Ship safety monitoring intelligent abnormity alarm system based on multi-dimensional analysis

Abstract

The invention discloses a ship safety monitoring intelligent abnormity alarm system based on multi-dimensional analysis, and relates to the technical field of shipping safety monitoring, the ship safety monitoring intelligent abnormity alarm system comprises a safety monitoring center, the safety monitoring center is in communication connection with the following modules: a data acquisition and analysis module used for collecting ship operation data from each sensor and equipment of a ship, and sending the collected ship operation data to an alarm module; feature analysis is carried out, alarm features reflecting ship abnormity are extracted, and an alarm feature sequence table is obtained through integration. According to the method, the ship operation data is collected and preprocessed from each sensor and equipment of the ship, and the abnormal condition in the ship operation can be accurately identified by combining the abnormal identification model and the alarm feature sequence table constructed based on the neural network model, so that the limitation possibly brought by a single data source is avoided; the accuracy of anomaly detection is improved, potential safety hazards can be found in time, and accidents are prevented.

Description

Technical Field

[0001] The present invention relates to the technical field of shipping safety monitoring, and in particular to an intelligent abnormality alarm system for ship safety monitoring based on multi-dimensional analysis. Background Art

[0002] The shipping industry faces a complex and ever-changing marine environment, such as severe weather, sea conditions, and waterway conditions. These factors may lead to safety accidents such as ship collisions, groundings, fires, and explosions. Once an accident occurs, it will not only cause casualties and property losses, but may also cause serious pollution to the marine environment. Therefore, a system is needed that can monitor the status of ships in real time, detect abnormal situations in a timely manner, and issue alarms to prevent accidents.

[0003] In the existing technology, ships may have multiple minor faults or abnormalities, which in turn frequently trigger alarms, causing alarm fatigue for relevant personnel, resulting in the neglect of important abnormal situations. At the same time, it is difficult to accurately assess the severity of the abnormalities, resulting in confusion in alarm priorities and affecting the efficiency of emergency response. Therefore, how to accurately assess the severity of the abnormalities, and construct a ship safety knowledge graph, associate the abnormalities with related processes, and determine the alarm direction through graph reasoning is the problem to be solved by the present invention. To this end, a ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis is proposed. Summary of the Invention

[0004] The purpose of the present invention is to provide a ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis to solve the problems raised in the above background technology.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis includes a safety monitoring center, which is connected to the following modules: The data acquisition and analysis module is used to collect ship operation data from various sensors and equipment on the ship, perform feature analysis, extract alarm features reflecting ship abnormalities, and integrate them to obtain an alarm feature sequence list; The ship anomaly detection module is used to identify abnormal conditions in ship operation by combining the anomaly recognition model built based on the neural network model and the alarm feature sequence table; The abnormality severity assessment module is used to analyze the detected abnormalities through multi-dimensional analysis, calculate the severity assessment index, distinguish the severity of different abnormalities, obtain the detected abnormality information from the ship abnormality detection module, analyze the degree of deviation of the abnormal alarm characteristics from their baseline values, and at the same time, collect the time of occurrence of the abnormality and the location of the ship. The collected multi-dimensional data is analyzed and, based on the actual needs and safety standards of the ship operation, a corresponding weight is assigned to each abnormal alarm characteristic, and then the severity assessment index is calculated. Based on the calculated severity assessment index, the abnormal situation is divided into different severity levels, namely, minor level, general level, severe level and emergency level, and a clear threshold for severity level division is set; The knowledge graph construction reasoning module is used to construct a ship safety knowledge graph and use the knowledge graph to perform reasoning analysis on detected anomalies to determine the direction of reporting to the police and the recommended handling measures; The alarm feedback module is used to combine the abnormal severity assessment results and the reasoning results of the knowledge graph to generate corresponding alarm information, and notify the alarm information to relevant personnel through various means, while collecting feedback from relevant personnel on the alarm information.

