Online alarm intelligent analysis system for mass production and final assembly process of aircrafts

Abstract

The invention provides an online alarm intelligent analysis system for an aircraft batch production final assembly process. The online alarm intelligent analysis system comprises a data import module, a parameter rule configuration module, an alarm algorithm management module, a real-time alarm monitoring module, a model and plant configuration module, an abnormal alarm intelligent analysis module, an alarm query processing module and a log push module. The system configures a parameter alarm rule by importing multi-source data such as a test report, telemetry data, a test log and a field image, reasones from an unstructured test log by using an AI large model, diagnoses time sequence telemetry data through a deep learning algorithm, identifies production field abnormity, and automatically generates an intelligent analysis aid decision report. According to the method, rapid identification and intelligent diagnosis of the abnormity in the aircraft final assembly process are realized, reliable and effective improvement measures and suggestions are generated, and the problems that existing quality control is not timely and excessively depends on artificial experience are solved.

Description

Technical Field

[0001] This invention relates to the technical field of aircraft quality inspection, specifically to an online alarm intelligent analysis system for the mass production and final assembly process of aircraft. Background Technology

[0002] Statistical process control (SPC) is an important tool for quality management in industrial production. It's a quality management technique that utilizes statistical principles and methods to monitor and control variations in the production process. This involves collecting, analyzing, and interpreting production process data, and then implementing corresponding control and improvement measures based on the analysis results. In the mass production assembly and testing of aircraft, there can be dozens of comprehensive test items. The value of each test item can serve as a key parameter for online alarms and anomaly diagnosis. Applying SPC to the mass production assembly process of aircraft can improve the quality and yield of mass-produced aircraft products, achieving digital quality management in the mass production assembly and testing process.

[0003] However, after diagnosing products with anomalies, manual inspection of the data charts for these products is currently required to provide feedback and corrective measures. During aircraft final assembly, test logs and time-series telemetry data generated during assembly testing can assist in diagnosing anomalies. However, test logs are unstructured data that logic programs cannot process; furthermore, telemetry data has a high sampling rate, with single telemetry parameter time-series data for a single test item reaching tens to hundreds of thousands of points. Manual analysis inevitably simplifies the analysis of telemetry data parameters, hindering the effective utilization and analysis of test data.

[0004] Online alarm systems in aircraft mass production and final assembly processes require significant manpower and time to handle abnormal products, resulting in untimely online quality control, sluggish task processing, and heavy reliance on experienced designers. The reliability of the results is often strongly correlated with the designer's skill level. Therefore, to achieve rapid identification of anomalies during aircraft final assembly and to intelligently analyze relevant data to obtain reliable and effective improvement suggestions, there is an urgent need for an intelligent online alarm analysis system for aircraft mass production and final assembly processes that combines the characteristics of aircraft testing with a general artificial intelligence model. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this invention is to provide an online alarm intelligent analysis system for the mass production and final assembly process of aircraft.

[0006] The present invention provides an online alarm intelligent analysis system for the mass production and final assembly process of an aircraft, comprising: a data import module, a parameter rule configuration module, an alarm algorithm management module, a real-time alarm monitoring module, a model and factory configuration module, an abnormal alarm intelligent analysis module, an alarm query and processing module, and a log push module; The data import module is used to import test report data, raw telemetry data files, telemetry database files, test logs, and on-site production image data generated during the mass production of aircraft. The parameter rule configuration module is used to configure alarm rules for various parameters in the test reports, supporting batch configuration and individual rule settings for critical parameters. The alarm algorithm management module is used to call and edit the algorithm module of parameter alarm rules, supporting adjustments to the control line algorithm and the addition of aircraft feature alarm rules. The real-time alarm monitoring module is a visual interface used to display real-time alarm status, statistical data charts, monitoring history of different aircraft models, and the final assembly plant. The system includes: digital monitoring of the workshop; a model and factory configuration module for configuring the final assembly plant and corresponding workshops, displaying the batch production process of different aircraft models; an intelligent analysis module for abnormal alarms that automatically performs intelligent analysis on abnormal product data triggered by rules, including test log inference, time-series telemetry data diagnosis, and on-site image recognition, generating auxiliary decision-making reports; an alarm query and processing module for querying historical alarm data, auxiliary decision-making reports, and processing status, allowing engineers to review and confirm alarm processing results; and a log push module for periodically pushing abnormal alarm information and intelligent analysis auxiliary decision-making reports to alarm processing engineers.

