Intelligent interlocking system that triggers the automatic start and stop of ship carbon capture devices when connected to shore power

Abstract

The present invention relates to the technical field of ship shore power access and carbon capture, and discloses an intelligent interlocking system for triggering the automatic start and stop of a ship's carbon capture device when shore power access is used. The system uses a shore power status monitoring module to obtain target electrical characteristic parameters such as voltage and current, as well as the access status, and index historical trigger data. A carbon capture device status acquisition module extracts historical operating parameters and start and stop records to generate a set of historical trigger parameters. After clustering and screening to obtain a historical trigger parameter sequence, a trigger trend prediction module combines sample data from multiple ships to predict the trigger response rate and delay rate. A dynamic interlocking correction module then matches the state characteristic group and corrects deviations. Finally, a start and stop decision generation module outputs a control instruction to execute automatic start and stop. The system implements intelligent interlocking control based on historical data and real-time electrical parameters, improving the accuracy and automation level of carbon capture device start and stop, and effectively optimizing the ship's energy conservation and emission reduction efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of ship shore power access and carbon capture, and specifically to an intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when shore power access is triggered. Background Art

[0002] The shipping industry faces significant challenges and urgent needs for green, low-carbon development. While docked, ships typically rely on their own fuel-powered generators for power. This generates significant amounts of pollutants, including carbon dioxide, sulfur oxides, and nitrogen oxides, severely impacting the air quality and ecological environment surrounding the port. Statistics show that pollutant emissions from docked ships account for a significant proportion of overall port emissions. This not only violates the concept of low-carbon environmental protection but also contradicts current global action to combat climate change.

[0003] With the development of shore power technology, connecting ships to shore power while docked has become an important means of reducing pollutant emissions. Shore power refers to the land-based power supply provided by ports to docked ships. Once connected to shore power, ships can stop operating their own fuel-powered generators, significantly reducing pollutant emissions. However, there are currently significant technical shortcomings in the coordinated control of ship shore power access and carbon capture devices.

[0004] Existing ship carbon capture device (CCD) start and stop controls mostly rely on manual operation or simple preset programs. Manual operation has obvious limitations. First, the experience and operating skills of operators vary widely, which can easily lead to operational errors and make it impossible to ensure that the CCD device starts and stops at the optimal time. Second, manual operation is difficult to accurately adjust to the actual shore power connection situation in real time, which may cause the CCD device to start and stop prematurely or delayed. Premature start and stop will cause the CCD device to operate unnecessarily, increasing energy consumption and equipment wear. Delayed start and stop will result in some pollutants not being effectively captured during the initial shore power connection period, affecting environmental protection.

[0005] Simple preset program control, due to its relatively fixed logic, cannot flexibly adapt to the complexities of different ships, different berthing conditions, and different shore power access conditions. The performance parameters of carbon capture devices vary between ships, and ships' power demands and carbon emissions vary under different berthing conditions. The electrical characteristic parameters of different shore power access conditions vary significantly. Preset programs fail to fully account for these factors, resulting in inaccurate start and stop control of the carbon capture device, which cannot fully utilize its carbon capture efficiency, and to some extent, affects the low-carbon emission reduction effect of ships while in port.

[0006] Current technologies lack in-depth research and effective utilization of the dynamic correlation between shore power access status and the start and stop of carbon capture devices. Electrical characteristic parameters during shore power access, such as voltage, current, and access timestamp, often contain rich information that is closely related to the optimal start and stop timing of the carbon capture device. However, existing technologies fail to organically integrate these electrical characteristic parameters with the start and stop control of the carbon capture device, making it impossible to achieve automatic and precise start and stop control of the carbon capture device based on shore power access status. Summary of the Invention

[0007] The purpose of the present invention is to provide an intelligent interlocking system for triggering the automatic start and stop of a ship's carbon capture device when shore power is connected, so as to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention provides the following technical solution: an intelligent interlocking system for triggering the automatic start and stop of a ship's carbon capture device when shore power is connected, the system comprising:

[0009] A shore power status monitoring module, which is used to obtain the current shore power access status information of the ship and collect target electrical characteristic parameters during the shore power access process. Based on the shore power access status information, the module indexes the historical triggering data of shore power access and carbon capture device start and stop within the ship's history.

[0010] a carbon capture device status collection module, the carbon capture device status collection module being configured to extract features of the carbon capture device operating status information in the historical trigger data to obtain a plurality of status feature groups, extract a set of historical operating parameters and a plurality of sets of historical start and stop operation records under the plurality of status feature groups, and process the extracted data to obtain a plurality of sets of historical trigger parameters;

[0011] A historical trigger data clustering module, the historical trigger data clustering module is used to cluster and filter the multiple historical trigger parameter sets respectively to obtain multiple filtered trigger parameter groups, and arrange them in chronological order to obtain multiple historical trigger parameter sequences;

[0012] a trigger trend prediction module, configured to perform trigger trend prediction based on the plurality of historical trigger parameter sequences to obtain a trigger response rate and a delay rate;

[0013] a dynamic interlock correction module, configured to match the target electrical characteristic parameters with the plurality of state characteristic groups to obtain a matching state characteristic group, and to correct the trigger response rate and delay rate based on a deviation between the target electrical characteristic parameters and a standard matching characteristic parameter of the matching state characteristic group to obtain a corrected response rate and a corrected delay rate;

[0014] A start-stop decision generation module is used to make a start-stop plan decision based on the corrected response rate and the corrected delay rate, obtain a start-stop control instruction, and execute automatic start-stop operations of the carbon capture device.

[0015] Preferably, the current shore power access status information of the ship is obtained, and the target electrical characteristic parameters during the shore power access process are collected. According to the shore power access status information, historical triggering data of shore power access and start and stop of the carbon capture device of the ship in the historical time are indexed, including:

[0016] Obtain the current voltage, current, and access timestamp of the ship's shore power connection as target electrical characteristic parameters;

[0017] According to the shore power access status information, an index is performed in the historical operation record of the ship to obtain historical triggering data of shore power access and start and stop of the carbon capture device.

