Satellite health management system and method used in cooperation with digital twin
Through a satellite health management system combined with digital twin technology, using multiple functions and fault inversion technologies, the problem of insufficient fault analysis and prediction capabilities in traditional satellite health management methods is solved, and high reliability and efficient maintenance of satellites in orbit operation are achieved.
Patent Information
- Application Number
- CN202411779484.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-05
- Publication Date
- 2025-03-11
AI Technical Summary
Traditional satellite health management methods lack in-depth analysis and prediction capabilities for fault phenomena, making it difficult to improve the reliability and maintenance efficiency of satellites in orbit.
Using a satellite health management system combined with digital twin technology, through a variety of functions such as health data collection, fault diagnosis and disposal, operation prediction and response, performance analysis and optimization, health status evaluation and maintenance, digital twin technology is used to inversion and plan verification, and correct algorithms and inventory data.
It improves the reliability and maintenance efficiency of satellites in orbit, realizes accurate positioning and accurate handling of faults, and improves the intelligent level of satellite health management.
Smart Images
Figure CN119598379A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of satellite health management technology, and in particular to a satellite health management system and method used in conjunction with a digital twin. Background Art
[0002] With the rapid development of aerospace technology, the health management of satellites in orbit has become increasingly important. Traditional satellite health management methods mainly rely on rules and experience, and lack the ability to deeply analyze and predict fault phenomena. In recent years, the development of digital twin technology has provided a new perspective for satellite health management. By building a virtual model of the satellite, the operating status of the satellite can be simulated in a virtual environment, and fault prediction and disposal inversion can be achieved. Summary of the invention
[0003] The purpose of the present invention is to provide a satellite health management system and method for use with digital twins, using digital twin technology to perform fault inversion and fault handling plan inversion to improve the reliability, safety and maintenance efficiency of satellite on-orbit operation.
[0004] To achieve the above purpose, the technical solution adopted by the present invention is as follows:
[0005] A satellite health management system used in conjunction with a digital twin includes a health data collection subsystem, a fault diagnosis and disposal subsystem, an operation prediction and response subsystem, an efficiency analysis and optimization subsystem, a health status assessment and maintenance subsystem, and a comprehensive management subsystem, among which:
[0006] Health data collection subsystem, which collects historical and real-time data from satellite, network, and ground station equipment;
[0007] Fault diagnosis and treatment subsystem: diagnose faults on real-time data in the health data collection subsystem, perform fault inversion and fault treatment plan verification through the digital twin system, and revise the fault diagnosis algorithm, expert knowledge base, and fault plan library;
[0008] Run the prediction and response subsystem, perform fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, perform predicted fault inversion and predicted fault disposal plan inversion through digital twins, and revise the predicted fault algorithm, expert knowledge base, and predicted fault plan library;
[0009] The performance analysis and optimization subsystem conducts performance analysis on historical data in the comprehensive management subsystem and improves the performance analysis algorithm and expert knowledge base;
[0010] Health status assessment and maintenance subsystem: conduct health status assessment on historical data in the comprehensive management subsystem and improve health status assessment algorithms and expert knowledge base;
[0011] The comprehensive management subsystem manages the historical data and real-time data of the health data collection subsystem.
[0012] A satellite health management method used in conjunction with a digital twin includes the following steps:
[0013] S1, the health data collection subsystem collects historical data from satellites, networks, and ground station equipment to form a knowledge base;
[0014] S2, the fault diagnosis and disposal subsystem, operation prediction and response subsystem, efficiency analysis and optimization subsystem, health status assessment and maintenance subsystem respectively perform algorithm training based on the data of the knowledge base to obtain the trained fault diagnosis algorithm, fault prediction algorithm, efficiency analysis algorithm, and health status assessment algorithm;
[0015] S3, the fault diagnosis and disposal subsystem performs fault diagnosis on the real-time data in the health data collection subsystem, performs fault inversion and fault disposal plan verification through the digital twin system, and modifies the fault diagnosis algorithm, expert knowledge base and fault plan library in S2; the operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and modifies the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2; the efficiency analysis and optimization subsystem performs efficiency analysis on the historical data in the comprehensive management subsystem, and improves the efficiency analysis algorithm and expert knowledge base in S2; the health status assessment and maintenance subsystem performs health status assessment on the historical data in the comprehensive management subsystem, and improves the health status assessment algorithm and expert knowledge base in S2.
[0016] Preferably, in S3, the fault diagnosis and treatment subsystem performs fault diagnosis on the real-time data in the health data acquisition subsystem, performs fault inversion and fault treatment plan verification through the digital twin system, and corrects the fault diagnosis algorithm, expert knowledge base and fault plan library in S2 in the following specific steps:
[0017] A1, the fault diagnosis and disposal subsystem preprocesses the real-time data, performs fault detection and identification on the preprocessed data, and obtains the characteristic quantity;
[0018] A2, after obtaining the characteristic quantity, select the fault diagnosis model to perform fault diagnosis, obtain the fault diagnosis result, and perform fault inversion through the digital twin system;
[0019] A3, match the fault handling plan according to the fault diagnosis result to handle the fault, and perform fault handling plan inversion through the digital twin system;
[0020] A4, modify the fault diagnosis algorithm, expert knowledge base and fault plan library in S2.
