the mirror model. This represents the predicted total system energy consumption. and These are weighting coefficients, and their values ​​are determined by priority, for example... It may be set to 0.8. The value is 0.2. The control parameter adjustment boundary in the collaborative strategy vector, such as "UV lamp power is between 70% and 110% of rated power", directly defines the search space of the optimization variables. The coupling relationship between parameters, such as the linkage rule between flow rate and filter pressure difference, is transformed into nonlinear constraints in the optimization process.

[0067] The latest acquired real-time multi-source sensor data is used as the boundary input condition for the ship's dynamic digital mirror. This means that before the simulation begins, the initial state of the mirror will be perfectly aligned with the physical entity. For example, if the current value measured by the turbidity sensor is 30 NTU, then the water quality model input for the virtual ballast water in the mirror will be 30 NTU; if the ship's roll angle fed back by the inertial measurement unit is 3 degrees, then the hydrodynamic solver in the mirror will load the additional gravity component generated by this roll angle, thus making the starting point of the simulation infinitely close to the real state of the physical world.

[0068] After defining the objective function, constraints, and initial boundary conditions, multiple rounds of forward simulation calculations will be performed on the ship's end dynamic digital image within this framework to evaluate and generate virtual control parameter combinations. A heuristic optimization algorithm, such as particle swarm optimization or a genetic algorithm, is typically used and executed on the ship's edge computing unit. The algorithm generates a population containing multiple virtual control parameter combinations; for example, one combination might be {ballast pump speed 85%, UV lamp power 98%, filter backwash interval 20 minutes}. For each combination, the ship's end dynamic digital image will perform a short-term predictive simulation, such as a 5-minute forecast, outputting the expected bio-inactivation rate and system energy consumption under that parameter combination. The algorithm is based on the objective function... The results of each combination are evaluated and scored. Through iteration, such as 100 to 500 iterations, inferior combinations are continuously eliminated, and new, potentially better combinations are generated based on the superior combinations, until a combination that satisfies the objective function is found. The optimal combination of virtual control parameters is achieved.

[0069] The optimized virtual control parameter combination is converted into a ship-shore cooperative mode equipment control instruction set. This is a formatted translation process that maps abstract percentages or time values ​​to specific engineering instructions that the underlying equipment can recognize. For example, "ballast pump speed is 85%" in the virtual control parameter combination will be converted into a message conforming to a specific inverter communication protocol, with the content "set output frequency to 51 Hz"; "UV lamp power is 98%" will be converted into "send an analog signal of 4.9 volts to the electronic ballast controller". These specific, time-sequential instructions together constitute the ship-shore cooperative mode equipment control instruction set, ready to be issued to the corresponding programmable logic controller (PLC) or equipment driver module to accurately execute the results of this optimization simulation. The cooperative strategy vector optimization simulation space is as follows: Figure 3 As shown.

[0070] For example, the system first parses the received cooperative strategy vector into an objective function and constraints for deduction and optimization based on the ship's dynamic digital image. During this process, the processing objective priority is transformed into a multi-objective optimization function. Set weight coefficients It is 0.8. The value is 0.2. Subsequently, the system injects the latest acquired real-time multi-source sensor data as boundary input conditions into the mirror image, for example, setting the current turbidity sensor measured value to be 30 NTU and the ship's roll angle to be 3 degrees. Within the framework of the objective function and constraints, the ship-side simulation algorithm generates a set of virtual control parameter combinations, including ballast pump speed, UV lamp power, and filter backwash interval. The mirror image performs short-term forward simulation calculations on this combination to predict the output biological inactivation rate. The system's total energy consumption is 99.99%. The value is 50 kWh. Substitute this value into the formula to calculate the target score. After multiple rounds of iterative evaluation and locking in, After determining the optimal parameter combination, the system converts it into a set of control instructions for ship-shore cooperative equipment. For example, 98% of the UV lamp power in the virtual parameters is converted into a 4.9-volt analog signal to be sent to the electronic ballast, and finally, the physical equipment is controlled to perform operations precisely through the industrial fieldbus.

[0071] S4. Monitor the communication status of the communication link. When the communication status changes from normal to abnormal, control the ship-end dynamic digital mirror to switch to autonomous evolution mode. In the autonomous evolution mode, the ship-end dynamic digital mirror uses the most recently received cooperative strategy vector as a fixed constraint, continuously performs simulation and deduction based on the latest acquired real-time multi-source sensor data, and generates an autonomous evolution mode device control instruction set.

[0072] Optionally, monitoring the communication status of the communication link includes:

[0073] The signal strength, data transmission delay, and packet loss rate parameters of the communication link are periodically acquired.

[0074] The signal strength, data transmission delay, and packet loss rate parameters are weighted and evaluated to calculate the communication quality score.

[0075] The communication quality score is compared with preset normal thresholds for determining normal states and abnormal thresholds for determining abnormal states to determine the communication status.