[0018] Preferably, feature extraction is performed on the carbon capture device operating status information in the historical trigger data to obtain multiple status feature groups, and historical operating parameter sets and multiple historical start-stop operation record sets under the multiple status feature groups are extracted, and the multiple historical trigger parameter sets are obtained by processing, including:

[0019] Acquiring operation status information of multiple carbon capture devices in the historical trigger data, performing feature extraction processing, and obtaining multiple state feature groups;

[0020] Extracting standard operating power and temperature thresholds of the carbon capture device from the historical trigger data under the multiple state feature groups to obtain a historical operating parameter set, and extracting multiple historical voltage threshold sets and multiple historical connection duration sets for triggering start and stop operations of the ship under the multiple state feature groups;

[0021] Classify and obtain multiple historical basic parameter sets according to the matching degree between the multiple historical voltage threshold sets and the historical operating parameter sets;

[0022] According to the ratio of the preset duration threshold and the multiple historical access duration sets, the multiple historical basic parameter sets are corrected and calculated to obtain multiple historical trigger parameter sets.

[0023] Preferably, cluster screening is performed on the multiple historical trigger parameter sets to obtain multiple filtered trigger parameter groups, and the groups are arranged in chronological order to obtain multiple historical trigger parameter sequences, including:

[0024] Selecting and obtaining a first benchmark trigger parameter from a first historical trigger parameter set in the plurality of historical trigger parameter sets;

[0025] Allocating weight coefficients according to the deviations between the other trigger parameters in the first historical trigger parameter set and the first benchmark trigger parameter to obtain a first basic weight distribution, wherein the deviations are negatively correlated with the weight coefficients;

[0026] performing cluster screening on the first historical trigger parameter set according to the first basic weight distribution to obtain a first screening trigger parameter group;

[0027] Sort the multiple trigger parameters in the first screening trigger parameter group according to their timestamp information to obtain a first historical trigger parameter sequence;

[0028] Cluster screening and time sorting are performed on other multiple historical trigger parameter sets to obtain multiple historical trigger parameter sequences.

[0029] Preferably, cluster screening is performed on the first historical trigger parameter set according to the first basic weight distribution to obtain a first screening trigger parameter group, including:

[0030] Randomly selecting a preset number of screening trigger parameters from the first historical trigger parameter set to obtain a first screening trigger parameter group;

[0031] Allocating weight coefficients according to deviations between the trigger parameters in the first screening trigger parameter group and the first benchmark trigger parameter to obtain a first screening weight distribution;

[0032] Calculating a matching degree between the first screening weight distribution and the first basic weight distribution as a first screening matching value;

[0033] Randomly selecting a preset number of trigger parameters from the first historical trigger parameter set again to obtain a second screening trigger parameter group, and processing to obtain a second screening matching value;

[0034] The cluster screening is continued until convergence, and the screening trigger parameter group with the largest screening matching value is output as the first screening trigger parameter group.

[0035] Preferably, performing trigger trend prediction based on the multiple historical trigger parameter sequences to obtain a trigger response rate and a delay rate includes:

[0036] According to the sample trigger data of multiple ships, a sample trigger parameter sequence set is collected, and a sample response rate set and a sample delay rate set are obtained according to the parameter change identifier in each sample trigger parameter sequence;

[0037] Using the sample trigger parameter sequence set as prediction input, and using the sample response rate set and the sample delay rate set as prediction output, to construct a trigger trend predictor;

[0038] Based on the trigger trend predictor, the trigger trend of the plurality of historical trigger parameter sequences is classified to obtain a plurality of characteristic response rates and a plurality of characteristic delay rates;

[0039] The similarities between the target electrical characteristic parameters and the multiple state characteristic groups are analyzed, and the multiple characteristic response rates and multiple characteristic delay rates are weightedly calculated according to the magnitude of the multiple characteristic similarities to obtain a trigger response rate and a delay rate.

[0040] Preferably, matching the target electrical characteristic parameter with the multiple state characteristic groups to obtain a matching state characteristic group, and correcting the trigger response rate and delay rate according to the deviation between the target electrical characteristic parameter and the standard matching characteristic parameter of the matching state characteristic group to obtain a corrected response rate and a corrected delay rate, including:

[0041] Selecting the state feature group with the greatest similarity as the matching state feature group, and obtaining the standard matching feature parameters of the matching state feature group;

[0042] setting an interlock correction coefficient according to a deviation between the target electrical characteristic parameter and a standard matching characteristic parameter of the matching state characteristic group;

[0043] The interlock correction coefficient is used to perform correction calculation on the trigger response rate and delay rate to obtain a corrected response rate and a corrected delay rate.

[0044] Preferably, making a start / stop plan decision based on the corrected response rate and the corrected delay rate, obtaining a start / stop control instruction, and executing the automatic start / stop operation of the carbon capture device include:

[0045] Collecting a set of sample corrected response rates and a set of sample corrected delay rates, and setting sample start / stop control instructions based on the magnitude of each sample corrected response rate and sample corrected delay rate to obtain a set of sample start / stop instructions, wherein each sample start / stop control instruction includes a start / stop time node, and the magnitude of the sample corrected response rate and the sample corrected delay rate is negatively correlated with the advance amount of the start / stop time node;

[0046] The sample corrected response rate set and the sample corrected delay rate set are used as decision inputs, and the sample start-stop instruction set is used as decision output to construct a start-stop plan decision maker;

[0047] The start-stop plan decision maker is used to make a start-stop plan decision on the corrected response rate and the corrected delay rate to obtain a start-stop control instruction.

[0048] Preferably, indexing is performed in the historical operation records of the vessel to obtain historical triggering data of shore power access and start and stop of the carbon capture device, including:

[0049] According to the voltage and current values ​​of the shore power connection, matching shore power connection events with the same voltage level and current range in the historical operation records;

[0050] The start and stop timestamps and operating status parameters of the carbon capture device in the matching events are extracted as the historical trigger data of shore power access and the start and stop of the carbon capture device.

[0051] Preferably, feature extraction processing is performed to obtain multiple state feature groups, including:

[0052] Normalizing the carbon capture device operating status information in the historical trigger data, and extracting power fluctuation coefficient and temperature stability as feature dimensions;

[0053] A density clustering algorithm is used to cluster the feature dimensions to obtain multiple state feature groups.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] The system uses the shore power status monitoring module to obtain shore power access status information and target electrical characteristic parameters, and indexes historical trigger data, providing a rich data source for subsequent intelligent control. The carbon capture device status acquisition module extracts and processes features from historical data to obtain multiple sets of historical trigger parameters, enabling the system to fully utilize the ship's previous operating experience and laying the foundation for precise control.