[0021] Preferably, in S3, the operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and corrects the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2 in the following specific steps:
[0022] B1, run the prediction and response subsystem to extract and fuse features of real-time data and historical data to obtain prediction feature quantities;
[0023] B2, after obtaining the prediction feature quantity, select the prediction fault model to predict the fault respectively, obtain the prediction fault result, and perform the prediction fault inversion through the digital twin system;
[0024] B3, match the predicted fault handling plan according to the predicted fault result to handle the predicted fault, and invert the predicted fault handling plan through the digital twin system;
[0025] B4, revise the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2.
[0026] Preferably, in S3, the performance analysis and optimization subsystem performs performance analysis on the historical data in the comprehensive management subsystem, and the specific steps of improving the performance analysis algorithm and expert knowledge base in S2 are:
[0027] C1, extract historical data from the comprehensive management subsystem, select the performance analysis model, and obtain the performance analysis results;
[0028] C2, matching performance optimization suggestions based on the analysis results, sending the performance optimization suggestions to satellites, websites, and ground station equipment, which then formulate response measures based on the performance optimization suggestions;
[0029] C3, improve the performance analysis algorithm and expert knowledge base in S2.
[0030] Preferably, in S3, the health status assessment and maintenance subsystem performs health status assessment on historical data in the comprehensive management subsystem, and the specific steps of improving the health status assessment algorithm and expert knowledge base in S2 are:
[0031] D1, extract historical data from the comprehensive management subsystem, select the health status assessment model, and obtain the health assessment results;
[0032] D2, matching health maintenance suggestions based on health assessment results, sending health maintenance suggestions to satellites, websites, and ground station equipment, which then formulate response measures based on the health maintenance suggestions;
[0033] D3, improve the health status assessment algorithm and expert knowledge base in S2.
[0034] The beneficial effects of the present invention are as follows:
[0035] 1. It integrates multiple functions such as fault diagnosis and disposal, operation prediction and response, health status assessment and health maintenance suggestions, performance analysis and optimization suggestions, etc.
[0036] 2. Use digital twin technology to perform fault inversion and fault handling plan inversion to verify the accuracy of fault diagnosis and the effectiveness of handling. According to the verification results, correct the fault diagnosis algorithm and optimize the handling measures of the fault plan library, and ultimately achieve the goal of precise fault location, accurate handling, and fault zeroing.
[0037] 3. Combining traditional management methods with modern artificial intelligence technology, the intelligence level of satellite health management is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 It is the overall architecture of the satellite health management system.
[0039] Figure 2 This is the interaction diagram between the satellite health management system and the digital twin system. DETAILED DESCRIPTION
[0040] In order to make the purpose, technical solution and advantages of the present invention more clear, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the present invention.
[0041] like Figure 1-Figure 2 As shown, a satellite health management system used in conjunction with a digital twin includes a health data collection subsystem, a fault diagnosis and disposal subsystem, an operation prediction and response subsystem, an efficiency analysis and optimization subsystem, a health status assessment and maintenance subsystem, and a comprehensive management subsystem, among which:
[0042] The health data collection subsystem collects historical and real-time data from satellites, networks, and ground station equipment.
[0043] The fault diagnosis and disposal subsystem performs fault diagnosis on the real-time data in the health data collection subsystem, performs fault inversion and fault disposal plan verification through the digital twin system, and modifies the fault diagnosis algorithm, expert knowledge base and fault plan library.
[0044] Run the prediction and response subsystem, predict faults on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, perform predicted fault inversion and predicted fault disposal plan inversion through digital twins, and revise the predicted fault algorithm, expert knowledge base and predicted fault plan library.
[0045] The performance analysis and optimization subsystem conducts performance analysis on historical data in the comprehensive management subsystem and improves the performance analysis algorithm and expert knowledge base.
[0046] The health status assessment and maintenance subsystem conducts health status assessment on historical data in the comprehensive management subsystem and improves the health status assessment algorithm and expert knowledge base.
[0047] The comprehensive management subsystem manages the historical data and real-time data of the health data collection subsystem.
[0048] 1. Health data collection subsystem: As the data source of the health management system, it receives raw health data from satellites, ground stations, and networks, and manages data from multiple sources. At the same time, it can output the received raw health data to the outside world, which is used as the data basis for fault diagnosis, performance analysis, operation prediction, and health status assessment of the health management system.
[0049] 2. Fault diagnosis and disposal subsystem: This includes fault diagnosis and disposal suggestions for satellite systems, network systems, and ground stations. Based on the collected health data, it combines traditional methods with intelligent methods, combines fault case data and expert knowledge base data, analyzes and processes the perceived massive data, realizes fault diagnosis, and gives disposal suggestions for the fault diagnosis results.
[0050] 2.1. Comprehensive fault diagnosis: It can perform comprehensive fault diagnosis on satellites, network equipment and ground station equipment. By analyzing the pre-processed health data, using a strategy that combines traditional methods with intelligent methods, using fault case data and expert knowledge base, and conducting in-depth analysis of a large amount of data, the system can achieve rapid diagnosis and precise positioning of faults. In addition, the system can also provide effective disposal suggestions for some faults.
[0051] 2.2. Integration of digital twin technology: The fault diagnosis and treatment functions are combined with the digital twin system to simulate the fault treatment process in a virtual environment and verify and optimize the treatment suggestions. Through this integration, the system can collect feedback information, continuously revise and improve the fault plan library, so as to achieve accurate positioning and treatment of faults, and ultimately achieve the goal of zero faults.
