Shafting fault diagnosis and life prediction intelligent monitoring system based on AI deep learning

Abstract

The invention discloses a shafting fault diagnosis and life prediction intelligent monitoring system based on AI deep learning, and the system comprises a shafting operation state information collection, processing and storage module which is used for collecting, processing and storing all kinds of data during the operation of a shafting in a centralized manner; the AI deep learning-based data analysis module is used for analyzing various types of data by using an AI deep learning model so as to realize shafting operation health state monitoring, shafting fault diagnosis and early warning and shafting life prediction; the man-machine interaction and decision module is used for visually displaying the operation state of the shaft system, pushing early warning information and providing disposal suggestions; and the data security management and control module is used for carrying out encryption and authority control on the system data. The system has the capabilities of real-time monitoring of the operation state of the ship shafting, fault diagnosis and life prediction, can provide auxiliary decision-making information for sailors, can provide emergency control strategies when necessary, and can greatly improve the informatization and intelligence level of the ship shafting.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of intelligent monitoring of ship propulsion systems, and particularly relates to an intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning. BACKGROUND

[0002] The ship shafting is an important part of the ship power system, and its running state is directly related to the safety and reliability of the ship. The traditional ship shafting is equipped with mechanical instruments and meters near the equipment, and when monitoring the running state of the shafting, the crew relies on daily inspection and records the instrument and meter information, and makes running state fault judgment and disposal based on the crew's experience. This method has many problems, such as incomplete monitoring, inaccurate fault diagnosis, and inaccurate life prediction. At the same time, the ship needs to be docked for maintenance regularly, and the key components such as bearing bush and sealing element are measured and repaired, so as to judge the service life of the key wear parts and develop a replacement plan. The traditional shafting state monitoring and maintenance method consumes a lot of manpower, material resources and time resources, which leads to high maintenance cost of the ship shafting, and it is difficult to meet the demand of low-cost use in the existing market. In addition, the traditional shafting monitoring method relies on the professional knowledge and engineering experience of the crew to a great extent, and it is not easy to realize automation, large-scale and high-level intelligent fault diagnosis. With the development of artificial intelligence technology, it is possible to use big data and intelligent models to monitor, diagnose and predict the ship shafting. However, the existing related technology still has defects in model training, data collection and processing, system integration, etc.

[0003] Therefore, in order to improve the informatization of the existing ship propulsion shafting system, develop the functions of shafting online monitoring, fault diagnosis and life prediction, and use the AI intelligent learning ability to iteratively optimize the function system, so as to make it have the ability of intelligent monitoring and state judgment, adapt to the future development trend of unmanned and intelligent ship, and it is an effective means for ship design to reduce cost and increase efficiency. SUMMARY

[0004] In view of the shortcomings of the traditional ship shafting state monitoring technology, the present application provides an intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning, which is mainly applied to the monitoring of the running state of the ship shafting, can collect the running data of the ship shafting, autonomously diagnose and locate the abnormal state in the running process, and propose treatment suggestions, and can further predict the service life of the key wear parts and propose disposal opinions.

[0005] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows: In a first aspect, the present application provides an intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning, which comprises: The shaft system operating status information acquisition, processing and storage module is used to centrally acquire, process and store various types of data during shaft system operation; The AI ​​deep learning-based data analysis module is used to analyze various types of data using AI deep learning models, thereby enabling shaft system operation health status monitoring, shaft system fault diagnosis and early warning, and shaft system life prediction. The human-computer interaction and decision-making module is used to visually display the shaft system's operating status, push early warning information, and provide handling suggestions. The data security management and control module is used to encrypt system data and control access permissions.

[0006] In the above scheme, the shaft system operating status information acquisition, processing and storage module includes: A distributed sensor network is used to monitor parameters of the operating status of various parts of the shaft system; The data acquisition module is used to centrally collect data from the distributed sensor network; Edge computing units are used to process data and generate file formats required for data analysis. Data storage unit, used for local storage of various types of collected and organized information.