[0076] Specifically, key performance parameters of the communication link are periodically obtained from the shipboard communication management unit, with the monitoring period typically set between 5 and 10 seconds. The parameters obtained mainly include three core indicators: signal strength, usually expressed as Received Signal Strength Indicator (RSSI) or Reference Received Power (RSRP); data transmission delay, i.e., the round-trip time from sending a data packet from the ship to receiving a shore-based acknowledgment; and packet loss rate, i.e., the proportion of data packets that fail to be successfully transmitted within a specific time window out of the total number of data packets sent.

[0077] These raw performance parameters, with different dimensions, are normalized and weighted for evaluation to calculate a single, comprehensive communication quality score. They need to be converted into dimensionless scores between 0 and 1. For example, data transmission latency... normalized score It can be calculated using the following formula:

[0078] ,

[0079] In this formula, It is the delay value measured in real time. and It is the preset best and worst acceptable latency for this type of communication technology, for example, for nearshore 5G communication. It can be set to 20 milliseconds. This can be set to 200 milliseconds. This formula ensures that the lower the latency, the higher the score. The closer to 1, the better. Similarly, signal strength and packet loss rate will also undergo similar normalization. The final communication quality score is calculated using the following weighted summation formula. :

[0080] ,

[0081] In this formula, The final communication quality score is also between 0 and 1, or multiplied by 100 to convert it to a percentage. These are the normalized scores for signal strength, data transmission delay, and packet loss rate. These are the weighting coefficients of the three factors, which sum to 1, and are pre-configured by the system based on the degree of communication dependence of the ballast water treatment task. For example, in stages where frequent distribution of collaborative strategy vectors is required, the weights of latency and packet loss rate are... and It may be increased, for example, set to 0.4 respectively.

[0082] The communication quality score calculated in real time The current communication state is determined by comparing the calculated score with preset thresholds for normal and abnormal states. These two thresholds are key parameters for system calibration; for example, the normal threshold can be set to 80 points, and the abnormal threshold to 40 points. When the value is greater than the normal threshold, the communication status is judged as normal; when If the value is below the abnormal threshold, it is considered abnormal; when When the communication state falls between these two extremes, it will be determined that the communication status has entered an intermediate warning state. This final determined communication state will serve as a trigger condition to drive subsequent control mode switching logic.

[0083] For example, the system acquires key performance parameters of the communication link through the shipborne communication management unit at 10-second intervals. Within a certain monitoring period, the system measures the real-time data transmission delay. The signal strength normalization score is 110 milliseconds. The normalized score for packet loss is 0.75. The value is 0.85. The optimal delay is set based on the preset communication technology type. The worst acceptable latency is 20 milliseconds. The delay is 200 milliseconds. First, the normalized fraction of the delay is calculated using the formula. The calculation process is as follows: =0.5. Subsequently, the system pre-configures weighting coefficients based on the degree of communication dependence of the ballast water treatment task. It is 0.2 It is 0.4. The value is 0.4. The final communication quality score is calculated using a weighted summation formula. The calculation process is as follows: The system compares the score with the preset normal threshold of 0.8 and the abnormal threshold of 0.4. Since 0.69 falls between 0.4 and 0.8, the system determines the current communication status to be in a warning state and sends a warning notification to both the ship and shore ends, indicating a decline in communication quality and a possible switch to autonomous propulsion mode.

[0084] Optionally, controlling the ship's end dynamic digital mirror to switch to autonomous propulsion mode includes:

[0085] The communication quality score is continuously calculated and recorded. When the communication quality score is lower than the normal threshold but higher than the abnormal threshold within a monitoring period, the communication status is determined to be an early warning status, and an early warning notification is sent to both the ship and shore ends.

[0086] When the communication quality score falls below the abnormal threshold in any monitoring period, the communication status is immediately determined to be abnormal, and a switching command is triggered.

[0087] Based on the switching command, the currently running cooperative strategy vector is locked and used as the fixed constraint in the autonomous evolution mode, and an autonomous control loop with local inference as the core is started.

[0088] Specifically, this switching control logic is first manifested as a continuously operating monitoring and early warning layer. The system continuously calculates and records the communication quality score defined in the prior claims, with a monitoring cycle of 5 to 10 seconds. When this communication quality score first drops to a preset normal threshold (e.g., below 80 points) within a monitoring cycle, but remains above a preset abnormal threshold (e.g., 40 points), the system determines that the communication status has entered an early warning state. This state triggers an early warning notification, which is simultaneously sent in the form of a structured message to the ship's operating interface and the shore-based monitoring center. The notification includes the current communication quality score, the specific indicators that caused the score to drop (e.g., data transmission latency increased to 150 milliseconds), and a clear prompt that "communication quality has deteriorated, and switching to automatic transition mode may be necessary," allowing operators to intervene or prepare in advance.