[0007] Preferably, the data import module includes a test data import module, a telemetry data import module, and a test log image information import module; The test data import module is used to import various test report data; the telemetry data import module is used to import raw telemetry data files and telemetry database files; the test log image information import module is used to import test logs and on-site production image data.

[0008] Preferably, the parameter rule configuration module includes a parameter rule setting preview module and a parameter rule batch configuration module; The parameter rule setting preview module is used to preview the alarm triggering situation after the parameter rule is set; the parameter rule batch configuration module is used to batch configure multiple parameters under the same rule or set rules separately for key parameters.

[0009] Preferably, the alarm algorithm management module includes a control type management module and an alarm condition configuration module; The control type management module is used to manage control line algorithm types; the alarm condition configuration module is used to configure alarm condition rules.

[0010] Preferably, the real-time alarm monitoring module includes an alarm log quick display module, a statistical chart data display module, a final assembly plant workshop display module, and an alarm statistics history module; The alarm log quick display module is used to display the production calendar, daily data push volume, and number of anomalies; the statistical chart data display module is used to display statistics on total production batches, product quantity, anomaly rate, and anomaly handling time; the final assembly plant workshop display module is used to display the digital monitoring of the final assembly plant workshop and the aircraft final assembly work status of each workshop; the alarm statistics history module is used to display the historical anomaly handling status of different models of batch-produced aircraft.

[0011] Preferably, the model and factory configuration module includes a model configuration module and a factory workshop configuration module; The model configuration module is used to configure aircraft model information; the factory and workshop configuration module is used to configure final assembly plant and workshop information.

[0012] Preferably, the intelligent analysis module for abnormal alarms includes an intelligent telemetry data parsing module, a test log image reasoning module, and a decision report auxiliary generation module; The telemetry data intelligent analysis module is used to analyze telemetry data from abnormal product testing processes and perform intelligent diagnosis based on a deep learning model; the test log image reasoning module is used to perform large-scale model reasoning through test logs and on-site images to identify the causes of abnormalities; and the decision report auxiliary generation module is used to generate a decision-aiding report on corrective measures based on a prior knowledge set.

[0013] Preferably, the alarm query and processing module includes an alarm history query module, an alarm parameter details display module, and an alarm processing module; The alarm history query module is used to query historical alarm information; the alarm parameter details display module is used to display raw test data, telemetry data, test logs and images; the alarm processing module is used by engineers to review and submit processing results.

[0014] Preferably, the AI ​​big model in the abnormal alarm intelligent analysis module is a replaceable model, which is trained by importing historical aircraft production quality problem handling case sets and aircraft component performance abnormality test datasets to adjust it into a more adaptable big model.

[0015] Preferably, the time-series telemetry data aircraft component performance anomaly diagnosis algorithm in the anomaly alarm intelligent analysis module classifies anomalies through processing result feedback and continuously iterates and adjusts the deep learning model parameters.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates AI large models and deep learning algorithms to achieve intelligent fusion analysis of diverse and heterogeneous data during the mass production and final assembly of aircraft. It can automatically diagnose the causes of anomalies and generate auxiliary decision-making reports, significantly reducing the cost and time of manual analysis and improving the efficiency and accuracy of anomaly handling. 2. The system of the present invention has a highly configurable alarm rule management function, which supports the flexible setting of batch or personalized alarm thresholds and control algorithms according to the characteristics of different aircraft models and the importance of key parameters. This enhances the adaptability and practicality of the system in different production scenarios and improves the refinement and intelligence of quality control. 3. By introducing a model iterative optimization mechanism, this invention enables the system to continuously optimize the reasoning and classification capabilities of the AI ​​large model and performance diagnostic algorithm using historical anomaly processing feedback data. This continuously improves the accuracy and reliability of anomaly identification and analysis, forming a closed-loop quality improvement, and providing strong support for the stability and product consistency of the aircraft mass production and assembly process. Attached Figure Description