[0007] In the above scheme, the parameters of the operating status of each part of the shaft system include at least: the bearing shell temperature, lubricating oil temperature, lubricating oil pressure, lubricating oil flow rate, machine foot vibration, and bearing shell wear of the thrust bearing; the flow rate, pressure, leakage, and sealing ring wear of the stern tube sealing device; the bearing shell temperature, bearing shell wear, and bearing vibration of the water-lubricated stern shaft bearing; the shaft section strain and vibration; and the pressure, temperature, and flow rate of the shaft system auxiliary systems.

[0008] In the above solution, the AI ​​deep learning-based data analysis module is a model library based on large models with deep learning and autonomous reasoning capabilities, including: The shaft system health status monitoring module is used to analyze the mapping relationship between various sensor parameters and shaft system health status, and to evaluate the shaft system health status through various parameters of the operating status of various parts of the shaft system. The fault diagnosis and early warning module is used to issue early warnings for parameters that exceed the healthy operating range, and to provide fault location, impact analysis, and handling suggestions to assist decision-making. The life prediction module is used to monitor the usage status of key loss components in the system in real time, analyze the wear cycle and estimate the remaining life, and provide disposal suggestions for replacing key loss components. A model management platform for model training, inference services, and iterative optimization.

[0009] In the above scheme, the data analysis module based on AI deep learning uses the Deepseek large model as the base, combines knowledge engineering information for adaptation, iteratively trains using test bench test data and actual ship test data, fine-tunes it using the LoRA method, and finally compresses it into a lightweight model through model distillation and deployment in the local single-machine monitoring system; among which, knowledge engineering information includes professional basic knowledge, design drawings, relevant standards and specifications, engineering usage experience, historical maintenance records, and expert experience.

[0010] In the above scheme, the shaft system health status monitoring module performs parameter analysis on the real-time monitored data, evaluates the health status of the shaft system operation in real time, outputs and displays key parameters of the system operation status to the crew in real time, and provides threshold ranges for key parameters under normal operating conditions. Among them, key parameters include at least the temperature parameters of key bearing parts, the pressure parameters of key system parts, the leakage parameters of sealing devices, the leakage of bearings or pressure pipelines, the vibration of key system parts, system speed, torque, power, and thrust.

[0011] In the above scheme, the fault diagnosis and early warning module performs shaft system operation status analysis and parameter analysis. Once a key parameter exceeds the threshold range, the fault analysis and processing program is activated to perform the following information processing, display, and control response: Perform condition diagnosis and information warning for potential faults in the shaft system, analyze the mapping relationship between the deviated data parameters and the health status of the shaft system, and quickly locate the fault point; Based on the parameter deviation values, a pattern analysis is performed to determine the severity of the fault hazard and to clarify its impact level on the system's operational safety. The output displays fault information, hazard mode, impact level, and handling recommendations to assist crew members in decision-making. In the event of a severe fault, proactive power reduction operation or emergency shutdown control strategies are provided to prevent the severity of the damage from escalating.

[0012] In the above scheme, the life prediction module monitors the usage status of key wear-prone components of the shafting system in real time and predicts their lifespan, while also providing disposal suggestions to assist crew decision-making, including at least: Predict the remaining life of the thrust bearing bush based on the wear parameters and provide recommendations for bush replacement. Predict the remaining life of the water-lubricated tail shaft bearing bush based on the wear parameters of the bush, and provide recommendations for bush replacement; Predict the remaining life of the seal based on the wear parameters of the sealing surface of the trans-tank sealing device, and provide recommendations for seal replacement; Predict the remaining life of the lubricating oil based on the oil composition data of the auxiliary lubricating oil system, and provide recommendations for oil replacement.

[0013] In the above solution, the human-computer interaction and decision-making module includes: Web / mobile visual dashboards are used to display the operating status of 3D axis systems; The alarm center is used to display parameters that exceed the threshold range and push the warning level with different colors; The maintenance work order system is used to provide work orders with suggestions for handling abnormal states based on the model analysis results.