[0089] If the communication status deteriorates further, a decisive switchover action will be executed. If the communication quality score falls below the abnormal threshold (e.g., below 40 points) in any monitoring period, the ship's control core will no longer wait but will immediately determine the communication status as abnormal and autonomously trigger an internal, high-priority switchover command. This command is the starting point for the entire mode transition.

[0090] Based on the switching command, a series of chain actions are immediately executed to initiate the autonomous evolution mode. First, the currently running cooperative strategy vector is locked, transforming the most recently received and verified cooperative strategy vector from the shore-based system from a dynamically awaiting-update input into an internal static parameter. This means that the processing target priority, control parameter adjustment boundaries, parameter coupling relationships, and anomaly response principles contained in this vector will become the highest criteria for local decision-making for the foreseeable future. Subsequently, this locked cooperative strategy vector is set as the fixed constraint for the ship's dynamic digital image in autonomous evolution mode. The autonomous control loop, centered on local simulation, is then formally initiated. In this loop, the ship's dynamic digital image will no longer seek or wait for shore-based strategy updates, but will rely entirely on this fixed constraint, continuously using the latest acquired real-time multi-source sensor data as boundary conditions to perform independent simulations and optimization calculations, uninterruptedly generating and issuing autonomous evolution mode equipment control command sets, thereby achieving intelligent autonomous operation decoupled from the shore-based system.

[0091] For example, the system continuously calculates and records communication quality scores. The system uses a switching control logic to drive the communication. Within a monitoring cycle, the system measures a real-time communication quality score. The score is 0.35. This rating... If the communication status falls below the preset anomaly threshold of 0.4, the ship's control core immediately determines the communication status as abnormal and triggers a switchover command. Based on this switchover command, the system immediately locks the currently running cooperative strategy vector, fixing the most recently received cooperative strategy vector from the shore-based system as an internal static parameter. This locked cooperative strategy vector includes the priority of the processing target. 0.8 and The system employs key constraints such as a value of 0.2 and a UV lamp power adjustment boundary of 70% to 110%. This vector is then used as a fixed constraint in the autonomous evolution mode, and the autonomous control loop centered on local simulation is formally initiated. In this loop, the ship's dynamic digital image no longer waits for shore-based updates but continuously performs independent simulations based on the latest acquired real-time multi-source sensor data, such as the current flow rate of 200 cubic meters per hour and turbidity of 30 NTU, within the fixed constraint framework. Through iterative calculations using a particle swarm optimization algorithm, the system generates the optimal combination of virtual control parameters and converts it into an autonomous evolution mode equipment control command set for execution, thereby achieving intelligent autonomous operation decoupled from the shore-based system during communication interruptions.

[0092] Specifically, when the communication status monitoring and autonomous evolution control module determines that the quality score of the communication link is below the abnormal threshold, it automatically triggers a logical switch from cooperative control to autonomous control, putting the system into autonomous evolution mode. In this mode, the shipboard control core first executes the switching command and locks the most recently acquired cooperative strategy vector from the shore-based system. The cooperative strategy vector is a structured guidance document that includes the priority of macro-level processing objectives, equipment operation adjustment boundaries, and abnormal response principles. Using it as a fixed constraint ensures that local decision-making remains within a safe and compliant framework even without real-time updates from the shore-based system. Subsequently, the latest real-time multi-source sensor data is continuously acquired through the shipboard data bus. This data covers turbidity and salinity reflecting the water quality environment, flow rate and differential pressure characterizing the system load, and ship attitude parameters affecting fluid stability. The shipboard dynamic digital mirror, a digital simulation model that can accurately map the physical system's operating state and fluid dynamic characteristics, uses this real-time data as initial boundary conditions for state alignment. Then, the locked cooperative strategy vector is parsed into an objective function for simulation derivation, driving the digital mirror to perform multiple rounds of forward simulation prediction. During the simulation, the shipborne computing unit generates a series of virtual control parameter combinations, and the mirror model evaluates the expected processing performance of each combination under the current operating conditions. Finally, based on the evaluation results, the optimal parameter combination is determined and translated into an autonomous evolution mode equipment control instruction set. This allows for precise control of the physical actuators via an industrial fieldbus, achieving intelligent closed-loop regulation even under communication anomalies.

[0093] S5. Under normal communication conditions, the ship-end ballast water treatment equipment is controlled to perform operations based on the ship-shore collaborative mode equipment control instruction set. Under abnormal communication conditions, the ship-end ballast water treatment equipment is controlled to perform operations based on the autonomous evolution mode equipment control instruction set.

[0094] Optionally, controlling the ship's ballast water treatment equipment to perform operations based on the ship-shore collaborative mode equipment control command set includes:

[0095] The ship-shore collaborative mode equipment control instruction set is parsed and converted into control signals for the corresponding specific execution equipment.