[0056] The historical trigger data clustering module clusters and filters historical trigger parameter sets, chronologically sorting them to generate a historical trigger parameter sequence. This process effectively eliminates interference from abnormal data, improving data reliability and validity, and enabling the system to better understand the patterns between shore power access and the start and stop of the carbon capture device. The trigger trend prediction module predicts trigger trends based on the historical trigger parameter sequence, obtaining trigger response rates and delay rates. This provides a scientific basis for predicting the start and stop decisions of the carbon capture device, enabling the system to predict the optimal start and stop timing in advance.

[0057] The dynamic interlock correction module matches the target electrical characteristic parameters with the state characteristic group and adjusts the trigger response rate and delay rate based on the deviation. This ensures that the system maintains high control accuracy under different shore power access conditions and ship operating conditions, improving the system's adaptability and flexibility. The start-stop decision generation module generates start-stop control commands based on the corrected response rate and delay rate, enabling automatic start and stop operations of the carbon capture device. This significantly reduces manual intervention and operator workload, while also improving the timeliness and accuracy of operations.

[0058] The system accurately predicts the trigger response rate and delay rate of the carbon capture device based on real-time shore power voltage, current, and other electrical characteristic parameters, combined with the ship's historical operating data. Through a dynamic correction mechanism, the predictions are more consistent with actual operating conditions. In practical applications, the system can achieve seamless linkage between the carbon capture device and shore power access. When the shore power access status changes, the system can quickly respond, automatically starting and stopping the carbon capture device, ensuring that the carbon capture device is put into operation at the optimal time, thereby maximizing carbon capture efficiency and effectively reducing the ship's carbon emissions.

[0059] By analyzing and learning from historical data, the system continuously optimizes control strategies to adapt to the operational needs of different ships and operating conditions. This data-driven intelligent control approach not only enhances the system's intelligence but also provides an efficient and reliable solution for energy conservation and emission reduction on board ships, playing a significant role in promoting green and low-carbon development in the shipping industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a working principle diagram of the intelligent interlocking system for triggering the automatic start and stop of a ship's carbon capture device by shore power access according to the present invention;

[0061] Figure 2 This is the design diagram of the shore power status monitoring module;

[0062] Figure 3 This is a design diagram of the carbon capture device status collection module;

[0063] Figure 4 This is the design diagram of the historical trigger data clustering module;

[0064] Figure 5 This is the design diagram for triggering the trend prediction module. DETAILED DESCRIPTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0066] See also Figure 1-Figure 5 The present invention relates to an intelligent interlocking system for triggering the automatic start and stop of a ship's carbon capture device by shore power access. The specific implementation steps are as follows:

[0067] The shore power status monitoring module is used to obtain the current shore power access status information of the ship and collect the target electrical characteristic parameters during the shore power access process. Based on the shore power access status information, the module indexes the historical trigger data of shore power access and carbon capture device start and stop within the ship's historical time.

[0068] The carbon capture device status acquisition module extracts features from the carbon capture device operating status information in the historical trigger data to obtain multiple status feature groups, extracts historical operating parameter sets and multiple historical start and stop operation record sets under the multiple status feature groups, and processes them to obtain multiple historical trigger parameter sets.

[0069] The historical trigger data clustering module performs cluster screening on multiple historical trigger parameter sets respectively to obtain multiple filtered trigger parameter groups, and arranges them in chronological order to obtain multiple historical trigger parameter sequences.

[0070] The trigger trend prediction module performs trigger trend prediction based on multiple historical trigger parameter sequences to obtain the trigger response rate and delay rate.

[0071] The dynamic interlock correction module matches the target electrical characteristic parameters with multiple state characteristic groups to obtain a matching state characteristic group, and corrects the trigger response rate and delay rate according to the deviation between the target electrical characteristic parameters and the standard matching characteristic parameters of the matching state characteristic group to obtain a corrected response rate and a corrected delay rate.

[0072] The start-stop decision generation module makes start-stop plan decisions based on the corrected response rate and the corrected delay rate, obtains start-stop control instructions, and executes automatic start-stop operations of the carbon capture device.

[0073] Example 1: In the specific implementation of the shore power status monitoring module, obtaining the current ship's shore power connection status information, collecting target electrical characteristic parameters, and indexing historical trigger data are the primary steps in the entire intelligent interlocking system to automatically start and stop the carbon capture device. The core function of this module is to obtain key electrical parameters during shore power connection in real time and retrieve related historical trigger data from the ship's historical operation records.

[0074] Specifically, when a vessel connects to shore power, the shore power status monitoring module collects the current shore power voltage, current, and access timestamp in real time. These three parameters constitute the target electrical characteristic parameters, which intuitively and accurately reflect the current electrical status of the shore power connection. The voltage and current values ​​directly reflect the power supply characteristics of the shore power. Different voltage levels and current ranges will have different impacts on the operation of the ship's equipment, especially the start and stop logic of the carbon capture device, which may have different triggering conditions. The access timestamp records the specific moment of shore power connection.

[0075] After obtaining these target electrical characteristic parameters, the module needs to index the historical operation records of the ship based on the shore power access status information. The shore power access status information here mainly includes key parameters such as the voltage and current values ​​of the current shore power access. The indexing process is not a simple full search, but a precise match based on the voltage and current values. Specifically, the system will search for shore power access events with the same voltage level and current range in the historical operation records based on the voltage and current values ​​of the current shore power access. For example, if the voltage of the current shore power access is 380V and the current is in the range of 200-250A, the system will search the historical records for shore power access events with a voltage level of around 380V and a current in a similar range.

[0076] After matching eligible historical shore power connection events, the system extracts key data related to the carbon capture device. Specifically, it extracts the start and stop timestamps and operating status parameters of the carbon capture device from these matching events. The start and stop timestamps record the specific time points when the carbon capture device started and stopped during the historical shore power connection process. The operating status parameters include various key indicators of the carbon capture device during historical operation, such as operating power and operating temperature. These parameters can reflect the operating status of the carbon capture device under different shore power connection conditions.