[0014] In a second aspect, the present invention provides a ship that includes the intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning as described in any one of the first aspects, wherein the ship's shafting is subjected to condition monitoring, fault diagnosis and life prediction through the intelligent monitoring system.

[0015] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: This invention proposes an intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning. This system has the capability for real-time monitoring of the ship's shafting operating status, fault diagnosis, and life prediction. It can provide auxiliary decision-making information to the crew and, when necessary, emergency control strategies. It can significantly improve the informatization and intelligence level of the ship's shafting system and has the following advantages: (1) Comprehensiveness: It can comprehensively monitor the operating status of the ship shafting, including thrust, temperature, wear, vibration and other aspects, providing all-round protection for the safe operation of the ship shafting. It can centrally collect, process and store various signal sources during the stable operation of the shafting. With the help of AI deep learning data analysis model, it can realize real-time monitoring of the operating status of the ship shafting, objectively evaluate the health status of the shafting and display key operating parameters. (2) Accuracy: By accessing the large model of artificial intelligence and carrying out private model training, the accuracy of shaft fault diagnosis and life prediction is improved. It can analyze the abnormal operating state of the shaft and judge the fault, display fault information, hazard mode, impact level and propose handling suggestions, and can also detect potential faults in time and accurately predict the service life of key components. (3) Intelligent local deployment: Through the local deployment of the system's large artificial intelligence model, information collection and intelligent auxiliary judgment can be realized in the state of the ship's single machine. It does not require the external conditions required for the artificial intelligence model to run on the network. This is beneficial to the protection of data and also reduces the problem of slow system response caused by network failures, data transmission interruptions and other factors.

[0016] The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning of the present invention has certain reference value for large power units. Attached Figure Description

[0017] Figure 1This is an overall architecture diagram of an intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning, provided in an embodiment of the present invention. Figure 2 A flowchart illustrating the model building process based on AI deep learning, as provided in an embodiment of the present invention. Figure 3 This is a diagram illustrating the configuration of a distributed sensor network according to an embodiment of the present invention. Detailed Implementation

[0018] Preferred embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that the invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0019] This invention addresses the shortcomings of traditional ship shafting condition monitoring technologies by proposing an intelligent monitoring system for shafting fault diagnosis and lifespan prediction based on AI deep learning. This system is primarily used for monitoring the operational status of ship shafting systems. It can collect operational data of the ship shafting system, autonomously diagnose and locate abnormal conditions during operation, and propose handling suggestions. Furthermore, it can predict the service life of key wear-prone components and provide recommendations for remediation.

[0020] like Figure 1 As shown, the intelligent monitoring system for shaft system fault diagnosis and life prediction based on AI deep learning of the present invention includes a system architecture comprising a shaft system operating status information acquisition, processing and storage module, an AI deep learning-based data analysis module, a human-computer interaction and decision-making module, and a data security management and control module. The system also includes the specific structure and connection relationships for implementing the functions of each module.

[0021] The shaft system operating status information acquisition, processing, and storage module centrally collects, processes, and stores various signals during stable shaft system operation, providing a basic data source for system operation. This module includes a distributed sensor network, a data acquisition module, an edge computing unit, and a data storage unit. The distributed sensor network primarily monitors parameters related to the operating status of various parts of the shaft system, covering various signals such as temperature, pressure, flow rate, vibration, speed, and displacement. The data acquisition module is responsible for centrally collecting data from various sensors. The edge computing unit primarily processes edge data signals and generates file formats required for data analysis. The data storage unit primarily stores the collected and processed information locally.