[0052] 2.3. Fault avoidance and optimization suggestions: By accumulating fault handling experience and combining the historical fault diagnosis data of satellites and key ground equipment, we can provide research and development optimization suggestions to help avoid potential failures of key satellite equipment.
[0053] 2.4. Fault diagnosis based on expert knowledge base: The system has fault diagnosis functions based on expert knowledge base and expert rules, and supports the editing and management of rules through rule editing interactive tools to ensure the flexibility and adaptability of the diagnosis process.
[0054] 2.5. Data-driven fault diagnosis: The system supports data-driven fault diagnosis methods, including common fault diagnosis algorithms such as support vector machines (SVMs), and intelligent algorithms such as Bayesian networks. The system also has a visual algorithm parameter tuning function and a knowledge graph-based fault diagnosis function. It completes fault diagnosis by building a fault knowledge database and using fault features for knowledge reasoning.
[0055] 2.6. Scalability of knowledge, rules and algorithms: The scalability of knowledge, rules and algorithms is taken into consideration during the system design, ensuring that the system can flexibly add new diagnostic methods and algorithms as technology develops and needs change.
[0056] 2.7. Generation and optimization of fault plans: For fault handling without a plan, the system can semi-automatically or automatically generate a new draft of a fault plan through the editing interface, providing a basis for rapid response to fault handling. For fault handling with a plan, the system can semi-automatically or automatically optimize the existing fault plan through the editing interface according to the handling effect, thereby improving the handling efficiency and effect.
[0057] 2.8. User interaction and feedback mechanism: The system provides a user interaction interface, allowing operators to input feedback based on actual conditions. The system conducts self-learning and optimization based on these feedbacks, and continuously improves the accuracy of fault diagnosis and disposal.
[0058] 2.9. Security and reliability: During the design and implementation process, the system will follow the highest security and reliability standards to ensure stable operation under various conditions and reduce false alarms and missed alarms.
[0059] 2.10 Real-time monitoring and early warning: The system has real-time monitoring capabilities, which can continuously track the status of equipment and issue early warnings when abnormalities are detected, providing support for early detection and handling of faults.
[0060] Figure 2 The following is the interaction diagram between the satellite health management system and the digital twin system. The interaction process between the fault diagnosis and disposal subsystem and the digital twin system is as follows:
[0061] The fault diagnosis and disposal subsystem sends the extracted fault feature value and fault inversion request to the digital twin system. After receiving the fault inversion request, the digital twin system inputs the fault feature value to the corresponding satellite model / ground station model / network model, performs fault inversion, and checks the subsequent operation of the satellite model / ground station model / network model. The operation of the satellite model / ground station model / network model is then returned to the fault diagnosis and disposal subsystem. The fault diagnosis and disposal subsystem compares the inversion results of the digital twin with the faults diagnosed in the fault diagnosis and disposal subsystem to verify the fault diagnosis accuracy of the fault diagnosis and disposal subsystem. The fault diagnosis algorithm / rules are then corrected according to the verification results, and the goal of accurate fault location is ultimately achieved.
[0062] The fault diagnosis and disposal subsystem matches the corresponding fault disposal plan according to the diagnosed fault and sends the fault disposal plan to the digital twin system. The digital twin system performs corresponding operations according to the fault disposal plan. After executing all operations, it feeds back the operation status of the satellite model / ground station model / network model to the fault diagnosis and disposal subsystem. The fault diagnosis and disposal subsystem determines whether the fault has been eliminated or alleviated based on the operation data of the satellite model / ground station model / network model after the fault disposal of the digital twin, so as to verify the effectiveness of the fault disposal plan of the fault diagnosis and disposal subsystem. Therefore, the disposal measures of the fault plan library are optimized according to the verification results, and the goal of accurate fault disposal and zero fault is finally achieved.
[0063] 3. Operation prediction and response subsystem: Perform life prediction and fault prediction on satellite platform key equipment, payload key equipment, ground station equipment, and network equipment, and use the prediction results to avoid faults in advance and improve operation efficiency.
[0064] 3.1. Satellite platform and payload equipment life and failure prediction: The system can perform real-time life and failure prediction for key equipment on the satellite platform and the payload. By analyzing the equipment's operating data and historical maintenance records, the system can predict the possible failure time and end of life of the equipment, and based on these prediction results, make advance recommendations for decommissioning to reduce unexpected downtime and extend the service life of the equipment.
[0065] 3.2. Failure and life prediction of key equipment in ground stations: The system uses the operating data and historical maintenance data of ground station equipment to conduct in-depth failure prediction and life prediction of key equipment in ground stations. These predictions can help the maintenance team identify potential failure risks in advance and formulate corresponding maintenance plans to ensure the stable operation of the ground station.
[0066] 3.3. Failure prediction of key network equipment: The system predicts failures of key network equipment by analyzing the operation data and historical failure data of network equipment. This prediction can timely discover weak links in the network and take measures in advance to avoid large-scale network failures.
[0067] 3.4. Matching of response plans based on prediction results: The system can automatically match or suggest corresponding response plans based on the results of fault prediction and life prediction. These plans may include equipment replacement, repair, upgrade or other preventive measures to ensure the continuity and reliability of equipment and systems.
[0068] 3.5. Prediction function based on big data:
[0069] Gradual Failure Prediction: The system supports model-based gradual failure prediction, which can identify the gradual decline in equipment performance and predict the time point when failure will occur.