[0017] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram of the module structure of an online alarm intelligent analysis system for the mass production and final assembly process of an aircraft, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the rapid data analysis process of an online alarm intelligent analysis system for the mass production and final assembly of an aircraft, provided in an embodiment of the present invention. Detailed Implementation

[0018] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0019] Example 1: Reference Figure 1 and Figure 2According to the present invention, an online alarm intelligent analysis system for the mass production and final assembly process of an aircraft includes: a data import module 101, a parameter rule configuration module 102, an alarm algorithm management module 103, a real-time alarm monitoring module 104, a model and factory configuration module 105, an abnormal alarm intelligent analysis module 106, an alarm query and processing module 107, and a log push module 108. The data import module 101 is used to import test report data, raw telemetry data files, telemetry database files, test logs, and on-site production image data generated during the mass production process of the aircraft. The parameter rule configuration module 102 is used to configure alarm rules for various parameters in the test reports, supporting batch configuration and individual rule settings for critical parameters. The alarm algorithm management module 103 is used to call and edit the algorithm module of the parameter alarm rules, supporting adjustments to the control line algorithm and additions. The system includes: aircraft feature alarm rules; a real-time alarm monitoring module 104 with a visual interface for displaying real-time alarm status, statistical data charts, monitoring history of different aircraft models, and digital monitoring of the final assembly plant workshop; a model and plant configuration module 105 for configuring the final assembly plant and corresponding workshops, displaying the batch production process flow of different aircraft models; an abnormal alarm intelligent analysis module 106 for automatically performing intelligent analysis on abnormal product data that triggers the rules, including test log inference, time-series telemetry data diagnosis, and on-site image recognition, generating auxiliary decision-making reports; an alarm query and processing module 107 for querying historical alarm data, auxiliary decision-making reports, and processing status for engineers to review and confirm alarm processing results; and a log push module 108 for periodically pushing abnormal alarm information and intelligent analysis auxiliary decision-making reports to alarm processing engineers.

[0020] The data import module 101 performs structured processing on the test report data, raw telemetry data files, telemetry database files, test logs, and on-site production image data generated during the mass production of the aircraft and outputs them as a data package a1. The data package contains the data's metadata a11 (such as model, batch, product number, test category, and person in charge), parameter list a12, and specific test data a13 and telemetry data a14.

[0021] The alarm algorithm management module 103 is an algorithm module for designing and editing parameter alarm rules. It supports adjusting the control line algorithm and adding aircraft feature alarm rules to form a rule list a31.

[0022] In the parameter rule configuration module 102, the bound alarm parameter rule a21 is configured and recorded based on the data packet metadata a11 and parameter list a12, and simultaneously based on rule list a31.

[0023] In the model and factory configuration module 105, the assembly plant and corresponding workshop of each model are configured according to the data packet metadata a11, forming a list of factories corresponding to metadata a51.

[0024] The anomaly alarm intelligent analysis module 106 takes data packet a1 and alarm parameter rule a21 as input, and uses statistical methods such as consistency analysis, envelope analysis, and deviation analysis to analyze the anomalies of discrete test data a13 of the same batch of products; it uses deep learning neural networks, dynamic time warping, outlier processing, and slice alignment to analyze the anomalies of time-series telemetry data a14 of the same batch of products; and finally forms a parameter anomaly alarm list a61.

[0025] The log push module 108 takes the data packet metadata a11 and the parameter anomaly alarm list a61 as input, and pushes the parameter anomaly alarm list a61 to the person in charge to form alarm log a81.

[0026] The alarm query and processing module 107 takes the data packet metadata a11, parameter list a12, and parameter abnormal alarm list a61 as input to form item-type alarm information a71; after receiving the alarm log a81, the model-related system enters the alarm query and processing module 107 to process the item-type alarm information a71 and form alarm processing feedback information a72.

[0027] The real-time alarm monitoring module 104 takes data packet element information a11, parameter list a12, and factory list a51 as inputs and displays the batch production process flow of different models of aircraft products in a visual interface; it displays the alarm and abnormal information generated in real time for each model, as well as the feedback information after processing, in the form of item-type alarm information a71 and alarm processing feedback information a72.