[0022] The AI ​​deep learning-based data analysis module is a model library built upon a large model with deep learning and autonomous reasoning capabilities. It incorporates professional fundamentals, design drawings, relevant standards and specifications, engineering experience, historical maintenance records, and expert experience, among other knowledge engineering information, and is iteratively trained using extensive test bench data and real-ship test data. After continuous optimization driven by experimental data, the model is further distilled based on the training results, compressing it into a lightweight model that can be deployed on a local standalone monitoring system.

[0023] The AI ​​deep learning-based data analysis module is the core knowledge base of the intelligent monitoring system for shaft fault diagnosis and life prediction. It includes a shaft health status monitoring module, a fault diagnosis and early warning module, a life prediction module, and a model management platform. It has the following technical features: 1) It can effectively analyze the mapping relationship between information parameters collected by various sensors and the health status of the shaft system, and can evaluate the health status of the shaft system through various parameters of the shaft system's operating status; 2) It clearly defines the healthy operating range of sensor information parameters, and can provide timely early warnings for information parameters that exceed the range, and provide auxiliary decision-making information such as fault location, impact analysis, and handling suggestions; 3) It performs real-time status monitoring, wear cycle analysis, and remaining life prediction of the usage of key wear components in the system, and provides handling suggestions for replacing key wear components in the system.

[0024] The shafting health status monitoring module analyzes parameters based on real-time monitoring data from the shafting operation status information acquisition, processing, and storage module to achieve real-time assessment of the shafting's operational health status. This module outputs and displays key system operation parameters to the crew in real time, and provides threshold ranges for these parameters under normal operating conditions. The system's key parameters include (but are not limited to) the following: 1) temperature parameters of key bearing components; 2) pressure parameters of key system components; 3) leakage parameters of sealing devices; 4) leakage of bearings or pressure pipelines; 5) vibration of key system components; and 6) conventional parameters such as system speed, torque, power, and thrust.

[0025] The fault diagnosis and early warning module can analyze the shaft system's operating status and parameters based on the shaft system operation status information acquisition, processing, and storage module, using an AI deep learning-based data analysis module. Once a key parameter exceeds a threshold range, the fault analysis and processing program is automatically initiated, performing the following information processing, display, and control responses: 1) Analyzing the mapping relationship between the deviated data parameters and the shaft system's health status to quickly locate the fault point; 2) Conducting pattern analysis based on the severity of the fault according to the parameter deviation value to clarify its impact level on system operational safety; 3) Outputting and displaying fault information, hazard patterns, impact levels, and handling suggestions to provide a basis for crew decision-making; 4) Providing control strategies for proactive power reduction or emergency shutdown when necessary to prevent further deterioration of the system's hazard level.

[0026] The shafting life prediction module can analyze shafting operation-related parameters obtained by the shafting operation status information acquisition, processing, and storage module. Using an AI deep learning-based data analysis module, it can monitor the usage status of key wear components in the shafting in real time and predict their lifespan, while also providing disposal suggestions to assist crew decision-making. Its lifespan prediction includes, but is not limited to, the following: 1) Predicting the remaining lifespan of thrust bearing bushes based on wear parameters and providing replacement suggestions; 2) Predicting the remaining lifespan of water-lubricated stern shaft bearing bushes based on wear parameters and providing replacement suggestions; 3) Predicting the remaining lifespan of seals based on wear parameters of the through-tank sealing device sealing surface and providing replacement suggestions; 4) Predicting the remaining lifespan of lubricating oil based on the oil composition data of the auxiliary lubricating oil system and providing lubricating oil replacement suggestions.

[0027] The model management platform has self-learning capabilities, can provide private training and professional inference services, and can continuously iterate and optimize the data analysis model by combining newly collected system parameters, fault information, human factors engineering data, etc.

[0028] The human-computer interaction and decision-making module includes a web / mobile visual dashboard, an alarm center, and a maintenance work order system. It provides a user-friendly interface that intuitively displays the shafting system's operating status digitally. Specifically, the web / mobile visual dashboard shows the crew a 3D view of the shafting system's operating status; the alarm center displays parameters that exceed threshold ranges and uses differentiated colors to push warning levels, providing users with fault alarms or lifespan prediction information; and the maintenance work order system uses AI model analysis to provide users with suggestions for handling abnormal conditions.