[0070] Sudden failure prediction: The system uses machine learning and data-driven methods to predict sudden failures, which may be caused by unexpected events or unknown factors.
[0071] Life prediction: The system supports life prediction based on empirical formulas, failure mechanisms and statistical reliability, providing a scientific basis for equipment maintenance and replacement.
[0072] 3.6. Optimization and updating of prediction fault algorithm: The system has the ability of self-learning and optimization, and can continuously update and optimize the prediction fault algorithm based on new data and feedback to improve the accuracy and reliability of the prediction.
[0073] 3.7. User customized prediction model: The system allows users to customize the prediction model according to their specific needs and device characteristics to adapt to different application scenarios and device types.
[0074] 3.8. Real-time monitoring and dynamic adjustment: The system provides real-time monitoring functions and can dynamically adjust prediction models and plans to cope with changes in equipment operating status and the impact of the external environment.
[0075] 3.9. Visual display of prediction results: The system can intuitively display the prediction results in the form of graphics and reports to help users quickly understand the prediction information and make decisions.
[0076] 3.10. Audit and record of prediction results: The system can record all prediction results and measures taken, and provide detailed historical data for equipment maintenance and fault analysis.
[0077] Figure 2 The following is the interaction diagram between the satellite health management system and the digital twin system. The interaction process between the operation prediction and response subsystem and the digital twin system is as follows:
[0078] The operation prediction and response subsystem sends the predicted fault characteristic value and the predicted fault inversion request to the digital twin system. After receiving the predicted fault inversion request, the digital twin system inputs the predicted fault characteristic value to the corresponding satellite model / ground station model / network model, performs predicted fault inversion, and checks the subsequent operation of the satellite model / ground station model / network model. The operation status of the satellite model / ground station model / network model is then returned to the operation prediction and response subsystem. The operation prediction and response subsystem compares the predicted inversion results of the digital twin with the predicted faults in the operation prediction and response subsystem to verify the predicted fault accuracy of the operation prediction and response subsystem. Therefore, the predicted fault algorithm / rules are corrected according to the verification results, and the goal of accurately predicting faults is finally achieved.
[0079] The operation prediction and response subsystem matches the predicted fault to the corresponding predicted fault disposal plan based on the predicted fault, and sends the predicted fault disposal plan and predicted fault disposal request to the digital twin system. The digital twin system performs corresponding operations according to the predicted fault disposal plan, and after executing all operations, it feeds back the operation status of the satellite model / ground station model / network model to the operation prediction and response subsystem. The operation prediction and response subsystem determines whether the predicted fault is eliminated or alleviated based on the satellite model / ground station model / network model operation data after the predicted fault disposal of the digital twin, so as to verify the effectiveness of the predicted fault disposal plan of the operation prediction and response subsystem. Therefore, the disposal measures of the predicted fault plan library are optimized according to the verification results, and the goal of early fault disposal and risk avoidance is finally achieved.
[0080] 4. Health status assessment and maintenance subsystem: Based on satellite-ground operation data and under the guidance of the health status assessment system, through the computational analysis of various assessment models, indicators, and algorithms, the health status assessment of different assessment objects in the satellite system, ground station, and network system is completed, and corresponding health maintenance suggestions are put forward.
[0081] 4.1. Comprehensive health assessment system:
[0082] The system supports comprehensive health assessment of satellite systems, network systems and ground station equipment.
[0083] Using advanced evaluation tools such as the Bayesian network model, the evaluation objects in each system are graded and a health status evaluation system is established.
[0084] 4.2. Templated service for health status assessment:
[0085] Provide evaluation template services, each template contains evaluation indicators, evaluation algorithms and confidence intervals, etc., to meet different evaluation needs.
[0086] It supports users to assign weights to health status assessment indicators based on subjective and objective factors to reflect the importance of different indicators.
[0087] The system can iteratively train health status assessment algorithms and models based on historical data, optimize confidence interval configuration, and improve the accuracy and reliability of the assessment.
[0088] 4.3. Health maintenance suggestion matching function: The system can automatically match corresponding health maintenance suggestions according to the health assessment results, including preventive maintenance, emergency repairs, etc., to ensure the optimal operation of the equipment.
[0089] 4.4. Creation and management of assessment task scenarios:
[0090] Provides assessment task scenario creation function, users can load, edit and save different assessment scenarios.
[0091] Users can expand and maintain the evaluation index parameters and algorithm information through the editing interface, which supports editing functions such as addition, deletion, modification, and query.
[0092] Provides evaluation and verification data sets with evaluation task verification functions to ensure the accuracy of evaluation results.
[0093] The evaluation results can be presented in the form of curves, data charts, etc., which is convenient for users to understand intuitively.
[0094] 4.5. Automatic generation of evaluation reports and information visualization:
[0095] The system has the function of automatically generating an assessment report based on the assessment results. The report contains detailed health status assessment, health maintenance recommendations and other information.
[0096] It provides a visual display function for health status assessment results, health maintenance recommendations and other information, allowing users to intuitively understand the assessment results and recommendations.
[0097] It supports management functions such as export, download, query, upload and update, making it convenient for users to manage data and provide decision support.
[0098] 5. Performance analysis and optimization subsystem: Based on the satellite-ground operation data, algorithms are designed to analyze and process the massive data sensed, to implement performance evaluation services for satellite platforms, payloads, satellite constellations, and networks, and to provide performance optimization suggestions and performance improvement suggestions for satellite development based on the performance analysis results.