[0077] By matching the target electrical characteristic parameters of the current shore power connection with relevant events in historical operation records and extracting the corresponding carbon capture device data, the shore power status monitoring module provides accurate and historically valuable data for subsequent carbon capture device status analysis, historical data clustering, trigger trend prediction, and other modules. This indexing method based on actual operating data ensures that the system can fully utilize the ship's previous operating experience, improving the accuracy and reliability of carbon capture device startup and shutdown decisions under the current shore power connection.

[0078] Throughout the entire process, data collection and indexing must be highly real-time and accurate. Real-time ensures the system can respond promptly to changes in shore power access, avoiding decision-making delays caused by data collection delays. Accuracy ensures the reliability of the underlying data for subsequent analysis and processing. Errors in collected voltage and current values, or mismatches in indexed historical data, will affect the overall system's decision-making. Therefore, the design and implementation of this module requires the use of high-precision sensors to collect electrical parameters and the establishment of an efficient and accurate historical data indexing mechanism.

[0079] Furthermore, the module also needed to be highly compatible and scalable. As a ship's operating time increases, historical operational records accumulate. The system needed to efficiently process this massive amount of historical data, ensuring that indexing speeds did not significantly decrease due to the increase in data volume. Furthermore, when a ship upgrades to a different specification of shore power equipment or carbon capture device, the module needed to be able to adapt to the new electrical parameter ranges and operating status parameters through appropriate parameter configuration and algorithm adjustments.

[0080] Example 2: In the specific implementation of the carbon capture device status acquisition module, feature extraction and processing of the carbon capture device operating status information within historical trigger data to obtain multiple historical trigger parameter sets is a key step in achieving accurate decision-making for the intelligent interlocking system. Through deep mining and feature analysis of historical data, this module transforms complex operating status information into standardized parameter sets that can be used for clustering and prediction, providing structured data support for subsequent modules such as historical trigger data clustering and trigger trend prediction.

[0081] Specifically, after obtaining the historical trigger data of shore power access and carbon capture device start and stop from the shore power status monitoring module, the carbon capture device status acquisition module first processes the operating status information of multiple carbon capture devices in these historical trigger data. Taking into account the possible dimensional differences and numerical ranges of different historical data, in order to ensure the accuracy and consistency of subsequent feature extraction, these operating status information need to be normalized. Normalization can convert parameters of different dimensions into a unified numerical range and eliminate the impact of dimensions on data analysis. Based on normalization, the module extracts power fluctuation coefficient and temperature stability as key feature dimensions. The power fluctuation coefficient reflects the amplitude of the power output of the carbon capture device during operation, which can reflect the stability of the device operation; temperature stability characterizes the device's ability to control temperature during operation, and is an important indicator affecting the performance and life of the device.

[0082] After extracting the power fluctuation coefficient and temperature stability, the module uses a density clustering algorithm to cluster these two feature dimensions. The core concept of the density clustering algorithm is to divide clusters based on the density distribution of data points. It can automatically identify dense areas in the dataset as cluster centers. It is suitable for discovering clusters of arbitrary shapes and is insensitive to noisy data. Using this algorithm, the module can group the operating status information of carbon capture devices with similar power fluctuation coefficients and temperature stability into a single category, thereby obtaining multiple state feature groups. Each state feature group represents the typical state of the carbon capture device under specific operating conditions.

[0083] After obtaining multiple state feature groups, the module needs to further extract the historical operating parameter set and historical start-stop operation record set under each state feature group. Specifically, the standard operating power and temperature threshold of the carbon capture device in the historical trigger data under multiple state feature groups are extracted to form a historical operating parameter set. The standard operating power reflects the typical working power of the carbon capture device under different state feature groups, and the temperature threshold limits the temperature range for safe operation of the device. At the same time, multiple historical voltage threshold sets and multiple historical access time sets for triggering the start-stop operation of the ship under multiple state feature groups are extracted. The historical voltage threshold set records the critical shore power voltage that triggers the start and stop of the carbon capture device under different state feature groups; the historical access time set represents the correlation between the duration of shore power access and the start and stop of the carbon capture device.

[0084] The module categorizes data based on the degree of match between multiple historical voltage threshold sets and historical operating parameter sets. This match is calculated based on a correlation analysis between historical voltage thresholds and standard operating power and temperature thresholds. By assessing the impact of voltage thresholds on the carbon capture device's operating status, the module categorizes historical data into multiple historical basic parameter sets. Each historical basic parameter set corresponds to a group of historical data with similar voltage trigger conditions and operating status characteristics.

[0085] The module calculates and modifies multiple historical basic parameter sets based on the ratio of a preset duration threshold to multiple historical access duration sets. The preset duration threshold is a reference duration set based on the design characteristics of the carbon capture device and actual operating experience. By calculating the ratio of the historical access duration to the preset duration, it is possible to assess the impact of the shore power access duration on the start and stop triggering of the carbon capture device. Modifying the historical basic parameter set based on this ratio can further optimize the accuracy of the parameter set, making it more consistent with the needs of actual operating scenarios, thereby obtaining multiple historical trigger parameter sets. These historical trigger parameter sets integrate multi-dimensional information such as the operating status of the carbon capture device, the shore power voltage threshold, and the access duration.

[0086] Throughout the entire processing process, data normalization and feature extraction must ensure accuracy and rationality. The choice of normalization method should be optimized based on the characteristics of the carbon capture device's operating state parameters to maximize the preservation of the data's original characteristics. The extraction of power fluctuation coefficient and temperature stability requires a combination of the device's operating principle and actual operating data to ensure that the selected features effectively characterize the device's operating state. The parameter settings of the density clustering algorithm also need to be adjusted based on the distribution characteristics of historical data to achieve a reasonable division of state feature groups and avoid data classification bias caused by overly fine or coarse clustering.

[0087] Furthermore, the setting of the preset duration threshold requires comprehensive consideration of common shore power access scenarios and the response characteristics of the carbon capture device. If the preset duration threshold is set too high, valid data from short shore power access scenarios may be overlooked; if it is set too low, the correction calculation may become overly sensitive, affecting the stability of the historical trigger parameter set. Therefore, in practical applications, it is necessary to analyze a large amount of historical data and summarize actual operating experience to reasonably determine the preset duration threshold to ensure the effectiveness and reliability of the correction calculation.