[0096] The control signal is sent to the corresponding ballast water treatment equipment;

[0097] The execution feedback signal of the ballast water treatment equipment is monitored in real time to confirm whether the equipment operates correctly according to the instructions;

[0098] The execution feedback signal is fed back to the ship's end dynamic digital mirror as part of the real-time multi-source sensor data.

[0099] Specifically, the system parses the ship-shore collaborative mode equipment control instruction set generated by the ship's dynamic digital mirror. This instruction set is a structured data packet containing a description of the target state of each device, such as "ballast pump P-101 speed set to 80% of rated value" or "UV disinfection unit UV-01 power set to 95%". The actuator drive module in the control system is responsible for parsing these semantic instructions and converting them into directly identifiable control signals for the specific actuators. For example, the aforementioned ballast pump speed instruction is converted into a specific frequency value, such as 48 Hz, to drive its inverter; the UV unit power instruction is converted into a specific 4 to 20 mA analog current signal, such as sending 19.2 mA of current to its electronic ballast controller.

[0100] After conversion, these precise control signals are sent to the corresponding ballast water treatment equipment via the ship's onboard industrial fieldbus. Each device or its associated local controller, such as a programmable logic controller (PLC), has a unique bus address. Control signals are broadcast on the bus or sent point-to-point in the form of messages with addresses and data, ensuring that each instruction accurately reaches the target device.

[0101] After the command is issued, the interaction does not terminate but immediately transitions to the monitoring phase. The system polls in real time or passively receives execution feedback signals from the ballast water treatment equipment via the bus. These signals are not processing effect data, but rather status confirmation signals for the equipment itself, used to confirm whether the equipment is operating correctly according to the command. For example, for an electric valve, the execution feedback signal is the limit switch signal at both ends of the valve's stroke. After the command "fully open," the system expects to receive a "fully open" limit switch closure signal within a preset time, such as 5 seconds. For a frequency converter, the feedback signal is the actual output frequency and motor operating current.

[0102] These collected execution feedback signals are formatted and fed back to the ship's dynamic digital mirror as an important component of real-time multi-source sensor data. This constitutes the inner closed loop of the control execution layer. When the ship's dynamic digital mirror receives a feedback signal such as "valve V-102 is open," it immediately updates the valve's state to "open" in its internal equipment and piping topology model. This ensures that the fluid path topology in the digital mirror is completely consistent with the physical world, laying the foundation for the next round of simulation based on a more accurate model state and avoiding model distortion caused by command execution delays or failures.

[0103] For example, the system first parses the ship-shore collaborative mode equipment control instruction set generated by the ship's dynamic digital mirror. This instruction set contains a description of the target state of specific equipment, such as setting the speed of ballast pump P-101 to 80% of its rated value and the power of ultraviolet disinfection unit UV-01 to 95%. The actuator drive module converts these semantic instructions into specific execution signals, calculates the specific frequency value of the ballast pump inverter to be 48Hz based on the rated frequency of 60Hz, and converts the ultraviolet unit power instruction into a standard current signal, which is calculated to be equal to 19.2mA. Subsequently, the system sends these control signals to the corresponding equipment via the ship's industrial fieldbus. After the instructions are sent, the system immediately enters the monitoring phase, polling the execution feedback signals of the ballast water treatment equipment in real time. For example, for the electric valve V-102 set to the fully open instruction, the system successfully receives the limit switch closing signal within the preset 5s time, and simultaneously monitors that the actual output frequency fed back by the ballast pump inverter is 48Hz. Finally, the system feeds these formatted execution feedback signals as part of the real-time multi-source sensor data back to the ship's dynamic digital mirror. Upon receiving the feedback, the mirror immediately updates the valve and pump status in the equipment and pipeline topology model to ensure that the topology of the digital mirror is completely consistent with the physical world, laying the foundation for the next round of simulation.