[0099] The performance analysis and optimization subsystem includes comprehensive performance analysis functions, data input and performance analysis algorithms, operation optimization and improvement suggestions, automatic generation of analysis reports and information visualization.
[0100] 5.1. Comprehensive performance analysis function:
[0101] The system has the ability to conduct comprehensive performance analysis of satellite platforms, satellite payloads, entire satellites and satellite systems, and network systems.
[0102] It supports hierarchical and multi-dimensional performance analysis, and can conduct in-depth research on performance at different levels and aspects.
[0103] It supports comprehensive performance analysis of multiple satellite systems and can perform performance comparison and comprehensive evaluation across systems.
[0104] 5.2 Data input and analysis algorithm:
[0105] The system supports product / equipment technical parameters, satellite telemetry data, network operation data, etc. as input.
[0106] It provides extraction and analysis algorithm functions for commonly used evaluation indicator parameters, and can quickly extract key indicators based on input data.
[0107] Supports cluster analysis and other methods to extract data features and help identify patterns and trends in data.
[0108] Supports performance analysis based on adaptive dynamic programming (ADC) to achieve more accurate performance prediction and evaluation.
[0109] 5.3. Operation optimization and improvement suggestions:
[0110] Based on the performance analysis results, the system can output operational optimization suggestions for satellite systems and network systems to improve system efficiency and performance.
[0111] Output design and manufacturing improvement suggestions for the satellite as a whole or a single unit to guide subsequent product iterations and process optimization.
[0112] 5.4. Automatic generation of analysis reports and information visualization:
[0113] The system has the function of automatically generating a detailed analysis report based on the analysis results, which includes performance analysis results, optimization and improvement suggestions, etc.
[0114] Provides a visual display function for analysis results, optimization and improvement suggestions, and other information, allowing users to intuitively understand the analysis results and suggestions.
[0115] It supports management functions such as export, download, query, upload and update, making it convenient for users to manage data and provide decision support.
[0116] 6. Comprehensive management subsystem: As a visualization carrier of the health management system, it has functions such as situation display, system configuration, data management, and algorithm training.
[0117] 6.1 Data management function
[0118] Data preprocessing: Preprocess the received status data, health data, etc., including data cleaning, data noise reduction, data formatting and other operations, to form a data set with unified time information, and store the preprocessed data in a centralized manner, which can intuitively express the health information of specific satellites and key equipment of ground stations. At the same time, the processed valid data is output to each core business as a basis for fault detection, operation prediction, and analysis and evaluation.
[0119] Data storage and management: It has the management functions of receiving, sending, storing, retrieving, querying, backing up, and restoring telemetry data, network data, ground station status data, and product data in the health management process, and realizes expert knowledge base management, fault plan library management, etc.
[0120] Supports the upgrade, maintenance and management of algorithms, models, knowledge bases, scripts and other data for fault diagnosis, operation prediction, analysis and evaluation.
[0121] 6.2 System configuration function
[0122] Environmental configuration: It has environmental configuration functions such as health management system communication and storage parameters.
[0123] User management: Provides functions such as system user registration, login, logout, system user information, and permission maintenance.
[0124] 6.3 Algorithm training management function
[0125] Algorithm training management: It has various algorithm training management functions and comprehensively dispatches various subsystems to complete the algorithm training function.
[0126] Algorithm training task scheduling: Schedule each subsystem to complete algorithm training tasks such as pre-training and approximate training according to existing samples or simulated samples, so that the algorithm has practical application capabilities.
[0127] 6.4. Situation display function
[0128] Visualization carrier: As a visualization carrier of the health management system, it has display function.
[0129] Fault display function: including fault alarm, fault phenomenon, fault handling suggestions, fault verification and handling execution status, fault handling effect, etc.
[0130] Constellation health status display function: constellation health status display, custom constellation area health status display.
[0131] Ground station health status display function: health status display of the gateway station and health status display of key equipment at each station.
[0132] Operation prediction display function: life prediction and failure prediction results of each system (equipment), countermeasures, etc.
[0133] Health status analysis and evaluation result display function: display performance analysis results and optimization suggestions, health classification, dimensional evaluation results and maintenance measures.
[0134] A satellite health management method used in conjunction with a digital twin includes the following steps:
[0135] S1, the health data collection subsystem collects historical data from satellites, networks, and ground station equipment to form a knowledge base.
[0136] The health data collection subsystem receives historical data from satellites, networks, and ground stations, including space environment parameters, historical satellite operation data (including historical satellite telemetry data, historical satellite fault data and disposal plans, basic satellite data, equipment status monitoring information, software status monitoring information, etc.), historical network operation data (including network and link operation load data, network utilization data, historical network fault data, basic network operation data, basic network equipment data, etc.), and historical ground station operation data (including ground station equipment status, basic ground station data, historical ground station fault data and disposal plans, etc.). These data are stored in the database to form a knowledge base, which serves as the input for various business function algorithms of the health management system.
[0137] S2, the fault diagnosis and disposal subsystem, operation prediction and response subsystem, efficiency analysis and optimization subsystem, health status assessment and maintenance subsystem respectively perform algorithm training based on the data of the knowledge base to obtain the trained fault diagnosis algorithm, fault prediction algorithm, efficiency analysis algorithm, and health status assessment algorithm.