[0088] Example 3: In the specific implementation of the historical trigger data clustering module, clustering and filtering multiple historical trigger parameter sets to obtain multiple historical trigger parameter sequences is a key step in enabling the intelligent interlocking system to effectively utilize historical data. By weighting, clustering, filtering, and chronologically sorting historical trigger parameters, this module transforms dispersed historical data into an ordered parameter sequence. This provides structured time series data for the trigger trend prediction module, thereby improving the system's accuracy in predicting the start and stop trigger trends of the carbon capture device.

[0089] After acquiring multiple historical trigger parameter sets from the carbon capture device status acquisition module, the historical trigger data clustering module processes each set individually. Taking the first historical trigger parameter set as an example, the module first selects a first baseline trigger parameter within that set. This first baseline trigger parameter can be selected in a variety of ways, such as selecting the trigger parameter with the latest timestamp in the set, or using statistical methods to select representative trigger parameters, such as those corresponding to the mean or median. Its core purpose is to provide a baseline reference point for subsequent weight assignment and cluster screening.

[0090] After determining the first baseline trigger parameter, the module assigns a weight coefficient to each trigger parameter based on the deviation between the other trigger parameters in the first historical trigger parameter set and the first baseline trigger parameter, thereby obtaining a first basic weight distribution. The deviation here is usually measured by calculating the absolute value or sum of squares of the difference between the trigger parameter and the baseline trigger parameter in each dimension. The size of the deviation is negatively correlated with the size of the weight coefficient, that is, the smaller the deviation between the trigger parameter and the baseline trigger parameter, the larger its weight coefficient, and vice versa. This weight distribution method reflects that the historical trigger parameters that are closer to the baseline parameters have higher reference value and can be given priority in subsequent cluster screening.

[0091] After obtaining the first basic weight distribution, the module begins cluster screening. A preset number of trigger parameters are randomly selected from the first historical trigger parameter set to form a first filtered trigger parameter group. The setting of the preset number of trigger parameters requires a comprehensive consideration of the size of the historical trigger parameter set and computational efficiency. It is usually set to a reasonable proportion of the number of trigger parameters in the set, such as 10%-30%, to ensure that the filtered parameter group is representative while avoiding excessive computational effort.

[0092] The module assigns weight coefficients to each trigger parameter based on the deviation between the trigger parameters in the first screening trigger parameter group and the first baseline trigger parameter, obtaining a first screening weight distribution. The module then calculates the degree of match between the first screening weight distribution and the first base weight distribution, using this as the first screening match value. The degree of match can be calculated using a variety of methods, such as cosine similarity or Euclidean distance, to measure the degree of consistency between the screening weight distribution and the base weight distribution. The higher the degree of match, the more the screened trigger parameter group meets the requirements of the base weight distribution.

[0093] After completing the processing of the first screening trigger parameter group, the module again randomly selects a preset number of screening trigger parameters in the first historical trigger parameter set to form a second screening trigger parameter group, and processes according to the same steps to obtain a second screening matching value. The clustering screening operation is repeated in this way until the convergence condition is met. The convergence condition can be that the change in the matching value after multiple consecutive screenings is less than a preset threshold, or reaches a preset maximum number of screening times. At this time, the screening trigger parameter group with the largest screening matching value is output as the first screening trigger parameter group. This clustering screening method based on random selection and iterative optimization can find the trigger parameter group that best matches the basic weight distribution in a larger search space, thereby improving the reliability of the clustering results.

[0094] After obtaining the first filtered trigger parameter set, the module sorts the trigger parameters within that set according to their timestamp information, thereby obtaining the first historical trigger parameter sequence. Timestamp sorting ensures the temporal order of the parameter sequence and reflects the patterns and trends of changes in the historical trigger parameters over time. For the remaining multiple historical trigger parameter sets, the module performs clustering screening and time sorting using the same methods used for the first historical trigger parameter set, ultimately obtaining multiple historical trigger parameter sequences. These historical trigger parameter sequences are arranged in chronological order, integrating trigger parameters from different historical periods and operating states, providing rich time series data input for the trigger trend prediction module.

[0095] Throughout the cluster screening and sorting process, the selection of baseline trigger parameters and the weighting mechanism are key factors influencing clustering results. The baseline trigger parameters should accurately represent the typical characteristics of the historical trigger parameter set. Failure to do so may lead to biased weighting, which in turn affects the accuracy of cluster screening. When assigning weights, the negative correlation between bias and weight coefficients must be appropriately set to ensure that trigger parameters with higher reference value receive greater weight.

[0096] The preset number of filters and the convergence criteria also need to be optimized based on the actual data. A too small preset number of filters may result in a less representative set of parameters, while a too large one will increase the computational effort. Overly loose convergence criteria may lead to inaccurate clustering results, while overly strict ones will prolong computation time. Therefore, in practical applications, it is necessary to determine the appropriate preset number of filters and convergence criteria through analysis of historical data and multiple experiments to strike a balance between computational efficiency and clustering accuracy.

[0097] This module needs to be able to process large amounts of historical data. As a ship's operating time increases, the number and size of historical trigger parameter sets may continue to expand. The module needs to use efficient algorithms and data structures to ensure the efficiency of clustering, screening, and sorting operations, and avoid the system response speed being reduced due to the increase in data volume.

[0098] Example 4: In the specific implementation of the trigger trend prediction module and the dynamic interlock correction module, the two work together to predict the trigger trend based on the historical trigger parameter sequence and make corrections based on the current electrical parameters. This is the core link for achieving precise control of the automatic start and stop of the carbon capture device. The following is a detailed description of the workflow with a specific example:

[0099] Suppose a container ship is connected to shore power at a port. The shore power status monitoring module records the current connection voltage as 440V and the current as 300A, with the connection timestamp as 10:00:00 on June 15, 2025. At this point, the trigger trend prediction module needs to predict the trigger response rate and delay rate under these operating conditions based on historical trigger parameter sequences. The module collects sample trigger data from the operating data of multiple similar ships. For example, a set of sample trigger parameter sequences from 100 ships under different shore power connection scenarios is collected. Each sequence contains data on the time-varying changes in parameters such as voltage, current, and connection duration. Furthermore, based on the parameter change indicators within each sample trigger parameter sequence, such as the time at which the carbon capture device starts after shore power connection, a set of sample response rates (for example, if the device starts 10 minutes after shore power connection, the response rate corresponds to a specific calculated value) and a set of sample delay rates (the ratio of the delay time to the standard time) are calculated.