[0104] Specifically, when the communication quality of the communication link drops below a preset threshold, a switch from ship-to-shore collaborative mode to autonomous evolution mode is executed, and the ballast water treatment equipment is precisely controlled based on the autonomous evolution mode equipment control instruction set. First, the shipboard communication management unit periodically acquires key performance parameters of the communication link, with the monitoring period typically set between 5 and 10 seconds. The acquired parameters include signal strength, data transmission latency, and packet loss rate. Using a weighted evaluation model, these raw data with different physical dimensions are normalized and transformed into a unified dimensionless score, thereby calculating a comprehensive communication quality score. Normalization ensures the logical comparability of each parameter; for example, the lower the latency and the stronger the signal, the closer the corresponding normalized score is to the ideal full score, allowing the final score to objectively reflect the true reliability of the communication link. When the communication quality score falls below a preset abnormal threshold, the shipboard control core determines the communication status as abnormal and immediately triggers a switchover command. Based on this instruction, the most recently received and verified cooperative strategy vector is locked as a fixed constraint, solidifying the processing target priority, control parameter adjustment boundaries, parameter coupling relationships, and anomaly response principles. Subsequently, an autonomous control loop centered on local simulation is initiated, driving the ship's dynamic digital mirror to continuously perform multiple rounds of forward simulation prediction based on the latest acquired real-time multi-source sensor data such as water quality, equipment, and ship attitude, independently generating an autonomous evolution mode equipment control instruction set. After obtaining the instruction set, the ballast water treatment equipment execution control module parses the instruction set and converts it into control signals for the corresponding specific execution equipment. For example, the ballast pump speed in the virtual control parameters is converted into a specific frequency setpoint conforming to the inverter communication protocol, or the UV lamp power is converted into an analog current or voltage signal recognizable by the electronic ballast controller. These precise control signals are sent to the corresponding hardware devices such as ballast pumps, filters, and UV disinfection units via the ship's industrial fieldbus. Meanwhile, the system monitors the execution feedback signals of each device in real time to confirm whether the physical devices are acting correctly according to the instructions, and sends the feedback data back to the ship's dynamic digital mirror. By updating the topology model state inside the mirror, it ensures high-fidelity synchronization between the digital world and the physical world.

[0105] Optionally, the method further includes:

[0106] Acquire sensor data to verify the effect;

[0107] The effect verification sensor data is fed back to the ship's dynamic digital mirror;

[0108] The ship's end dynamic digital mirror compares the preset expected processing effect with the actual processing effect obtained based on the effect verification sensor data to generate effect deviation data.

[0109] Based on the effect deviation data, the ship-side dynamic digital mirror adaptively adjusts the local parameters of its internal inference algorithm and re-performs the simulation to generate a corrected ship-shore cooperative mode equipment control instruction set.

[0110] Specifically, the acquisition of effect verification sensor data, which is dedicated to evaluating the final effectiveness of ballast water treatment, comes from specialized sensors deployed on the outlet pipeline of the ballast water treatment system. These sensors include, for example, an adenosine triphosphate (ATP) fluorescence detector for real-time assessment of microbial activity, or a chlorophyll fluorescence probe for measuring phytoplankton concentration. Readings from these sensors, such as relative fluorescence units per milliliter of water sample, directly reflect the survival status of pathogens after treatment.

[0111] The acquired raw effect verification sensor data is fed back to the ship's dynamic digital mirror in real time. The mirror compares the preset expected processing effect with the actual processing effect calculated based on this sensor data. The expected processing effect is the target value that the ship's dynamic digital mirror should achieve based on its internal model in the previous round of simulation, according to the optimal control strategy, such as the predicted biological inactivation rate. The actual processing effect This is calculated by processing the readings of the sensors before and after processing. The calculation method is as follows:

[0112] ,

[0113] Here, For pretreatment of water quality sensor readings, such as turbidity meters. To validate the effects, sensor data—readings from an ATP analyzer or similar device after treatment—were normalized to reflect the relative change in biomass. Deviation data were then calculated. :

[0114] ,

[0115] This effect deviation data This intuitively quantifies the gap between model predictions and physical reality. A positive value... A negative value indicates that the actual effect did not meet expectations, while a negative value indicates that the effect exceeded expectations.

[0116] Based on this performance deviation data, the ship's dynamic digital mirror will adaptively adjust the local parameters of its internal inference algorithm. This process does not directly modify the equipment control commands, but rather corrects the physical or chemical model of the digital mirror itself. For example, if A continuously positive value indicates poor actual inactivation effect, and the system will activate a PI controller to... As input, a key internal parameter of the UV disinfection model in the digital mirror is fine-tuned, such as the UV attenuation coefficient of water. The adjustment process follows this logic:

[0117] ,

[0118] in, and These are the UV attenuation coefficients before and after adjustment. and These are the proportional and integral adjustment coefficients, which are preset empirical values, typically ranging from 0.01 to 0.5. By increasing... In subsequent simulations, the digital mirror assumes that the water body absorbs ultraviolet light more strongly. Therefore, to achieve the same inactivation rate target, a control strategy requiring a higher ultraviolet dose must be generated. After completing the adaptive adjustment of the internal model parameters, the ship-side dynamic digital mirror immediately uses the updated model and combines it with the latest real-time multi-source sensor data to re-simulate and generate a set of corrected ship-shore cooperative mode equipment control commands, which are then issued to the physical equipment for execution, thus forming a fast-responding and self-improving closed-loop control loop.