[0029] The data security management and control module includes data encryption and access control, providing corresponding levels of control permissions for crew members of different levels, which facilitates the protection of core data or access level control.

[0030] In summary, the intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning proposed in this invention has the ability to monitor the operating status of ship shafting in real time, diagnose faults, and predict life. It can provide crew members with auxiliary decision-making information and provide emergency control strategies when necessary, which can significantly improve the informatization and intelligence level of ship shafting.

[0031] Specifically, such as Figure 1 As shown, the present invention discloses an intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning. Its system architecture comprises four layers: an edge layer, a platform layer, an application layer, and a security layer. The components and functions of each layer are as follows: the edge layer is a module for collecting, processing, and storing shaft operating status information; the platform layer is a data analysis module based on AI deep learning; the application layer is a human-computer interaction and decision-making module; and the security layer is a data security management and control module. For specific components and functions, see [link to documentation]. Figure 1 .

[0032] The shaft system operating status information acquisition, processing, and storage module includes a distributed sensor network, a data acquisition module, an edge computing unit, and a data storage unit, used for centralized acquisition, processing, and storage of various signal sources within the system. For example... Figure 3 As shown, the distributed sensor network is mainly used for parameter monitoring of the operating status of various parts of the shaft system, including but not limited to the following signals: 1) thrust bearing bearing shell temperature, lubricating oil temperature, lubricating oil pressure, lubricating oil flow rate, machine foot vibration, bearing shell wear, etc.; 2) stern tube sealing device flow rate, pressure, leakage, sealing ring wear, etc.; 3) water-lubricated stern shaft bearing shell temperature, bearing shell wear, bearing vibration, etc.; 4) shaft section strain, vibration, etc.; 5) shaft system auxiliary systems pressure, temperature, flow rate, etc.; the data acquisition module is used for centralized acquisition of various sensor data; the edge computing unit is mainly used for processing edge data signals; the data storage unit is mainly used for local storage of various types of information collected and processed.

[0033] The platform layer consists of a lightweight model based on AI deep learning data analysis modules deployed in a local standalone monitoring system. It includes a shaft health status monitoring module, a fault diagnosis and early warning module, a life prediction module, and a model management platform. It is the core database of the intelligent monitoring system for shaft fault diagnosis and life prediction.

[0034] The human-computer interaction and decision-making module provides a user-friendly interface that displays the shafting system's operating status digitally and intuitively. This includes a web / mobile visual dashboard, an alarm center, and a maintenance work order system. The web / mobile visual dashboard displays the 3D shafting system's operating status to the crew; the alarm center displays parameters exceeding threshold ranges and provides users with fault alarms or lifespan predictions, using differentiated colors to indicate warning levels; and the maintenance work order system leverages AI models to provide users with suggestions for handling abnormal conditions.

[0035] The data security management and control module includes data encryption and access control, providing corresponding levels of control permissions for crew members of different levels, which facilitates the protection of core data or access level control.

[0036] like Figure 2 As shown, this embodiment uses Deepseek's large model as a foundation and proposes specific implementation steps for an intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning: 1. Environmental preparation Hardware requirements: GPU: At least 1×NVIDIA A10 (24GB VRAM); CPU: ≥8 cores; RAM ≥ 32GB.

[0037] Software requirements: Install the latest version of Deepseek model.

[0038] 2. Data Preparation and Preprocessing Prepare relevant knowledge base materials on ship shafting and perform data preprocessing based on the aforementioned data sources. The data sources include: (1) Provide relevant materials in the marine field: maintenance manual (PDF / Word), failure case library (JSON), design specifications (Markdown); (2) Provide structured data: sensor logs (CSV), maintenance work orders (SQL), and knowledge graph triples (RDF) obtained in the laboratory or in actual ship operation.