[0100] The module uses a set of sample trigger parameter sequences as prediction inputs and a set of sample response rates and sample delay rates as prediction outputs to construct a trigger trend predictor. This predictor can employ machine learning models such as recurrent neural networks (RNNs) or long short-term memory networks (LSTMs). Through training, the model learns the temporal relationship between shore power parameter changes and the carbon capture device's startup and shutdown responses. For example, after learning from a large amount of historical data, the model can identify that when the voltage is between 400-450V and the current exceeds 250A, the device has a high startup response rate and a low delay rate.

[0101] Based on the trained trigger trend predictor, the module processes multiple historical trigger parameter sequences for the current ship. For example, assume the ship has five historical trigger parameter sequences similar to the current operating conditions of 440V voltage and 300A: 430V / 280A in January 2024, 445V / 310A in May 2024, and so on. The predictor classifies these sequences by trigger trend, analyzing the voltage and current variation patterns within each sequence to obtain the corresponding characteristic response rate and delay rate. For example, under the 430V / 280A operating condition, the characteristic response rate is 0.85 (indicating an 85% probability of triggering within a specific timeframe), and the delay rate is 0.2 (indicating a delay time of 20% of the standard timeframe). Under the 445V / 310A operating condition, the characteristic response rate is 0.92, and the delay rate is 0.15.

[0102] The module analyzes the similarity between the current target electrical characteristic parameters (440V / 300A) and multiple state feature groups output by the carbon capture device state acquisition module. Assume that Group A, corresponding to the voltage 400-450V and current 250-350A operating conditions, has the smallest Euclidean distance between its power fluctuation coefficient and temperature stability characteristics and the current parameters. The module then weights the response rates and delay rates of multiple features based on their similarities (e.g., 90% similarity for voltage and 85% similarity for current). For example, the response rates for the 430V / 280A and 445V / 310A operating conditions are weighted by similarity weights (e.g., 0.4 and 0.6), resulting in an initial trigger response rate of 0.85 × 0.4 + 0.92 × 0.6 = 0.892, and a delay rate of 0.2 × 0.4 + 0.15 × 0.6 = 0.17.

[0103] Enter the dynamic interlock correction module. The module matches the current target electrical characteristic parameters with multiple state characteristic groups, selects Group A with the highest similarity as the matching state characteristic group, and obtains the standard matching characteristic parameters for this group, such as standard voltage of 440V, standard current of 300A, standard power fluctuation coefficient of 0.3, and standard temperature stability of 0.9. The module calculates the deviation between the current parameters and the standard parameters: voltage deviation of 0 (matching 440V) and current deviation of 0 (matching 300A). However, assume that the actual collected power fluctuation coefficient value is 0.35 (standard 0.3) and temperature stability is 0.85 (standard 0.9). Interlock correction factors are set based on these deviations. For example, for every 0.01 deviation in the power fluctuation coefficient, a correction factor of -0.02 is applied, and for every 0.01 deviation in the temperature stability, a correction factor of -0.01 is applied. The total correction factor is (0.35 - 0.3) × (-2) + (0.85 - 0.9) × (-1) = -0.1 - 0.05 = -0.15.

[0104] This correction factor is used to calculate the trigger response rate and delay rate: Corrected response rate = 0.892 + (-0.15) × 0.892 × 0.1 ≈ 0.878 (rounded to three decimal places), Corrected delay rate = 0.17 + (-0.15) × 0.17 × 0.2 ≈ 0.165. This correction process takes into account the difference between the current operating parameters and the standard state, making the prediction results more consistent with actual operating conditions.

[0105] For example, if a ship is connected to shore power at 380V and 200A, the historical data shows a characteristic response rate of 0.7 and a delay rate of 0.3 for similar operating conditions. The standard voltage for the matching state characteristic group is 380V, but the standard current is 220A. The current current deviation is -20A, corresponding to a correction factor of +0.08 (because the current below the standard may cause the device to start faster). This increases the corrected response rate to 0.7 + 0.08 × 0.7 × 0.1 ≈ 0.706, and reduces the corrected delay rate to 0.3 - 0.08 × 0.3 × 0.2 ≈ 0.295.

[0106] Throughout the entire process, the diversity of sample data and the quality of predictor training are crucial. Insufficient sample data coverage (e.g., a lack of low-voltage, high-current scenarios) can lead to prediction bias. During dynamic corrections, the mapping relationship between deviations and correction coefficients must be statistically derived from a large amount of historical data. For example, by analyzing the correlation between more than 1,000 sets of parameter deviations and actual response changes, reasonable correction coefficient calculation rules can be determined to ensure that the correction logic conforms to actual operating rules. At the same time, the system must regularly update sample data and correction rules to adapt to changes such as aging ship equipment and upgrades to carbon capture devices to maintain the accuracy of predictions and corrections.

[0107] Example 5: In the specific implementation of the start-stop decision generation module, this module constructs a sample data set, trains a decision model, and generates start-stop control instructions based on the corrected response rate and delay rate. This is the final execution step for automatically starting and stopping the carbon capture device. The following details its workflow with specific examples:

[0108] Suppose a bulk carrier connects to shore power at a port. The dynamic interlocking correction module outputs a corrected response rate of 0.91 and a corrected delay rate of 0.18. The start-stop decision generation module then generates specific start-stop control instructions based on these two parameters. The module collects a set of sample corrected response rates and sample corrected delay rates and sets the sample start-stop control instructions accordingly. For example, consider 1,000 sets of samples from historical data under different operating conditions. One set of samples has a corrected response rate of 0.85 and a corrected delay rate of 0.2, corresponding to a carbon capture device starting 15 minutes after shore power connection. This start-up time is then used as part of the sample start-stop control instructions. Another set of samples has a corrected response rate of 0.95 and a corrected delay rate of 0.1, corresponding to a device starting 10 minutes after shore power connection. Each sample start-stop control instruction includes a start-stop time node. The magnitude of the sample corrected response rate and corrected delay rate is negatively correlated with the lead time of the start-stop time node. That is, a higher response rate and a lower delay rate result in an earlier start-stop time node.