[0119] For example, the system first acquires effect verification sensor data using an adenosine triphosphate (ATP) fluorescence detector deployed on the outlet pipeline. In this verification phase, the initial biomass measured by the water quality sensor before treatment is set. The processing efficiency is 1000 RLU, and the effect is verified by analyzing the sensor data. The value is 20 RLU. The system feeds this data back to the ship's dynamic digital mirror, using the formula... Calculate the actual processing effect, that is 98 Subsequently, the mirror image will reflect the expected processing results preset in the previous simulation. Set to 99.9%, and combine the actual processing effect with the formula. Generate performance deviation data, i.e. Based on this positive deviation, the ship's dynamic digital mirror initiates an adaptive adjustment mechanism to correct local parameters of its internal extrapolation algorithm. An initial value for the ultraviolet attenuation coefficient is set. The proportional adjustment coefficient is 0.15. The integral adjustment coefficient is 0.1. The value is 0.02, and it is assumed that the cumulative value of the current integral term is 0.5. According to the formula... Perform calculations, that is Finally, the mirror image uses the corrected parameters to re-simulate and generate a set of corrected ship-shore cooperative mode equipment control commands that require higher power output, thereby completing the self-improvement of the control loop.

[0120] Optionally, the method further includes:

[0121] When the communication status of the communication link is detected to have returned to normal from abnormal, the autonomous evolution log recorded in the autonomous evolution mode is extracted. The autonomous evolution log includes an environmental data sequence, a locally generated autonomous evolution mode device control instruction set sequence, an effect verification sensor data sequence, and the local optimization strategy variant information.

[0122] Upload the auto-action log to the shore-based data center;

[0123] The shore-based data center uses the autonomous evolution log to update the preset regional historical database, and retrains and optimizes the shore-based policy generation model based on the updated regional historical database.

[0124] Specifically, when the communication quality score consistently exceeds the preset normal threshold for several consecutive monitoring periods, such as three consecutive periods, it is confirmed that communication has returned to a highly reliable state. At this time, the shipboard system automatically performs a data extraction task, packaging the autonomous evolution log stored in local non-volatile memory. This autonomous evolution log is not a single file, but a structured dataset containing multi-dimensional time-series data. Its core contents include: first, environmental data sequences, i.e., time series of complete water quality parameters and ship attitude parameters recorded during the autonomous evolution mode; second, locally generated autonomous evolution mode equipment control command set sequences, i.e., all equipment control commands generated by the shipboard dynamic digital mirror after each round of autonomous simulation without shore-based guidance; third, effect verification sensor data sequences, i.e., time series of actual treatment effect data fed back by equipment such as ATP detectors; and fourth, local optimization strategy variant information, which is key learning data, recording the entire process of the shipboard dynamic digital mirror adaptively adjusting its internal model parameters to cope with effect deviations.

[0125] This compressed and encrypted autonomous evolution log data packet is uploaded to the shore-based data center via the restored high-bandwidth communication link. Upon receiving the log, the data center first decompresses and verifies it, then initiates the background data processing flow. The shore-based data center utilizes the environmental data sequences, autonomous evolution mode device control command set sequences, and effect verification sensor data sequences from the autonomous evolution log to construct one or more entirely new historical operational case studies with complete causal chains. The unique value of this case study lies in its representation of a process of autonomous closed-loop solution and verification by the system under specific real-world environmental constraints. This new case study is added and indexed into a pre-defined regional historical database, greatly enriching the knowledge base for that specific navigation area or sea state.

[0126] The shore-based data center periodically retrains and optimizes the shore-based policy generation model based on an updated regional historical database, such as during off-peak hours each day or after accumulating a certain number of new logs. During retraining, the local optimization policy variant information recorded in the autonomous evolution logs plays a crucial role. It provides valuable supervision signals for the machine learning model, revealing the gap between the initial and optimal policies under specific operating conditions, as well as the adjustment path to the optimal solution. By adding these high-value samples to the training set, the shore-based policy generation model can learn more refined operating condition recognition capabilities and more robust policy generation logic. This allows it to respond more proactively and accurately to similar operating conditions when issuing cooperative policy vectors to other vessels in the future, fundamentally improving the intelligence level of the entire ship-shore linkage system.

[0127] For example, when the system monitors the communication quality score When the turbidity remains consistently above the preset normal threshold of 0.8 for three consecutive monitoring cycles, the communication status is confirmed to have returned to normal. The shipboard system immediately extracts the autonomous propulsion log recorded in autonomous propulsion mode. This log includes water quality parameters such as turbidity (45 NTU) recorded in the environmental data sequence, ultraviolet power (105%) recorded in the autonomous propulsion mode equipment control command set sequence, and actual biomass measured in the effect verification sensor data sequence. For 15 RLU and local model parameters Information on the optimized strategy variant, adjusted from 0.15 to 0.38. The system uploads the encrypted data packet to the shore-based data center via a high-bandwidth link, where the data center first utilizes actual biomass. Compared with biomass before treatment For data of 1200 RLU, using the formula Calculate the actual processing effect under this working condition, i.e. This equals 98.75%. Subsequently, the shore-based data center uses this data, which contains complete causal chains, to update the pre-set regional historical database, and retrains and optimizes the shore-based policy generation model based on the updated database. During the retraining process, the system compares the initial policy with the final optimal parameters. The path is 0.38. By using these high-value samples to correct the model logic, the collaborative strategy vectors issued in the future can more accurately preset the processing parameters when facing high turbidity conditions such as 45NTU.