[0039] 3. System Initialization First, complete the hardware installation and software deployment of each module, including the shaft system operating status information acquisition, processing, and storage module; the AI ​​deep learning-based data analysis module; the human-computer interaction and decision-making module; and the data security management and control module. After installation and software deployment, ensure normal communication connections between the subsystems. Additionally, calibrate all sensors in the information acquisition and processing system to ensure the accuracy of the collected data.

[0040] 4. Model Installation and LoRA Configuration Using Deepseek as the base model, the LORA method is applied for fine-tuning in the field of intelligent monitoring of ship shafting. LORA (Low-Rank Adaptation) is a parameter-efficient fine-tuning method that injects a low-rank matrix into the original model (instead of updating all parameters). This preserves the original knowledge of the base model (Deepseek) while adapting it to fault diagnosis in the specific field of ship shafting.

[0041] First, the interface standard between the ship shafting operation data and the Deepseek large model is determined, and a data transmission channel is established to ensure that the data can be accurately transmitted from the shafting operation status information acquisition, processing and storage module to the input port of the Deepseek large model. The transmitted data is then formatted and encrypted to ensure data security and compatibility.

[0042] Secondly, based on the characteristics and requirements of ship shafting operation, the initial parameters of the Deepseek large model are adjusted. For example, the weight parameters related to key factors such as thrust, temperature, and wear in shafting operation are adjusted. Through pre-training and testing with small batches of data, the model parameters are continuously optimized to better suit tasks such as monitoring ship shafting operation status, fault diagnosis, and life prediction.

[0043] Finally, the adjusted Deepseek model was integrated with other subsystems in the ship shafting localization deployment control system to ensure that the model's output data could be correctly received and used by these subsystems. The integrated system was then validated using simulated and actual ship shafting operation data to check the model's accuracy and reliability, and any issues identified were promptly adjusted and optimized.

[0044] 5. Model adaptation and training First, the collected knowledge base data related to ship shafting is labeled to clarify the shafting operating status (normal or fault type) and wear of key components, thereby providing accurate labeled data for model training.

[0045] Secondly, based on the Deepseek large-scale model framework, a private model structure suitable for monitoring the operating status, fault diagnosis, and life prediction of ship shafting systems is constructed. This model structure focuses on the representation and processing of key physical quantities (such as thrust, temperature, and vibration) during shafting operation. Preliminary cleaning and preprocessing of data collected from existing databases are performed, including removing outliers and noisy data, and standardizing the data format. Then, the labeled data is divided into training, validation, and test sets. The private model is trained using the training set, and during training, the model parameters are continuously adjusted using the validation set to avoid overfitting. Once the model achieves satisfactory performance on the validation set, the model is finally tested using the test set to evaluate its generalization ability and accuracy.

[0046] 6. Lightweight Model Processing and Local Deployment After training, the model undergoes distillation, compressing the large Deepseek model into a lightweight format to generate an AI-based deep learning data analysis module, which is then deployed on a real ship shafting monitoring device. Simultaneously, during model usage, new ship shafting operation data is continuously collected and added to the training dataset. The model is regularly optimized and updated to enhance its ability to identify new situations and faults.

[0047] 7. Operation of the AI-based deep learning-based intelligent monitoring system for shaft system fault diagnosis and life prediction When the shafting operation status information acquisition, processing, and storage module is running, it collects various data in real time during the ship's shafting operation. The collected data then undergoes preliminary cleaning and preprocessing. This process includes removing outliers and noisy data, and standardizing the data format to facilitate subsequent processing. Data cleaning methods mainly include missing value interpolation and imputation, outlier identification and removal, and noise filtering; data format standardization methods mainly include standardizing the sampling frequency, aligning timestamps, and standardizing the data output format.

[0048] The following work was carried out on the shaft system using the trained AI deep learning-based data analysis module: (1) Monitoring of shaft system operating health status Perform parameter analysis on the real-time monitoring data to assess the health status of the shaft system in real time; It outputs and displays key parameters of the system's operating status to the crew in real time, and provides threshold ranges for key parameters under normal operating conditions.