[0006] A further improvement of the technical solution of the present invention is that: the data acquisition and analysis module includes a multi-source data acquisition unit and a feature analysis unit; The multi-source data acquisition unit is used to collect and pre-process ship operation data from various sensors and equipment on the ship, including the ship's navigation status, equipment operation status, environmental data and crew operating behavior; The feature analysis unit is used to perform feature analysis on the pre-processed ship operation data, extract alarm features, and determine the baseline value of each alarm feature based on the ship alarm requirements, and then integrate them to obtain an alarm feature sequence table.

[0007] A further improvement of the technical solution of the present invention is that the multi-source data acquisition unit specifically includes: Establish connections with various sensors and equipment on board the vessel through multiple interface protocols (CAN bus, RS485, Ethernet, etc.) to collect ship operation data including the vessel's navigation status, equipment operating status, environmental data, and crew operating behavior; The collected ship operation data is preprocessed, including data cleaning and formatting. At the same time, the preprocessed ship operation data is integrated and organized according to a unified data structure to form a complete ship operation data set. During the integration process, the data from different sources are timestamp aligned to ensure the consistency of the data in the time dimension, and the integrated ship operation data set is then stored in the local database.

[0008] A further improvement of the technical solution of the present invention is that the feature analysis unit specifically includes: The pre-processed ship operation data is comprehensively scanned and analyzed to identify alarm features related to ship safety and abnormal status. By analyzing the statistical characteristics of the data and combining it with the ship's operating mechanism and safety standards, alarm features related to navigation status, equipment operation, environmental conditions, and crew operating behavior are screened out. Among them, the navigation status alarm features include abnormal speed, course deviation, and position deviation; the equipment operation status alarm features include abnormal engine speed, abnormal equipment temperature, and abnormal equipment pressure; the environmental data alarm features include abnormal meteorological conditions and abnormal sea conditions; and the crew operating behavior alarm features include abnormal operation response time, abnormal operation frequency, and illegal operation behavior; Based on the ship alarm requirements and the normal operation data of the ship, the extracted alarm features are analyzed one by one to determine the benchmark value. After determining the benchmark value of each alarm feature, all alarm features and their corresponding benchmark values ​​are integrated to form an alarm feature sequence table.

[0009] A further improvement of the technical solution of the present invention is that the ship anomaly detection module specifically includes: Receive real-time collected and pre-processed ship operation data, call a pre-built alarm feature sequence table, perform a preliminary match between the received ship operation data and the alarm feature sequence table, and extract data related to the alarm features, where the data related to the alarm features include the ship's navigation status, equipment operation status, environmental data, and ship operation data of crew operating behavior; Load the anomaly recognition model built based on the neural network, obtain the alarm feature data of the current ship operation from the alarm feature sequence table, and input it into the anomaly recognition model for inference. Based on the input alarm feature data, the multi-layer neural network is calculated to compare each alarm feature with its corresponding benchmark value to identify the alarm features that deviate from the benchmark value; According to the inference results of the anomaly recognition model, the ship's operating status is judged as abnormal. If the actual value of an alarm feature deviates from the baseline value by more than the set deviation threshold, it will be marked as an abnormal state.

[0010] A further improvement of the technical solution of the present invention is that: the knowledge graph construction and reasoning module includes a ship safety knowledge graph construction unit and a graph reasoning unit; The ship safety knowledge graph construction unit is used to construct a ship safety knowledge graph based on ship domain knowledge, historical fault cases, equipment maintenance manuals and other information. It structures entities such as ship equipment, fault types, emergency response processes and relevant personnel, and the relationships between them to form a knowledge network, thereby achieving systematic integration and visual display of ship safety knowledge. The graph reasoning unit is used to use knowledge graph reasoning technology to query and reason in the knowledge graph based on the currently detected abnormal information, and infer the processing direction and emergency response strategy of the current alarm.