[0028] The data import module 101 includes a test data import module 109, a telemetry data import module 110, and a test log image information import module 111; the test data import module 109 is used to import various test report data; the telemetry data import module 110 is used to import original telemetry data files and telemetry database files; the test log image information import module 111 is used to import test logs and on-site production image data.

[0029] The parameter rule configuration module 102 includes a parameter rule setting preview module 112 and a parameter rule batch configuration module 113; the parameter rule setting preview module 112 is used to preview the alarm triggering situation after the parameter rule is set; the parameter rule batch configuration module 113 is used to batch configure multiple parameters under the same rule or to set rules separately for key parameters.

[0030] The alarm algorithm management module 103 includes a control type management module 114 and an alarm condition configuration module 115; the control type management module 114 is used to manage control line algorithm types; the alarm condition configuration module 115 is used to configure alarm condition rules.

[0031] The real-time alarm monitoring module 104 includes an alarm log quick display module 116, a statistical chart data display module 117, a final assembly plant workshop display module 118, and an alarm statistics history module 119. The alarm log quick display module 116 is used to display the production calendar, daily data push volume, and number of anomalies. The statistical chart data display module 117 is used to display statistics on total production batches, product quantity, anomaly rate, and anomaly handling time. The final assembly plant workshop display module 118 is used to display the digital monitoring of the final assembly plant workshop and the aircraft final assembly work status of each workshop. The alarm statistics history module 119 is used to display the historical anomaly handling status of different models of batch-produced aircraft.

[0032] The model and factory configuration module 105 includes a model configuration module 120 and a factory and workshop configuration module 121; the model configuration module 120 is used to configure aircraft model information; the factory and workshop configuration module 121 is used to configure final assembly plant and workshop information.

[0033] The intelligent analysis module 106 for abnormal alarms includes a telemetry data intelligent parsing module 122, a test log image reasoning module 123, and a decision report auxiliary generation module 124. The telemetry data intelligent parsing module 122 is used to parse telemetry data from abnormal product testing processes and perform intelligent diagnosis based on a deep learning model. The test log image reasoning module 123 is used to perform large-scale model reasoning through test logs and on-site images to identify the causes of abnormalities. The decision report auxiliary generation module 124 is used to generate a decision-making auxiliary report on corrective measures based on a prior knowledge set.

[0034] The alarm query and processing module 107 includes an alarm history query module 125, an alarm parameter details display module 126, and an alarm processing module 127. The alarm history query module 125 is used to query historical alarm information. The alarm parameter details display module 126 is used to display raw test data, telemetry data, test logs, and images. The alarm processing module 127 is used by engineers to review and submit processing results.

[0035] The AI ​​model in the anomaly alarm intelligent analysis module 106 is a replaceable model. It is trained by importing historical aircraft production quality problem handling case sets and aircraft component performance anomaly test datasets, and adjusted to become a more adaptable large model. The time-series telemetry data aircraft component performance anomaly diagnosis algorithm in the anomaly alarm intelligent analysis module 106 classifies anomalies through processing result feedback and continuously iterates and adjusts the deep learning model parameters.

[0036] Example 2: This invention provides an online alarm intelligent analysis system for the mass production and final assembly process of aircraft, which enables rapid identification of product anomalies during the aircraft final assembly process, and intelligent analysis of relevant data of abnormal products to obtain reliable and effective improvement measures and suggestions, providing a high-level and highly automated technical means for quality control in the mass production process of aircraft.