[0049] (2) Shaft system fault diagnosis and early warning Perform condition diagnosis and information warning for potential faults in the shaft system, analyze the mapping relationship between the deviated data parameters and the health status of the shaft system, and quickly locate the fault point; Based on parameter deviations, a pattern analysis is performed to determine the severity of the fault and its impact on system operational safety. The output displays fault information, hazard patterns, impact levels, and handling suggestions to assist crew members in decision-making. Provide proactive power reduction operation or emergency shutdown control strategies when necessary to prevent the system damage from further escalating.

[0050] (3) Shaft life prediction Real-time monitoring and life prediction of the service status of key wear-prone components in the shafting system, providing handling suggestions to assist crew decision-making: Predict the remaining life of the thrust bearing bush based on the wear parameters and provide recommendations for bearing bush replacement. Predict the remaining life of the water-lubricated tail shaft bearing bush based on the wear parameters of the bushing bush and provide bushing replacement recommendations; Predict the remaining lifespan of the seals based on the wear parameters of the sealing surface of the trans-tank sealing device, and provide recommendations for seal replacement. Based on the oil composition data of the auxiliary lubricating oil system, the remaining life of the lubricating oil is predicted, and oil replacement recommendations are provided.

[0051] 8. Model Optimization and System Update During model usage, new data on ship shafting operation are continuously collected and incorporated into the training dataset. The model is regularly optimized and updated to improve its ability to identify new situations and faults.

[0052] In summary, this invention presents a design scheme for an intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning. By employing the specific implementation steps described above, the large model access method based on Deepseek, and the private model training method, the system can comprehensively and accurately monitor, diagnose, and predict ship shafting, thereby improving the safety and reliability of ship shafting operation.

[0053] In addition, the present invention also provides a ship that includes the above-mentioned intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning. The ship's shafting is monitored for condition, diagnosed for fault, and predicted for life through the intelligent monitoring system.

[0054] The present invention has been described in detail above with reference to the accompanying drawings. In the above embodiments, the descriptions of each embodiment have their own emphasis; for parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. Those skilled in the art should also understand that the actions and modules involved in the specification are not necessarily essential to the present invention. Furthermore, it is understood that the steps in the method of the embodiments of the present invention can be adjusted, combined, and deleted according to actual needs, and the modules in the device of the embodiments of the present invention can be combined, divided, and deleted according to actual needs.

[0055] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. An intelligent monitoring system for shaft system fault diagnosis and life prediction based on AI deep learning, characterized in that, The system includes: The shaft system operating status information acquisition, processing and storage module is used to centrally acquire, process and store various types of data during shaft system operation; The AI ​​deep learning-based data analysis module is used to analyze various types of data using AI deep learning models, thereby enabling shaft system operation health status monitoring, shaft system fault diagnosis and early warning, and shaft system life prediction. The human-computer interaction and decision-making module is used to visually display the shaft system's operating status, push early warning information, and provide handling suggestions. The data security management and control module is used to encrypt system data and control access permissions.

2. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning as described in claim 1, characterized in that, The shaft system operating status information acquisition, processing and storage module includes: A distributed sensor network is used to monitor parameters of the operating status of various parts of the shaft system; The data acquisition module is used to centrally collect data from the distributed sensor network; Edge computing units are used to process data and generate file formats required for data analysis. Data storage unit, used for local storage of various types of collected and organized information.

3. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning as described in claim 2, characterized in that, The parameters for the operating status of each part of the shaft system include at least the following: thrust bearing bearing shell temperature, lubricating oil temperature, lubricating oil pressure, lubricating oil flow rate, machine foot vibration, bearing shell wear; stern tube sealing device flow rate, pressure, leakage, and sealing ring wear; water-lubricated stern shaft bearing bearing shell temperature, bearing shell wear, and bearing vibration; shaft section strain and vibration; and pressure, temperature, and flow rate of shaft system auxiliary systems.

4. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 2, characterized in that, The AI ​​deep learning-based data analysis module is a model library built upon large models with deep learning and autonomous reasoning capabilities, including: The shaft system health status monitoring module is used to analyze the mapping relationship between various sensor parameters and shaft system health status, and to evaluate the shaft system health status through various parameters of the operating status of various parts of the shaft system. The fault diagnosis and early warning module is used to issue early warnings for parameters that exceed the healthy operating range, and to provide fault location, impact analysis, and handling suggestions to assist decision-making. The life prediction module is used to monitor the usage status of key loss components in the system in real time, analyze the wear cycle and estimate the remaining life, and provide disposal suggestions for replacing key loss components. A model management platform for model training, inference services, and iterative optimization.

5. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 4, characterized in that, The AI-based deep learning data analysis module uses the Deepseek large model as its foundation, adapts it with knowledge engineering information, iterates and trains using test bench data and real ship test data, fine-tunes it using the LoRA method, and finally compresses it into a lightweight model through model distillation and deployment in a local single-machine monitoring system. The knowledge engineering information includes professional basic knowledge, design drawings, relevant standards and specifications, engineering experience, historical maintenance records, and expert experience.

6. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 4, characterized in that, The shaft system health status monitoring module performs parameter analysis on the real-time monitored data, evaluates the health status of the shaft system operation in real time, outputs and displays key parameters of the system operation status to the crew in real time, and provides threshold ranges for key parameters under normal operating conditions. Among them, key parameters include at least the temperature parameters of key bearing parts, the pressure parameters of key system parts, the leakage parameters of sealing devices, the leakage of bearings or pressure pipelines, the vibration of key system parts, system speed, torque, power, and thrust.

7. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 4, characterized in that, The fault diagnosis and early warning module performs shaft system operating status analysis and parameter analysis. Once a key parameter exceeds the threshold range, the fault analysis and processing program is activated to perform the following information processing, display, and control response: Perform condition diagnosis and information warning for potential faults in the shaft system, analyze the mapping relationship between the deviated data parameters and the health status of the shaft system, and quickly locate the fault point; Based on the parameter deviation values, a pattern analysis is performed to determine the severity of the fault hazard and to clarify its impact level on the system's operational safety. The output displays fault information, hazard mode, impact level, and handling recommendations to assist crew members in decision-making. In the event of a severe fault, proactive power reduction operation or emergency shutdown control strategies are provided to prevent the severity of the damage from escalating.

8. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 4, characterized in that, The life prediction module monitors the usage status of critical wear components in the shafting system in real time and predicts their lifespan, while also providing handling suggestions to assist crew decision-making, including at least: Predict the remaining life of the thrust bearing bush based on the wear parameters and provide recommendations for bush replacement. Predict the remaining life of the water-lubricated tail shaft bearing bush based on the wear parameters of the bush, and provide recommendations for bush replacement; Predict the remaining life of the seal based on the wear parameters of the sealing surface of the trans-tank sealing device, and provide recommendations for seal replacement; Predict the remaining life of the lubricating oil based on the oil composition data of the auxiliary lubricating oil system, and provide recommendations for oil replacement.

9. The intelligent monitoring system for shaft fault diagnosis and life prediction based on AI deep learning according to claim 4, characterized in that, Web / mobile visual dashboards are used to display the operating status of 3D axis systems; The alarm center is used to display parameters that exceed the threshold range and push the warning level with different colors; The maintenance work order system is used to provide work orders with suggestions for handling abnormal states based on the model analysis results.

10. A ship, characterized in that, The vessel includes the intelligent monitoring system for shafting fault diagnosis and life prediction based on AI deep learning as described in any one of claims 1 to 9, through which the vessel's shafting performs condition monitoring, fault diagnosis, and life prediction.

Images