[0011] A further improvement of the technical solution of the present invention is that the ship safety knowledge graph construction unit specifically includes: Collect various knowledge sources in the ship field, including ship design documents, equipment maintenance manuals, historical failure cases and safety regulations, covering the operating parameters, failure modes, emergency response procedures and responsibilities of relevant personnel of ship equipment. Then classify and organize the collected knowledge sources, extract key information including equipment name, failure type, treatment measures and related personnel, and use the organized key information as the basic data for constructing the knowledge graph; Based on the organized knowledge sources, various entities and their relationships within the field of ship safety are identified and defined. Entities include ship equipment, fault types, emergency response procedures, and relevant personnel. At the same time, the relationships between entities are clarified. Using natural language processing (NLP) technology, entities and relationships are extracted from text data and structured as nodes and edges in a knowledge graph. The identified entities and relationships are integrated into the ship safety knowledge graph to form a structured knowledge network, in which the knowledge graph represents entities with nodes and the relationships between entities with edges, and the knowledge graph is displayed through visualization tools.

[0012] A further improvement of the technical solution of the present invention is that the graph reasoning unit specifically includes: Receive the current anomaly information from the ship anomaly detection module, map the anomaly information with the entities and relationships in the ship safety knowledge graph, and determine the associated nodes of the anomaly information in the knowledge graph. Through mapping analysis, convert the actual anomaly problem into a query object in the knowledge graph; Based on the mapping results, deep queries are performed in the knowledge graph. Using the query language of the graph database, the relationships between nodes and edges are traversed to find all information related to the current anomaly. At the same time, based on the existing rules and patterns in the knowledge graph, potential risks and related issues are inferred. Based on the query and reasoning results, the graph reasoning unit conducts comprehensive analysis and infers the processing direction and emergency response strategy of the current alarm. The processing direction clearly defines the overall idea of ​​solving the problem, and the emergency response strategy specifies the specific operation steps, required resources, and responsibilities of relevant personnel.

[0013] A further improvement of the technical solution of the present invention is that the alarm feedback module specifically includes: Obtain the severity level obtained by the abnormal severity assessment module and the processing direction and emergency response strategy provided by the knowledge graph reasoning unit, and then generate corresponding alarm information, clearly indicating the abnormal situation, severity, recommended processing measures and emergency response steps, and notify the relevant personnel of the alarm information through various means; After notifying relevant personnel, the alarm feedback module collects feedback from relevant personnel on the alarm information, including whether the relevant personnel received the alarm in a timely manner, whether the alarm content is accurate and clear, and whether the emergency measures taken are effective. It also categorizes and organizes the collected feedback information and summarizes it according to feedback type, relevant personnel position, etc. Based on the collected feedback information, the alarm feedback module optimizes and adjusts the knowledge graph, anomaly recognition model and alarm strategy. If the feedback shows that the alarm content is inaccurate and unclear, the cause is analyzed and the relevant entities and relationships in the knowledge graph are corrected, and the parameters of the anomaly recognition model are adjusted. If the feedback shows that the emergency measures are not effective, the emergency handling process in the knowledge graph is optimized based on the actual situation and professional knowledge, and the alarm strategy is adjusted at the same time.

[0014] Due to the adoption of the above technical solution, the present invention has the following technical advancements compared to the prior art: The present invention provides an intelligent anomaly alarm system for ship safety monitoring based on multi-dimensional analysis. By collecting and preprocessing ship operation data from various sensors and equipment on the ship, combined with an anomaly recognition model and an alarm feature sequence table constructed based on a neural network model, it can accurately identify abnormal situations in ship operation, avoiding the limitations that may be brought about by a single data source, improving the accuracy of anomaly detection, and helping to promptly discover potential safety hazards and prevent accidents.