[0128] Based on the same inventive concept, such as Figure 4 As shown, the present invention also provides a control system for real-time treatment of pathogens in ballast water in a ship-shore integrated manner, the system comprising:

[0129] The data acquisition and digital image construction module is used to acquire basic configuration information and real-time multi-source sensor data, and to construct and maintain a dynamic digital image of the ship's end that represents the real-time state.

[0130] The shore-based strategy generation and communication module is used to generate a collaborative strategy vector for the ship-end ballast water treatment process by calculating based on the mirror state data of the ship-end dynamic digital mirror and the acquired ship navigation information through a preset shore-based strategy generation model.

[0131] The ship-side strategy simulation and control command generation module is used to inject the cooperative strategy vector into the ship-side dynamic digital image and, in combination with the latest acquired real-time multi-source sensor data, drive the ship-side dynamic digital image to perform simulation simulation and generate a set of ship-shore cooperative mode equipment control commands.

[0132] The communication status monitoring and autonomous evolution control module is used to monitor the communication status of the communication link. When the communication status changes from normal to abnormal, it controls the ship-end dynamic digital mirror to switch to autonomous evolution mode. In the autonomous evolution mode, the ship-end dynamic digital mirror uses the most recently received cooperative strategy vector as a fixed constraint, continuously performs simulation and deduction based on the latest acquired real-time multi-source sensor data, and generates an autonomous evolution mode device control instruction set.

[0133] The ballast water treatment equipment execution control module is used to control the ship-end ballast water treatment equipment to perform operations based on the ship-shore collaborative mode equipment control instruction set when communication is normal, and to control the ship-end ballast water treatment equipment to perform operations based on the autonomous evolution mode equipment control instruction set when communication is abnormal.

[0134] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0135] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A control method for real-time treatment of pathogens in ballast water using a ship-shore integrated system, characterized in that, The method includes: Acquire basic configuration information and real-time multi-source sensor data, and construct and maintain a dynamic digital image of the ship's end that represents the real-time state; Based on the mirror state data of the ship's end dynamic digital mirror and the acquired ship navigation information, a collaborative strategy vector for the ship's end ballast water treatment process is generated by calculating through a preset shore-based strategy generation model. The collaborative strategy vector is injected into the ship-side dynamic digital image, and combined with the latest acquired real-time multi-source sensor data, the ship-side dynamic digital image is driven to perform simulation and generate a set of ship-shore collaborative mode equipment control instructions. Monitor the communication status of the communication link. When the communication status changes from normal to abnormal, control the ship-end dynamic digital mirror to switch to autonomous evolution mode. In the autonomous evolution mode, the ship-end dynamic digital mirror uses the most recently received cooperative strategy vector as a fixed constraint, continuously performs simulation and deduction based on the latest acquired real-time multi-source sensor data, and generates an autonomous evolution mode device control instruction set. Under normal communication conditions, the ship-end ballast water treatment equipment is controlled to perform operations based on the ship-shore collaborative mode equipment control instruction set. Under abnormal communication conditions, the ship-end ballast water treatment equipment is controlled to perform operations based on the autonomous evolution mode equipment control instruction set.

2. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The construction and maintenance of the ship's end dynamic digital mirror representing the real-time state includes: Based on the aforementioned basic configuration information, establish a topology model of the equipment and pipelines; The system receives real-time multi-source sensor data, which includes water quality parameters, equipment operating parameters, and ship attitude parameters. The real-time multi-source sensor data is input into the device and pipeline topology model for calculation, and the current system state vector composed of multiple state dimension indicators is output in real time. Based on a preset update frequency to ensure synchronization, the ship's dynamic digital image is constructed and updated using the current system state vector.

3. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The generated cooperative strategy vector for the ship's end ballast water treatment process includes: Key operating condition features are extracted from the mirror state data and associated with the navigation area and planned route in the ship navigation information to form a real-time operating condition context at the ship's end. Based on the real-time operating condition context at the ship's end, a set of historical similar operating condition cases is retrieved and generated; The shipboard real-time operating condition context and the set of historical similar operating condition cases are input into the shore-based strategy generation model. The output of the shore-based strategy generation model includes a collaborative strategy vector that includes the priority of processing targets, the boundary of control parameter adjustment, the coupling relationship between parameters, and the principle of anomaly response.

4. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The process of driving the ship-side dynamic digital mirror to perform simulation and generate a set of control instructions for ship-shore collaborative mode equipment includes: The collaborative strategy vector is parsed into the objective function and constraints of the ship's dynamic digital image for deduction and optimization. The newly acquired real-time multi-source sensor data is used as the boundary input condition for the ship's end dynamic digital image. Within the framework of the objective function and constraints, multiple rounds of forward simulation calculations are performed on the ship's end dynamic digital image to evaluate and generate virtual control parameter combinations; The virtual control parameter combination is converted into a set of control instructions for ship-shore cooperative mode equipment.

5. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The method further includes: Acquire sensor data to verify the effect; The effect verification sensor data is fed back to the ship's dynamic digital mirror; The ship's end dynamic digital mirror compares the preset expected processing effect with the actual processing effect obtained based on the effect verification sensor data to generate effect deviation data. Based on the effect deviation data, the ship-side dynamic digital mirror adaptively adjusts the local parameters of its internal inference algorithm and re-performs the simulation to generate a corrected ship-shore cooperative mode equipment control instruction set.

6. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The monitoring of the communication status of the communication link includes: The signal strength, data transmission delay, and packet loss rate parameters of the communication link are periodically acquired. The signal strength, data transmission delay, and packet loss rate parameters are weighted and evaluated to calculate the communication quality score. The communication quality score is compared with preset normal thresholds for determining normal states and abnormal thresholds for determining abnormal states to determine the communication status.

7. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 6, characterized in that, The control of switching the ship's dynamic digital mirror to the automatic propulsion mode includes: The communication quality score is continuously calculated and recorded. When the communication quality score is lower than the normal threshold but higher than the abnormal threshold within a monitoring period, the communication status is determined to be an early warning status, and an early warning notification is sent to both the ship and shore ends. When the communication quality score falls below the abnormal threshold in any monitoring period, the communication status is immediately determined to be abnormal, and a switching command is triggered. Based on the switching command, the currently running cooperative strategy vector is locked and used as the fixed constraint in the autonomous evolution mode, and an autonomous control loop with local inference as the core is started.

8. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The operation of the ship-end ballast water treatment equipment based on the ship-shore collaborative mode equipment control command set includes: The ship-shore collaborative mode equipment control instruction set is parsed and converted into control signals for the corresponding specific execution equipment. The control signal is sent to the corresponding ballast water treatment equipment; The execution feedback signal of the ballast water treatment equipment is monitored in real time to confirm whether the equipment operates correctly according to the instructions; The execution feedback signal is fed back to the ship's end dynamic digital mirror as part of the real-time multi-source sensor data.

9. The control method for real-time treatment of pathogens in ballast water using a ship-shore linkage system according to claim 1, characterized in that, The method further includes: When the communication status of the communication link is detected to have returned to normal from abnormal, the autonomous evolution log recorded in the autonomous evolution mode is extracted. The autonomous evolution log includes an environmental data sequence, a locally generated autonomous evolution mode device control instruction set sequence, an effect verification sensor data sequence, and the local optimization strategy variant information. Upload the auto-action log to the shore-based data center; The shore-based data center uses the autonomous evolution log to update the preset regional historical database, and retrains and optimizes the shore-based policy generation model based on the updated regional historical database.

10. A control system for real-time treatment of pathogens in ballast water using a ship-shore integrated system, applied to the control method for real-time treatment of pathogens in ballast water using a ship-shore integrated system as described in any one of claims 1-9, characterized in that, The system includes: The data acquisition and digital image construction module is used to acquire basic configuration information and real-time multi-source sensor data, and to construct and maintain a dynamic digital image of the ship's end that represents the real-time state. The shore-based strategy generation and communication module is used to generate a collaborative strategy vector for the ship-end ballast water treatment process by calculating based on the mirror state data of the ship-end dynamic digital mirror and the acquired ship navigation information through a preset shore-based strategy generation model. The ship-side strategy simulation and control command generation module is used to inject the cooperative strategy vector into the ship-side dynamic digital image and, in combination with the latest acquired real-time multi-source sensor data, drive the ship-side dynamic digital image to perform simulation simulation and generate a set of ship-shore cooperative mode equipment control commands. The communication status monitoring and autonomous evolution control module is used to monitor the communication status of the communication link. When the communication status changes from normal to abnormal, it controls the ship-end dynamic digital mirror to switch to autonomous evolution mode. In the autonomous evolution mode, the ship-end dynamic digital mirror uses the most recently received cooperative strategy vector as a fixed constraint, continuously performs simulation and deduction based on the latest acquired real-time multi-source sensor data, and generates an autonomous evolution mode device control instruction set. The ballast water treatment equipment execution control module is used to control the ship-end ballast water treatment equipment to perform operations based on the ship-shore collaborative mode equipment control instruction set when communication is normal, and to control the ship-end ballast water treatment equipment to perform operations based on the autonomous evolution mode equipment control instruction set when communication is abnormal.

Images