[0037] An online alarm intelligent analysis system for the mass production and assembly process of aircraft is characterized by comprising: a data import module, a parameter rule configuration module, an alarm algorithm management module, a real-time alarm monitoring module, a model and factory configuration module, an abnormal alarm intelligent analysis module, an alarm query and processing module, and a log push module. The data import module imports all data files generated during the mass production process, including various test report data, raw telemetry data files collected during testing, corresponding telemetry database files, test logs, and on-site production image data. The parameter rule configuration module configures alarm rules corresponding to various parameters in the test reports, allowing for batch configuration of the same rule or the establishment of stricter alarm rules for critical parameters. The alarm algorithm management module calls the algorithm module and edits and processes parameter alarm rules based on the selected module, such as adjusting control line algorithms, and can also add alarm rules more suited to the characteristics of aircraft mass production. The real-time alarm monitoring module displays real-time alarm information through a visual interface. This interface displays various statistical data charts, historical monitoring data for different models of mass-produced aircraft, and digital monitoring of the final assembly plant workshop. The model and plant configuration module configures each final assembly plant and its corresponding workshop, providing a more intuitive view of the mass production process for different aircraft models. The intelligent analysis module for abnormal alarms automatically analyzes report data from products that trigger rules in the online alarm system. It uses AI models to reason from unstructured test logs, identifying any violations in testing procedures that caused the abnormalities. Through deep learning algorithms, it diagnoses time-series telemetry data to determine if the product's performance is abnormal. It also identifies images of the final assembly site to analyze for any abnormal personnel operations or material misuse that do not conform to normal production processes. The alarm query and processing module allows querying historical alarm data, decision support reports, and processing status. Engineers can review abnormal data and corresponding intelligent analysis reports in this module to confirm alarm processing results. The log push module pushes abnormal alarm information and intelligent analysis decision support reports to alarm processing engineers at fixed intervals (daily or weekly). Engineers review and approve these reports to complete the processing of the abnormal alarms. The AI ​​model in the abnormal alarm intelligent analysis module is a replaceable model. After selecting the best model based on the principle of optimal performance, it is trained by importing knowledge such as historical aircraft production quality problem handling case sets and aircraft component performance abnormality test datasets to adjust it into a more adaptable large model. The time-series telemetry data aircraft component performance anomaly diagnosis algorithm in the anomaly alarm intelligent analysis module classifies anomalies based on feedback from subsequent processing results and iteratively adjusts the parameters of the deep learning model. This invention, by binding and archiving batch production test data, telemetry data, test logs, and test site images of aircraft, enables more accurate and convenient tracing of the data needed for AI large-scale model inference and deep learning model diagnosis when product anomalies occur. This invention utilizes intelligent analysis algorithms combining AI large-scale model inference and deep learning algorithm diagnosis to perform multi-dimensional heterogeneous data fusion diagnosis of abnormal alarm products discovered during the batch production and assembly of aircraft, and quickly generates auxiliary decision-making reports. This solves the problem that existing methods for analyzing and recommending corrective measures for abnormal products in aircraft batch production heavily rely on manual labor, leading to untimely processing and extremely high labor costs.

[0038] Reference Figure 1An online alarm intelligent analysis system for the mass production and assembly process of an aircraft includes a data import module 101, a parameter rule configuration module 102, an alarm algorithm management module 103, a real-time alarm monitoring module 104, a model and factory configuration module 105, an abnormal alarm intelligent analysis module 106, an alarm query and processing module 107, and a log push module 108. The data import module 101 is used to import all data files generated during the mass production process of the aircraft, including various test report data, raw telemetry data files collected during testing, corresponding telemetry database files, as well as test logs and on-site production image data. The various parameters in the test report data reflect the actual state of the aircraft during production and are the key data monitored by this system. The system includes online alarms based on corresponding rules; raw telemetry data files collected during testing are parsed into readable time-series telemetry data through a telemetry database file, which, along with test logs and on-site production image data, are used for intelligent analysis after anomaly alarms; the parameter rule configuration module 102 is used to configure alarm rules corresponding to various parameters in the test report, allowing for batch configuration of the same rule or setting stricter alarm rules for critical parameters; the alarm algorithm management module 103 is used to call the algorithm module and edit and process parameter alarm rules based on the selected module, such as adjusting control line algorithms, and can also add alarm rules that better suit the characteristics of the aircraft for mass production; the real-time alarm monitoring module 104 is used to display real-time alarm status as a visual interface. This interface displays various statistical data charts, historical monitoring data for different models of mass-produced aircraft, and digital monitoring of the final assembly plant workshop. The model and plant configuration module 105 configures each final assembly plant and its corresponding workshop, providing a more intuitive view of the mass production process flow for different aircraft models. The anomaly alarm intelligent analysis module 106 automatically performs intelligent analysis on the report data of abnormal products triggered by rules in the online alarm system. It infers from unstructured test logs to determine whether there are any violations in the testing process that led to the anomaly. Through deep learning algorithms, it analyzes time-series telemetry data... The system performs component performance diagnostics to determine if the product itself is abnormal; it identifies images of the product's assembly site to analyze whether there are any abnormal personnel operations or material misuses that do not conform to normal production processes; the alarm query and processing module 107 is used to query historical alarm data, auxiliary decision-making reports, and processing status. Engineers can review abnormal data and corresponding intelligent analysis reports in this module to confirm the alarm processing results; the log push module 108 is used to push abnormal alarm information and intelligent analysis auxiliary decision-making reports to alarm processing engineers at fixed intervals (daily or weekly). After the engineers review and approve the reports, the abnormal alarms are processed.