[0015] The present invention provides an intelligent abnormality alarm system for ship safety monitoring based on multi-dimensional analysis. Through the abnormality severity assessment module, it performs multi-dimensional analysis on the detected abnormal situations, calculates the severity assessment index, and distinguishes the severity of different abnormal situations accordingly. It effectively solves the alarm fatigue problem caused by frequent alarms in the existing technology, enables important alarms to receive timely attention and processing, and improves the efficiency of emergency response. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0017] Figure 1 It is a schematic diagram of the workflow of the present invention; Figure 2 Schematic diagram of the workflow of the abnormality severity assessment module of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0019] Example 1, as Figure 1 、 Figure 2 As shown, the present invention provides a ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis, including a safety monitoring center, which is communicatively connected to the following modules, wherein: The data acquisition and analysis module is used to collect ship operation data from various sensors and equipment on the ship, perform feature analysis, extract alarm features reflecting ship abnormalities, and integrate them to obtain an alarm feature sequence list. The data acquisition and analysis module includes a multi-source data acquisition unit and a feature analysis unit; Among them, the multi-source data acquisition unit is used to collect and pre-process ship operation data from various sensors and equipment on the ship, including the ship's navigation status, equipment operation status, environmental data and crew operation behavior, through multiple interface protocols (CAN bus, RS485, Ethernet, etc.) to establish connections with various sensors and equipment on the ship to collect ship operation data including the ship's navigation status, equipment operation status, environmental data, and crew operation behavior. Sensors include but are not limited to ship navigation status sensors (GPS, compass, depth sounder, etc.), equipment operation status sensors (engine speed sensor, temperature sensor, pressure sensor, etc.), environmental monitoring sensors (meteorological instrument, sea state monitor, etc.), and cameras and operation recorders used to monitor crew operation behavior. The collected ship operation data is pre-processed, including data cleaning and formatting. Data cleaning removes noise and outliers from the data, and missing values ​​are filled through interpolation, mean filling, etc., and data in different formats are converted to a unified format for subsequent processing. At the same time, the pre-processed ship operation data is integrated and organized according to a unified data structure to form a complete ship operation data set. During the integration process, the data from different sources are timestamp-aligned to ensure data consistency in the time dimension. The integrated ship operation data set is then stored in a local database; The feature analysis unit is used to perform feature analysis on the pre-processed ship operation data, extract alarm features, and determine the baseline value of each alarm feature based on the ship alarm requirements, and then integrate the alarm feature sequence table to comprehensively scan and analyze the pre-processed ship operation data to identify alarm features related to ship safety and abnormal status. By analyzing the statistical characteristics of the data, combined with the ship's operation mechanism and safety standards, the alarm features related to navigation status, equipment operation, environmental conditions and crew operating behavior are screened out. Among them, the navigation status alarm features include speed abnormality, course deviation and position deviation, and the equipment operation status alarm features are screened out. Features include abnormal engine speed, abnormal equipment temperature, and abnormal equipment pressure. For environmental data alarm features, they include abnormal meteorological conditions and abnormal sea conditions. For crew operation behavior alarm features, they include abnormal operation response time, abnormal operation frequency, and illegal operation behavior. Based on the ship's alarm requirements and the ship's normal operation data, the extracted alarm features are analyzed one by one to determine the baseline value. After determining the baseline value of each alarm feature, all alarm features and their corresponding baseline values ​​are integrated to form an alarm feature sequence table. At the same time, detailed description information is added to each alarm feature, including the meaning of the feature, the basis for determining the baseline value, and the possible consequences. The ship anomaly detection module is used to combine the anomaly recognition model built based on the neural network model and the alarm feature sequence table to identify abnormal situations in the ship operation, receive the ship operation data collected and pre-processed in real time, call the pre-built alarm feature sequence table, preliminarily match the received ship operation data with the alarm feature sequence table, and extract data related to the alarm features, wherein the data related to the alarm features include the ship's navigation status, equipment operation status, environmental data, and ship operation data of the crew's operating behavior, load the anomaly recognition model built based on the neural network, obtain the alarm feature data of the current ship operation from the alarm feature sequence table, and input it into the anomaly recognition model for inference. According to the input alarm feature data, each alarm feature is compared with its corresponding baseline value through the calculation of the multi-layer neural network to identify the alarm features that deviate from the baseline value. A large amount of historical ship operation data is collected, including navigation status, equipment operation, environmental conditions and crew operation behavior data under normal and abnormal conditions, and the data is cleaned and pre-processed to remove noise and outliers. Fill missing values ​​and convert data in different formats into a unified format. Extract alarm features related to ship safety and abnormal status from the preprocessed data. Based on the extracted alarm features, build an anomaly recognition model in combination with a neural network model. Divide the historical data into a training set, a validation set, and a test set. Use the training set to train the model. Adjust the model parameters through backpropagation and optimization algorithms so that the model can accurately identify normal and abnormal patterns. During the training process, use the validation set to monitor model performance to prevent overfitting and ensure the generalization ability of the model. Use the test set to evaluate the trained model. Calculate indicators such as accuracy, recall rate, and F1 value to evaluate the model's anomaly detection performance. Based on the evaluation results, optimize the model structure, parameters, or feature selection to further improve the accuracy and reliability of the model. After multiple iterations and optimizations, finally obtain an anomaly recognition model that can accurately identify abnormal ship status. Based on the inference results of the anomaly recognition model, judge the ship's operating status as abnormal. If the actual value of an alarm feature deviates from the baseline value by more than the set deviation threshold, it is marked as abnormal. The abnormality severity assessment module is used to analyze the detected abnormalities through multi-dimensional analysis, calculate the severity assessment index, distinguish the severity of different abnormalities, obtain the detected abnormality information from the ship abnormality detection module, analyze the degree of deviation of the abnormal alarm characteristics from their baseline values, and at the same time, collect the time of occurrence of the abnormality and the location of the ship. The collected multi-dimensional data is analyzed and, based on the actual needs and safety standards of the ship operation, a corresponding weight is assigned to each abnormal alarm characteristic, and then the severity assessment index is calculated. Based on the calculated severity assessment index, the abnormal situation is divided into different severity levels, namely, minor level, general level, severe level and emergency level, and a clear threshold for severity level division is set; In addition, the expression of severity evaluation index is as follows: ; Where, is the severity assessment index, which is used to measure the severity of abnormal situations. is the total number of abnormal alarm features, For the The index of an abnormal alarm feature, For the The weight of an abnormal alarm feature indicates the importance of the feature to ship safety. For the The actual value of the abnormal alarm feature, For the The baseline value of an abnormal alarm feature, For the The absolute deviation between the actual value and the reference value of an abnormal alarm feature, is the normalized deviation, which indicates the deviation ratio of the actual value relative to the reference value. When the deviation degree between the actual value of the feature and the reference value increases, The value of will increase accordingly, indicating that the severity of the anomaly increases; Multiple severity levels correspond to multiple severity level thresholds in a one-to-one manner. The corresponding relationships are as follows: Mild level: ; General Grade: ; Severity Level: ; Emergency Level: ; in, is the severity assessment index, The upper threshold for the mild level and the lower threshold for the general level, is the upper threshold of the general level and the lower threshold of the severe level, The upper threshold of the severity level and the lower threshold of the emergency level; The knowledge graph construction reasoning module is used to construct a ship safety knowledge graph and use the knowledge graph to perform reasoning analysis on detected anomalies to determine the direction of reporting to the police and the recommended handling measures; The alarm feedback module is used to combine the anomaly severity assessment results and the reasoning results of the knowledge graph to generate corresponding alarm information, and notify the alarm information to relevant personnel in various ways. At the same time, it collects feedback from relevant personnel on the alarm information, including whether the alarm was received in a timely manner, whether the alarm content is accurate and clear, and whether the emergency measures taken are effective, so as to optimize and adjust the knowledge graph, anomaly recognition model and alarm strategy.