[0109] When constructing a sample data set, it is necessary to ensure that the samples cover different operating conditions such as voltage, current, and connection time. For example, the sample includes a voltage of 380V, a current of 200A, a connection time of 30 minutes, a corrected response rate of 0.7, a delay rate of 0.25, and a device startup time of 20 minutes; a voltage of 440V, a current of 300A, a connection time of 60 minutes, a corrected response rate of 0.9, a delay rate of 0.15, and a device startup time of 10 minutes, etc. Through the accumulation of a large number of samples, a sample corrected response rate set, a sample corrected delay rate set, and a sample start-stop instruction set are formed. The start-stop time node of each sample must be determined based on the start-stop timestamp of the carbon capture device in the actual historical operation records to ensure the authenticity and reliability of the data.

[0110] The module uses a set of sample corrected response rates and a set of sample corrected delay rates as decision inputs, and a set of sample start-stop instructions as decision outputs, to construct a start-stop plan decision maker. This decision maker can use a regression model or a decision tree model in machine learning. Through training, the model learns the mapping relationship between the corrected response rate, the corrected delay rate, and the start-stop time nodes. For example, using the random forest algorithm, with an input of a corrected response rate of 0.85 and a delay rate of 0.2, the model outputs a corresponding start time node of 15 minutes; with an input of a corrected response rate of 0.95 and a delay rate of 0.1, the output starts at a time node of 10 minutes. During the training process, the model parameters are adjusted to minimize the error between the start-stop time nodes predicted by the model and the actual time nodes in the historical samples.

[0111] Based on the trained start / stop plan decider, the module processes the current ship's corrected response rate of 0.91 and corrected delay rate of 0.18. Using the learned mapping, the decider calculates the corresponding start / stop time nodes. Assume that in the training data, samples with a corrected response rate of approximately 0.9 and a corrected delay rate between 0.15 and 0.2 have start / stop time nodes concentrated between 12 and 14 minutes. The decider, combining the current parameter values ​​with the model, calculates the start / stop time node to be 13 minutes. This generates the start / stop control command: activate the carbon capture device 13 minutes after shore power is connected.

[0112] For example, when a container ship is connected to shore power, the corrected response rate is 0.75 and the corrected delay rate is 0.22. Based on sample data, when the corrected response rate is 0.7 and the delay rate is 0.25, the device startup time is 20 minutes, while when the corrected response rate is 0.8 and the delay rate is 0.2, the startup time is 18 minutes. The decision maker, through interpolation or model calculation, determines that the startup time corresponding to the current parameters is 19 minutes, indicating that the control instruction is to start the device 19 minutes after the shore power is connected.

[0113] After generating the start / stop control command, the module sends it to the carbon capture device's control system, executing the automatic start / stop operation. The control system verifies the command, checking for a stable shore power connection. If the shore power connection is disconnected before the start time, the command is suspended. If the connection is normal, the device is started as instructed.

[0114] Throughout the entire process, the quality and quantity of sample data directly impact the accuracy of the decision maker. If the sample lacks data for a specific operating condition (such as low voltage and high current), the decision maker may err when handling that condition. Therefore, the system needs to continuously accumulate new sample data and regularly update the decision maker's training model to adapt to a wider range of operating scenarios. For example, if a ship installs a new carbon capture device, its start-up and shutdown logic may change. In this case, operating data from the new device must be collected and the decision maker retrained to ensure the accuracy of the generated instructions.

[0115] The negative correlation between start-up and shutdown time nodes and the modified response rate and modified delay rate must be established based on actual operating patterns. For example, analysis of extensive historical data revealed that for every 0.1 increase in the modified response rate, the device startup time is shortened by an average of 2 minutes; for every 0.05 decrease in the modified delay rate, the startup time is shortened by an average of 1 minute. This relationship must be determined based on statistical analysis, rather than subjective assumptions, to ensure that the decision logic aligns with the actual response characteristics of the carbon capture device.

[0116] The module needs to have a fault-tolerance mechanism. When the input correction response rate or correction delay rate exceeds the range of historical samples, it can trigger an early warning and adopt a default control strategy to avoid erroneous instructions caused by data anomalies. For example, if the correction response rate reaches 1.0 (the theoretical maximum), the module can refer to the start time node corresponding to the operating condition with the highest response rate in the sample and fine-tune it based on expert experience to generate reasonable control instructions.

[0117] The start-stop decision generation module constructs a sample data set and trains a decision model, translating the corrected response rate and delay rate into specific start-stop time nodes, enabling precise control of the carbon capture device's automatic start and stop. This process relies closely on historical operating data and model algorithms. Through continuous optimization and updates, the system ensures it can adapt to start and stop requirements under different operating conditions, enhancing the intelligence and low-carbonization of ship operations.

[0118] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0119] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power, characterized in that: The system comprises: A shore power status monitoring module, which is used to obtain the current shore power access status information of the ship and collect target electrical characteristic parameters during the shore power access process. Based on the shore power access status information, the module indexes the historical triggering data of shore power access and carbon capture device start and stop within the ship's history. a carbon capture device status collection module, the carbon capture device status collection module being configured to extract features of the carbon capture device operating status information in the historical trigger data to obtain a plurality of status feature groups, extract a set of historical operating parameters and a plurality of sets of historical start and stop operation records under the plurality of status feature groups, and process the extracted data to obtain a plurality of sets of historical trigger parameters; A historical trigger data clustering module, the historical trigger data clustering module is used to cluster and filter the multiple historical trigger parameter sets respectively to obtain multiple filtered trigger parameter groups, and arrange them in chronological order to obtain multiple historical trigger parameter sequences; a trigger trend prediction module, configured to perform trigger trend prediction based on the plurality of historical trigger parameter sequences to obtain a trigger response rate and a delay rate; a dynamic interlock correction module, configured to match the target electrical characteristic parameters with the plurality of state characteristic groups to obtain a matching state characteristic group, and to correct the trigger response rate and delay rate based on a deviation between the target electrical characteristic parameters and a standard matching characteristic parameter of the matching state characteristic group to obtain a corrected response rate and a corrected delay rate; A start-stop decision generation module is used to make a start-stop plan decision based on the corrected response rate and the corrected delay rate, obtain a start-stop control instruction, and execute automatic start-stop operations of the carbon capture device.

2. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 1 is characterized in that: Obtain the current shore power access status information of the ship, and collect the target electrical characteristic parameters during the shore power access process. Based on the shore power access status information, index the historical triggering data of shore power access and carbon capture device start and stop within the ship's historical time, including: Obtain the current voltage, current, and access timestamp of the ship's shore power connection as target electrical characteristic parameters; According to the shore power access status information, an index is performed in the historical operation record of the ship to obtain historical triggering data of shore power access and start and stop of the carbon capture device.

3. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 1 is characterized in that: Feature extraction is performed on the carbon capture device operating status information in the historical trigger data to obtain multiple status feature groups, historical operating parameter sets and multiple historical start-stop operation record sets under the multiple status feature groups are extracted, and multiple historical trigger parameter sets are obtained through processing, including: Acquiring operation status information of multiple carbon capture devices in the historical trigger data, performing feature extraction processing, and obtaining multiple state feature groups; Extracting standard operating power and temperature thresholds of the carbon capture device from the historical trigger data under the multiple state feature groups to obtain a historical operating parameter set, and extracting multiple historical voltage threshold sets and multiple historical connection duration sets for triggering start and stop operations of the ship under the multiple state feature groups; Classify and obtain multiple historical basic parameter sets according to the matching degree between the multiple historical voltage threshold sets and the historical operating parameter sets; According to the ratio of the preset duration threshold and the multiple historical access duration sets, the multiple historical basic parameter sets are corrected and calculated to obtain multiple historical trigger parameter sets.

4. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 1 is characterized in that: Clustering and screening the multiple historical trigger parameter sets respectively to obtain multiple filtered trigger parameter groups, and arranging them in chronological order to obtain multiple historical trigger parameter sequences, including: Selecting and obtaining a first benchmark trigger parameter from a first historical trigger parameter set in the plurality of historical trigger parameter sets; Allocating weight coefficients according to the deviations between the other trigger parameters in the first historical trigger parameter set and the first benchmark trigger parameter to obtain a first basic weight distribution, wherein the deviations are negatively correlated with the weight coefficients; performing cluster screening on the first historical trigger parameter set according to the first basic weight distribution to obtain a first screening trigger parameter group; Sort the multiple trigger parameters in the first screening trigger parameter group according to their timestamp information to obtain a first historical trigger parameter sequence; Cluster screening and time sorting are performed on other multiple historical trigger parameter sets to obtain multiple historical trigger parameter sequences.

5. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 4 is characterized in that: Clustering and screening the first historical trigger parameter set according to the first basic weight distribution to obtain a first screening trigger parameter group includes: Randomly selecting a preset number of screening trigger parameters from the first historical trigger parameter set to obtain a first screening trigger parameter group; Allocating weight coefficients according to deviations between the trigger parameters in the first screening trigger parameter group and the first benchmark trigger parameter to obtain a first screening weight distribution; Calculating a matching degree between the first screening weight distribution and the first basic weight distribution as a first screening matching value; Randomly selecting a preset number of screening trigger parameters from the first historical trigger parameter set again to obtain a second screening trigger parameter group, and processing to obtain a second screening matching value; The cluster screening is continued until convergence, and the screening trigger parameter group with the largest screening matching value is output as the first screening trigger parameter group.

6. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 1 is characterized in that: Performing trigger trend prediction based on the multiple historical trigger parameter sequences to obtain a trigger response rate and a delay rate includes: According to the sample trigger data of multiple ships, a sample trigger parameter sequence set is collected, and a sample response rate set and a sample delay rate set are obtained according to the parameter change identifier in each sample trigger parameter sequence; Using the sample trigger parameter sequence set as prediction input, and using the sample response rate set and the sample delay rate set as prediction output, to construct a trigger trend predictor; Based on the trigger trend predictor, the trigger trend of the plurality of historical trigger parameter sequences is classified to obtain a plurality of characteristic response rates and a plurality of characteristic delay rates; The similarities between the target electrical characteristic parameters and the multiple state characteristic groups are analyzed, and the multiple characteristic response rates and multiple characteristic delay rates are weightedly calculated according to the magnitude of the multiple characteristic similarities to obtain a trigger response rate and a delay rate.

7. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 6 is characterized in that: Matching the target electrical characteristic parameter with the multiple state characteristic groups to obtain a matching state characteristic group, and correcting the trigger response rate and the delay rate according to a deviation between the target electrical characteristic parameter and a standard matching characteristic parameter of the matching state characteristic group to obtain a corrected response rate and a corrected delay rate, including: Selecting the state feature group with the greatest similarity as the matching state feature group, and obtaining the standard matching feature parameters of the matching state feature group; setting an interlock correction coefficient according to a deviation between the target electrical characteristic parameter and a standard matching characteristic parameter of the matching state characteristic group; The interlock correction coefficient is used to perform correction calculation on the trigger response rate and delay rate to obtain a corrected response rate and a corrected delay rate.

8. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 1 is characterized in that: According to the corrected response rate and the corrected delay rate, a start / stop plan decision is made, a start / stop control instruction is obtained, and an automatic start / stop operation of the carbon capture device is executed, including: Collecting a set of sample corrected response rates and a set of sample corrected delay rates, and setting sample start / stop control instructions based on the magnitude of each sample corrected response rate and sample corrected delay rate to obtain a set of sample start / stop instructions, wherein each sample start / stop control instruction includes a start / stop time node, and the magnitude of the sample corrected response rate and the sample corrected delay rate is negatively correlated with the advance amount of the start / stop time node; The sample corrected response rate set and the sample corrected delay rate set are used as decision inputs, and the sample start-stop instruction set is used as decision output to construct a start-stop plan decision maker; The start-stop plan decision maker is used to make a start-stop plan decision on the corrected response rate and the corrected delay rate to obtain a start-stop control instruction.

9. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 2 is characterized in that: Indexing the historical operation records of the vessel to obtain historical triggering data for shore power access and carbon capture device startup and shutdown, including: According to the voltage and current values ​​of the shore power connection, matching shore power connection events with the same voltage level and current range in the historical operation records; The start and stop timestamps and operating status parameters of the carbon capture device in the matching events are extracted as the historical trigger data of shore power access and the start and stop of the carbon capture device.

10. The intelligent interlocking system for automatically starting and stopping a ship's carbon capture device when connected to shore power according to claim 3 is characterized in that: Perform feature extraction processing to obtain multiple state feature groups, including: Normalizing the carbon capture device operating status information in the historical trigger data, and extracting power fluctuation coefficient and temperature stability as feature dimensions; A density clustering algorithm is used to cluster the feature dimensions to obtain multiple state feature groups.

Images