[0138] The fault diagnosis and disposal subsystem, operation prediction and response subsystem, efficiency analysis and optimization subsystem, and health status assessment and maintenance subsystem use the data in the knowledge base as training data and select algorithms from the model library of the comprehensive management subsystem for training. When the algorithm reaches a certain accuracy, the trained algorithm is saved as a model in the model library.
[0139] S3, the fault diagnosis and disposal subsystem performs fault diagnosis on the real-time data in the health data collection subsystem, performs fault inversion and fault disposal plan verification through the digital twin system, and modifies the fault diagnosis algorithm, expert knowledge base and fault plan library in S2; the operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and modifies the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2; the efficiency analysis and optimization subsystem performs efficiency analysis on the historical data in the comprehensive management subsystem, and improves the efficiency analysis algorithm and expert knowledge base in S2; the health status assessment and maintenance subsystem performs health status assessment on the historical data in the comprehensive management subsystem, and improves the health status assessment algorithm and expert knowledge base in S2.
[0140] (1) The fault diagnosis and treatment subsystem performs fault diagnosis on the real-time data in the health data collection subsystem, performs fault inversion and fault treatment plan verification through the digital twin system, and modifies the fault diagnosis algorithm, expert knowledge base, and fault plan library in S2. The specific steps are as follows:
[0141] A1, the fault diagnosis and disposal subsystem preprocesses the real-time data, performs fault detection and identification on the preprocessed data, and obtains feature quantities.
[0142] The health data collection subsystem receives satellite real-time telemetry data, real-time network status data, and real-time ground status data sent by satellites, networks, and ground station equipment, triggers fault diagnosis, and the fault diagnosis and disposal subsystem pre-processes the real-time telemetry data, real-time network status data, and real-time ground status data.
[0143] Fault detection and identification are performed on the preprocessed data, including preliminary fault determination through alarm code analysis and feature extraction based on various status information combined with data mining and other algorithms.
[0144] A2, after obtaining the characteristic quantity, select the fault diagnosis model to perform fault diagnosis, obtain the fault diagnosis result, and perform fault inversion through the digital twin system.
[0145] After obtaining the characteristic quantity, relevant data in the expert knowledge base and the model library are extracted, and the fault diagnosis model is selected to perform fault diagnosis to obtain the fault diagnosis result.
[0146] After obtaining the fault feature quantity, the feature quantity is sent to the digital twin system for fault inversion to check whether the digital twin system will fail. The specific process is as follows Figure 2 shown.
[0147] The fault diagnosis and disposal subsystem sends the extracted fault feature value and fault inversion request to the digital twin system. After receiving the fault inversion request, the digital twin system inputs the fault feature value to the corresponding satellite model / ground station model / network model, performs fault inversion, and checks the subsequent operation of the satellite model / ground station model / network model. The operation of the satellite model / ground station model / network model is then returned to the fault diagnosis and disposal subsystem. The fault diagnosis and disposal subsystem compares the inversion results of the digital twin with the faults diagnosed in the fault diagnosis and disposal subsystem to verify the fault diagnosis accuracy of the fault diagnosis and disposal subsystem. The fault diagnosis algorithm / rules are then corrected according to the verification results, and the goal of accurate fault location is ultimately achieved.
[0148] A3, match the fault handling plan according to the fault diagnosis result to handle the fault, and at the same time, perform fault handling plan inversion through the digital twin system.
[0149] The fault handling plan is matched according to the fault diagnosis results and sent to the satellite, website, and ground station equipment. The satellite, website, and ground station equipment perform fault handling operations according to the fault handling plan. The satellite health management system monitors the entire fault handling process, compares the consistency of the fault handling process results with the fault handling plan, and verifies the fault handling effect in real time.
[0150] Match the fault handling plan according to the fault diagnosis results, and simulate and verify the fault handling plan with the digital twin system, and provide relevant recommended handling solutions such as fault isolation and function recovery. Figure 2 shown.
[0151] The fault diagnosis and disposal subsystem matches the corresponding fault disposal plan according to the diagnosed fault and sends the fault disposal plan to the digital twin system. The digital twin system performs corresponding operations according to the fault disposal plan. After executing all operations, it feeds back the operation status of the satellite model / ground station model / network model to the fault diagnosis and disposal subsystem. The fault diagnosis and disposal subsystem determines whether the fault has been eliminated or alleviated based on the operation data of the satellite model / ground station model / network model after the fault disposal of the digital twin, so as to verify the effectiveness of the fault disposal plan of the fault diagnosis and disposal subsystem. Therefore, the disposal measures of the fault plan library are optimized according to the verification results, and the goal of accurate fault disposal and zero fault is finally achieved.
[0152] A4, modify the fault diagnosis algorithm, expert knowledge base and fault plan library in S2
[0153] The fault diagnosis results, fault handling plans and fault handling process records are displayed for users to use.
[0154] According to the fault handling effect, the fault diagnosis algorithm, expert knowledge base and fault plan library in S2 are modified.
[0155] Provide development and optimization suggestions for fault avoidance for key satellite equipment.
[0156] The fault diagnosis data, disposal suggestion data, etc. are saved in the database.
[0157] (2) The operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and modifies the predicted fault algorithm, expert knowledge base, and predicted fault plan library in S2. The specific steps are as follows:
[0158] B1. Run the prediction and response subsystem to extract and fuse features of real-time data and historical data to obtain prediction feature quantities.