[0020] Example 2, as Figure 1 、 Figure 2 As shown, based on Example 1, the present invention provides a technical solution: preferably, the knowledge graph construction reasoning module includes a ship safety knowledge graph construction unit and a graph reasoning unit; Among them, the ship safety knowledge graph construction unit is used to construct a ship safety knowledge graph based on ship field knowledge, historical fault cases and equipment maintenance manuals, etc., to structure entities such as ship equipment, fault types, emergency response procedures and related personnel and the relationships between them, to form a knowledge network, to achieve systematic integration and visual display of ship safety knowledge, to collect various knowledge sources in the ship field, including ship design documents, equipment maintenance manuals, historical fault cases and safety specifications, covering the operating parameters of ship equipment, fault modes, emergency response procedures and the responsibilities of related personnel, and to classify and organize the collected knowledge sources, to extract key information including equipment name, fault type, treatment measures and related personnel, and the organized key information is used as the basis for knowledge graph construction. Basic data, based on the organized knowledge sources, identifies and defines various entities and their relationships in the field of ship safety. Entities include ship equipment, fault types, emergency response processes, and relevant personnel. At the same time, the relationships between entities are clarified. Natural language processing (NLP) technology is used to extract entities and relationships from text data, and these are structured as nodes and edges in the knowledge graph. The identified entities and relationships are integrated into the ship safety knowledge graph to form a structured knowledge network. The knowledge graph uses nodes to represent entities and edges to represent relationships between entities. The knowledge graph is displayed using visualization tools. At the same time, the knowledge graph supports dynamic updates. As new fault cases and knowledge accumulate, new information can be incorporated into the graph to ensure the timeliness and accuracy of the knowledge graph. The graph reasoning unit is used to use knowledge graph reasoning technology to query and reason in the knowledge graph based on the currently detected anomaly information, infer the processing direction and emergency response strategy of the current alarm, receive the current anomaly information from the ship anomaly detection module, map the anomaly information with the entities and relationships in the ship safety knowledge graph, and determine the associated nodes of the anomaly information in the knowledge graph. Through mapping analysis, the actual anomaly problem is converted into a query object in the knowledge graph. Based on the mapping results, a deep query is performed in the knowledge graph. The query language of the graph database is used to traverse along the relationships between nodes and edges to find all information related to the current anomaly. At the same time, based on the existing rules and patterns in the knowledge graph, the potential risks and associated problems are inferred. Based on the query and reasoning results, the graph reasoning unit conducts comprehensive analysis and infers the processing direction and emergency response strategy of the current alarm. The processing direction clearly defines the overall idea of ​​solving the problem, and the emergency response strategy details the specific operation steps, required resources, and responsibilities of relevant personnel. Finally, the processing direction and emergency response strategy are output in a clear and easy-to-understand manner to provide decision support for ship safety management personnel; The alarm feedback module specifically includes: Obtain the severity level obtained by the abnormal severity assessment module and the processing direction and emergency response strategy provided by the knowledge graph reasoning unit, and then generate corresponding alarm information, clearly point out the abnormal situation, severity, recommended processing measures and emergency response steps, and notify the alarm information to relevant personnel through various means, including text messages, emails, ship internal broadcasting systems, etc., to ensure that the information can be conveyed in time. At the same time, the time and method of notification are recorded. After notifying the relevant personnel, the alarm feedback module collects the feedback of the relevant personnel on the alarm information, including whether the relevant personnel received the alarm in time, whether the alarm content is accurate and clear, and whether the emergency measures taken are effective. The module collects information such as effectiveness, classifies and organizes the collected feedback information, and summarizes it according to the feedback type, relevant personnel positions, etc. Based on the collected feedback information, the alarm feedback module optimizes and adjusts the knowledge graph, anomaly recognition model and alarm strategy. If the feedback shows that the alarm content is inaccurate and unclear, the cause is analyzed and the relevant entities and relationships in the knowledge graph are corrected, and the parameters of the anomaly recognition model are adjusted. If the feedback shows that the emergency measures are not effective, the emergency handling process in the knowledge graph is optimized in combination with actual conditions and professional knowledge, and the alarm strategy is adjusted at the same time to improve the reliability and effectiveness of the system and ensure the safety of the ship.

[0021] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. The ship safety monitoring intelligent abnormal alarm system based on multi-dimensional analysis includes a safety monitoring center, which is characterized by: The security monitoring center is connected to the following modules: The data acquisition and analysis module is used to collect ship operation data from various sensors and equipment on the ship, perform feature analysis, extract alarm features reflecting ship abnormalities, and integrate them to obtain an alarm feature sequence list; The ship anomaly detection module is used to identify abnormal conditions in ship operation by combining the anomaly recognition model built based on the neural network model and the alarm feature sequence table; The abnormality severity assessment module is used to analyze the detected abnormalities through multi-dimensional analysis, calculate the severity assessment index, distinguish the severity of different abnormalities, obtain the detected abnormality information from the ship abnormality detection module, analyze the degree of deviation of the abnormal alarm characteristics from their baseline values, and at the same time, collect the time of occurrence of the abnormality and the location of the ship. The collected multi-dimensional data is analyzed and, based on the actual needs and safety standards of the ship operation, a corresponding weight is assigned to each abnormal alarm characteristic, and then the severity assessment index is calculated. Based on the calculated severity assessment index, the abnormal situation is divided into different severity levels, namely, minor level, general level, severe level and emergency level, and a clear threshold for severity level division is set; The knowledge graph construction reasoning module is used to construct a ship safety knowledge graph and use the knowledge graph to perform reasoning analysis on detected anomalies to determine the direction of reporting to the police and the recommended handling measures; The alarm feedback module is used to combine the abnormal severity assessment results and the reasoning results of the knowledge graph to generate corresponding alarm information, and at the same time collect feedback from relevant personnel on the alarm information.

2. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 1 is characterized by: The data acquisition and analysis module includes a multi-source data acquisition unit and a feature analysis unit; The multi-source data acquisition unit is used to collect and pre-process ship operation data from various sensors and equipment of the ship; The feature analysis unit is used to perform feature analysis on the pre-processed ship operation data, extract alarm features, and determine the baseline value of each alarm feature based on the ship alarm requirements, and then integrate them to obtain an alarm feature sequence table.

3. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 2 is characterized by: The multi-source data acquisition unit specifically includes: Establish connections with various sensors and equipment on board ships through multiple interface protocols to collect ship operation data including the ship's navigation status, equipment operating status, environmental data, and crew operating behavior; The collected ship operation data is preprocessed, including data cleaning and formatting. At the same time, the preprocessed ship operation data is integrated to form a complete ship operation data set. During the integration process, the timestamps of data from different sources are aligned, and the integrated ship operation data set is stored in the local database.

4. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 2 is characterized by: The feature analysis unit specifically includes: The pre-processed ship operation data is comprehensively scanned and analyzed. By analyzing the statistical characteristics of the data and combining it with the ship's operation mechanism and safety standards, alarm features related to navigation status, equipment operation, environmental conditions, and crew operating behavior are screened out. Among them, the navigation status alarm features include abnormal speed, course deviation, and position deviation; the equipment operation status alarm features include abnormal engine speed, abnormal equipment temperature, and abnormal equipment pressure; the environmental data alarm features include abnormal meteorological conditions and abnormal sea conditions; and the crew operating behavior alarm features include abnormal operation response time, abnormal operation frequency, and illegal operation behavior; Based on the ship alarm requirements and the normal operation data of the ship, the extracted alarm features are analyzed one by one to determine the benchmark value. After determining the benchmark value of each alarm feature, all alarm features and their corresponding benchmark values ​​are integrated to form an alarm feature sequence table.

5. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 4 is characterized by: The ship anomaly detection module specifically includes: Receive real-time collected and pre-processed ship operation data, call a pre-built alarm feature sequence table, perform a preliminary match between the received ship operation data and the alarm feature sequence table, and extract data related to the alarm features, where the data related to the alarm features include the ship's navigation status, equipment operation status, environmental data, and ship operation data of crew operating behavior; Load the anomaly recognition model built based on the neural network, obtain the alarm feature data of the current ship operation from the alarm feature sequence table, and input it into the anomaly recognition model for inference. Based on the input alarm feature data, the multi-layer neural network is calculated to compare each alarm feature with its corresponding benchmark value to identify the alarm features that deviate from the benchmark value; According to the inference results of the anomaly recognition model, the ship's operating status is judged as abnormal. If the actual value of an alarm feature deviates from the baseline value by more than the set deviation threshold, it will be marked as an abnormal state.

6. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 1 is characterized by: The knowledge graph construction and reasoning module includes a ship safety knowledge graph construction unit and a graph reasoning unit; The ship safety knowledge graph construction unit is used to construct a ship safety knowledge graph based on ship domain knowledge, historical fault cases, equipment maintenance manuals and other information, and to structure the entities of ship equipment, fault types, emergency response processes and related personnel and the relationships between them to form a knowledge network; The graph reasoning unit is used to use knowledge graph reasoning technology to query and reason in the knowledge graph based on the currently detected abnormal information, and infer the processing direction and emergency response strategy of the current alarm.

7. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 6 is characterized by: The ship safety knowledge graph construction unit specifically includes: Collect various knowledge sources in the ship field, including ship design documents, equipment maintenance manuals, historical failure cases and safety regulations, covering the operating parameters, failure modes, emergency response procedures and responsibilities of relevant personnel of ship equipment. Then classify and organize the collected knowledge sources, extract key information including equipment name, failure type, treatment measures and related personnel, and use the organized key information as the basic data for constructing the knowledge graph; Based on the organized knowledge sources, various entities and their relationships in the field of ship safety are identified and defined. Entities include ship equipment, fault types, emergency response procedures, and relevant personnel. At the same time, the relationships between entities are clarified. Natural language processing technology is used to extract entities and relationships from text data and structure them as nodes and edges in the knowledge graph. The identified entities and relationships are integrated into the ship safety knowledge graph to form a structured knowledge network, in which the knowledge graph represents entities with nodes and the relationships between entities with edges, and the knowledge graph is displayed through visualization tools.

8. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 7 is characterized by: The graph reasoning unit specifically includes: Receive the current anomaly information from the ship anomaly detection module, map the anomaly information with the entities and relationships in the ship safety knowledge graph, and determine the associated nodes of the anomaly information in the knowledge graph. Through mapping analysis, convert the actual anomaly problem into a query object in the knowledge graph; Based on the mapping results, deep queries are performed in the knowledge graph. Using the query language of the graph database, the relationships between nodes and edges are traversed to find all information related to the current anomaly. At the same time, based on the existing rules and patterns in the knowledge graph, potential risks and related issues are inferred. Based on the query and reasoning results, the graph reasoning unit conducts comprehensive analysis and infers the processing direction and emergency response strategy of the current alarm. The processing direction clearly defines the overall idea of ​​solving the problem, and the emergency response strategy specifies the specific operation steps, required resources, and responsibilities of relevant personnel.

9. The ship safety monitoring intelligent abnormality alarm system based on multi-dimensional analysis according to claim 8 is characterized by: The alarm feedback module specifically includes: Obtain the severity level obtained by the abnormal severity assessment module and the processing direction and emergency response strategy provided by the knowledge graph reasoning unit, and then generate corresponding alarm information, clearly indicating the abnormal situation, severity, recommended processing measures and emergency response steps, and notify the relevant personnel of the alarm information through various means; After notifying relevant personnel, the alarm feedback module collects feedback from relevant personnel on the alarm information and classifies and organizes the collected feedback information; Based on the collected feedback information, the alarm feedback module optimizes and adjusts the knowledge graph, anomaly recognition model and alarm strategy. If the feedback shows that the alarm content is inaccurate and unclear, the cause is analyzed and the relevant entities and relationships in the knowledge graph are corrected, and the parameters of the anomaly recognition model are adjusted. If the feedback shows that the emergency measures are not effective, the emergency handling process in the knowledge graph is optimized based on the actual situation and professional knowledge, and the alarm strategy is adjusted at the same time.

Images