[0159] B101, the health data collection subsystem collects real-time data from satellites, networks, and ground station equipment, triggers operation predictions, and extracts historical data from the comprehensive management subsystem of health management.
[0160] B102, extract and fuse features of real-time data and historical data to obtain prediction feature quantities.
[0161] B2, after obtaining the prediction feature quantity, select the prediction fault model to predict the fault respectively, obtain the prediction fault result, and perform the prediction fault inversion through the digital twin system.
[0162] After obtaining the prediction feature quantity, a prediction fault model is selected to predict the fault and obtain the prediction fault result.
[0163] After obtaining the predicted feature quantity, the predicted feature quantity is sent to the digital twin system for predicted fault inversion to check whether the digital twin system will have the predicted fault. The specific process is as follows Figure 2 shown.
[0164] The operation prediction and response subsystem sends the predicted fault characteristic value and the predicted fault inversion request to the digital twin system. After receiving the predicted fault inversion request, the digital twin system inputs the predicted fault characteristic value to the corresponding satellite model / ground station model / network model, performs predicted fault inversion, and checks the subsequent operation of the satellite model / ground station model / network model. The operation status of the satellite model / ground station model / network model is then returned to the operation prediction and response subsystem. The operation prediction and response subsystem compares the predicted inversion results of the digital twin with the predicted faults in the operation prediction and response subsystem to verify the predicted fault accuracy of the operation prediction and response subsystem. Therefore, the predicted fault algorithm / rules are corrected according to the verification results, and the goal of accurately predicting faults is finally achieved.
[0165] B3, match the predicted fault handling plan according to the predicted fault result to carry out predicted fault handling, and at the same time, invert the predicted fault handling plan through the digital twin system.
[0166] The predicted fault handling plan is matched according to the predicted fault result and sent to the satellite, website, and ground station equipment. The satellite, website, and ground station equipment perform predicted fault handling according to the predicted fault handling plan.
[0167] The predicted fault handling plan is sent to the digital twin system to perform inversion of the predicted fault handling plan and verify that the predicted fault handling plan can solve the predicted fault. Figure 2 shown.
[0168] The operation prediction and response subsystem matches the predicted fault to the corresponding predicted fault disposal plan based on the predicted fault, and sends the predicted fault disposal plan and predicted fault disposal request to the digital twin system. The digital twin system performs corresponding operations according to the predicted fault disposal plan, and after executing all operations, it feeds back the operation status of the satellite model / ground station model / network model to the operation prediction and response subsystem. The operation prediction and response subsystem determines whether the predicted fault is eliminated or alleviated based on the satellite model / ground station model / network model operation data after the predicted fault disposal of the digital twin, so as to verify the effectiveness of the predicted fault disposal plan of the operation prediction and response subsystem. Therefore, the disposal measures of the predicted fault plan library are optimized according to the verification results, and the goal of early fault disposal and risk avoidance is finally achieved.
[0169] B4, revise the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2.
[0170] The comprehensive management subsystem displays and saves the predicted fault results, and modifies the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2.
[0171] (3) The performance analysis and optimization subsystem performs performance analysis on the historical data in the comprehensive management subsystem (the historical data in the comprehensive management subsystem also includes the latest real-time data in the health data collection subsystem that has been stored in the database) at the current time / period, and the specific steps for improving the performance analysis algorithm and expert knowledge base in S2 are as follows:
[0172] C1, extract historical data from the comprehensive management subsystem, select the performance analysis model, and obtain the performance analysis results.
[0173] After the performance analysis process is manually triggered by the user or periodically and automatically triggered, historical data is extracted from the comprehensive management subsystem.
[0174] Extract the analysis indicator system according to the analysis object, and select and match the performance analysis model.
[0175] The performance analysis model combines data, algorithms and analysis methods to output performance analysis results.
[0176] C2, matches performance optimization suggestions based on the analysis results, and sends the performance optimization suggestions to satellites, websites, and ground station equipment. Satellites, websites, and ground station equipment formulate response measures based on the performance optimization suggestions.
[0177] C3, improve the performance analysis algorithm and expert knowledge base in S2.
[0178] Display and save the performance analysis results, and improve the performance analysis algorithm and expert knowledge base in S2.
[0179] (4) The health status assessment and maintenance subsystem assesses the health status of the historical data in the comprehensive management subsystem (the historical data in the comprehensive management subsystem also includes the latest real-time data in the health data collection subsystem that has been stored in the database) at the current time / period, and the specific steps for improving the health status assessment algorithm and expert knowledge base in S2 are as follows:
[0180] D1, extract historical data from the comprehensive management subsystem, select the health status assessment model, and obtain the health assessment results.
[0181] After the health status assessment process is manually triggered by the user or periodically and automatically triggered, historical data is extracted from the comprehensive management subsystem of health management.
[0182] Extract and analyze the indicator system based on the assessment object, and select and match the health status assessment model.
[0183] The model combines data, algorithms and evaluation methods to output health assessment results.
[0184] D2, match health maintenance recommendations based on health assessment results, send health maintenance recommendations to satellites, websites, and ground station equipment, and satellites, websites, and ground station equipment formulate response measures based on the health maintenance recommendations.
[0185] D3, improve the health status assessment algorithm and expert knowledge base in S2.
[0186] Display and save the health assessment results, and improve the health status assessment algorithm and expert knowledge base in S2.
Claims
1. A satellite health management system used in conjunction with a digital twin, characterized in that: It includes health data collection subsystem, fault diagnosis and disposal subsystem, operation prediction and response subsystem, performance analysis and optimization subsystem, health status assessment and maintenance subsystem, and comprehensive management subsystem, among which: Health data collection subsystem, which collects historical and real-time data from satellite, network, and ground station equipment; Fault diagnosis and treatment subsystem: diagnose faults on real-time data in the health data collection subsystem, perform fault inversion and fault treatment plan verification through the digital twin system, and revise the fault diagnosis algorithm, expert knowledge base, and fault plan library; Run the prediction and response subsystem, perform fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, perform predicted fault inversion and predicted fault disposal plan verification through digital twins, and revise the predicted fault algorithm, expert knowledge base, and predicted fault plan library; The performance analysis and optimization subsystem conducts performance analysis on historical data in the comprehensive management subsystem and improves the performance analysis algorithm and expert knowledge base; Health status assessment and maintenance subsystem: conduct health status assessment on historical data in the comprehensive management subsystem and improve health status assessment algorithms and expert knowledge base; The comprehensive management subsystem manages the historical data and real-time data of the health data collection subsystem.
2. A satellite health management method used in conjunction with digital twins, characterized in that: According to claim 1, a satellite health management system used in conjunction with a digital twin is implemented, comprising the following steps: S1, the health data collection subsystem collects historical data from satellites, networks, and ground station equipment to form a knowledge base; S2, the fault diagnosis and disposal subsystem, operation prediction and response subsystem, efficiency analysis and optimization subsystem, health status assessment and maintenance subsystem respectively perform algorithm training based on the data of the knowledge base to obtain the trained fault diagnosis algorithm, fault prediction algorithm, efficiency analysis algorithm, and health status assessment algorithm; S3, the fault diagnosis and disposal subsystem performs fault diagnosis on the real-time data in the health data collection subsystem, performs fault inversion and fault disposal plan verification through the digital twin system, and modifies the fault diagnosis algorithm, expert knowledge base and fault plan library in S2; the operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and modifies the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2; the efficiency analysis and optimization subsystem performs efficiency analysis on the historical data in the comprehensive management subsystem, and improves the efficiency analysis algorithm and expert knowledge base in S2; the health status assessment and maintenance subsystem performs health status assessment on the historical data in the comprehensive management subsystem, and improves the health status assessment algorithm and expert knowledge base in S2.
3. The satellite health management method used in conjunction with digital twins according to claim 2, characterized in that: In S3, the fault diagnosis and treatment subsystem performs fault diagnosis on the real-time data in the health data acquisition subsystem, performs fault inversion and fault treatment plan verification through the digital twin system, and corrects the fault diagnosis algorithm, expert knowledge base and fault plan library in S2 in the following specific steps: A1, the fault diagnosis and disposal subsystem preprocesses the real-time data, performs fault detection and identification on the preprocessed data, and obtains the characteristic quantity; A2, after obtaining the characteristic quantity, select the fault diagnosis model to perform fault diagnosis, obtain the fault diagnosis result, and perform fault inversion through the digital twin system; A3, match the fault handling plan according to the fault diagnosis result to handle the fault, and perform fault handling plan inversion through the digital twin system; A4, modify the fault diagnosis algorithm, expert knowledge base and fault plan library in S2.
4. The satellite health management method used in conjunction with digital twins according to claim 2, characterized in that: In S3, the operation prediction and response subsystem performs fault prediction on the real-time data in the health data collection subsystem and the historical data in the comprehensive management subsystem, performs predicted fault inversion and predicted fault disposal plan inversion through digital twins, and corrects the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2. The specific steps are: B1, run the prediction and response subsystem to extract and fuse features of real-time data and historical data to obtain prediction feature quantities; B2, after obtaining the prediction feature quantity, select the prediction fault model to predict the fault respectively, obtain the prediction fault result, and perform the prediction fault inversion through the digital twin system; B3, match the predicted fault handling plan according to the predicted fault result to handle the predicted fault, and invert the predicted fault handling plan through the digital twin system; B4, revise the predicted fault algorithm, expert knowledge base and predicted fault plan library in S2.
5. The satellite health management method used in conjunction with digital twins according to claim 2, characterized in that: In S3, the performance analysis and optimization subsystem performs performance analysis on the historical data in the comprehensive management subsystem, and the specific steps of improving the performance analysis algorithm and expert knowledge base in S2 are: C1, extract historical data from the comprehensive management subsystem, select the performance analysis model, and obtain the performance analysis results; C2, matching performance optimization suggestions based on the analysis results, sending the performance optimization suggestions to satellites, websites, and ground station equipment, which then formulate response measures based on the performance optimization suggestions; C3, improve the performance analysis algorithm and expert knowledge base in S2.
6. The satellite health management method used in conjunction with digital twins according to claim 2, characterized in that: In S3, the health status assessment and maintenance subsystem performs health status assessment on the historical data in the comprehensive management subsystem, and the specific steps of improving the health status assessment algorithm and expert knowledge base in S2 are: D1, extract historical data from the comprehensive management subsystem, select the health status assessment model, and obtain the health assessment results; D2, matching health maintenance suggestions based on health assessment results, sending health maintenance suggestions to satellites, websites, and ground station equipment, which then formulate response measures based on the health maintenance suggestions; D3, improve the health status assessment algorithm and expert knowledge base in S2.