[0039] The data import module 101 includes a test data import module 109, a telemetry data import module 110, and a test log image information import module 111; the parameter rule configuration module 102 includes a parameter rule setting preview module 112 and a parameter rule batch configuration module 113; the alarm algorithm management module 103 includes a control type management module 114 and an alarm condition configuration module 115; the real-time alarm monitoring module 104 includes an alarm log quick display module 116, a statistical chart data display module 117, a final assembly plant workshop display module 118, and an alarm statistics history module 119; the model and plant configuration module 105 includes a model configuration module 120 and a plant workshop configuration module 121; the abnormal alarm intelligent analysis module 106 includes a telemetry data intelligent parsing module 122, a test log image reasoning module 123, and a decision report auxiliary generation module 124; the alarm query and processing module 107 includes an alarm history query module 125, an alarm parameter details display module 126, and an alarm processing module 127.

[0040] See Figure 2 In this embodiment of the invention, the operator first performs initial configuration: firstly, in the model and factory configuration module 105, each final assembly plant and its corresponding workshop are configured, including model configuration and factory / workshop configuration; secondly, in the alarm algorithm management module 103, the algorithm modules to be called are edited, and parameter alarm rules are edited and processed based on the selected modules, adding feature alarm rules for the characteristics of aircraft mass production; finally, in the parameter rule configuration module 102, alarm rules corresponding to various parameters in the test report are configured; during the mass production final assembly process, after each test is completed, the test equipment automatically imports test data, telemetry data, and test log image information into the data import module 101; the system automatically judges the imported test data in the background according to the initial configuration, and checks online whether the product test data triggers alarm rules. If there is no abnormality, it continues to wait for the next data import; if an abnormality occurs when an alarm rule is triggered, the abnormal product number, product test process, and triggered alarm rule are recorded in the alarm query and processing module 107.

[0041] In the telemetry data intelligent analysis module 122 of the abnormal alarm intelligent analysis module 106, the telemetry data in the abnormal product testing process is analyzed, and intelligent diagnosis of telemetry data is performed based on a deep learning model.

[0042] The telemetry data intelligent analysis module 122 can complete data image conversion with one click and generate model input image packages. Users can select a pre-trained model or retrain the model. If retraining, the required training set samples need to be configured. If a single unit or the entire product is identified as an anomaly, a conclusion about the performance of a specific component can be directly drawn and an auxiliary decision-making report generated. If a component performance anomaly cannot be identified, the test log image reasoning module 123 in the anomaly alarm intelligent analysis module 106 uses test logs and test site images to perform large-scale model reasoning. If the cause of the anomaly can be identified, the decision report auxiliary generation module 124 automatically infers feasible corrective measures based on prior knowledge sets. If the cause of the anomaly cannot be identified, the aircraft anomaly is directly pushed. The log push module 108 pushes the anomaly alarm information to the alarm processing engineer for review. The engineer processes the anomaly alarm information in the alarm query and processing module 107 and enters the alarm history. The historical query module 125 displays the pushed information. If the performance anomaly diagnosis model can identify performance anomalies in aircraft components or the large model can infer and identify production process anomalies, the auxiliary decision-making report generated after intelligent analysis will also be pushed. For anomalies with auxiliary decision-making reports, engineers will quickly review and complete the anomaly alarm processing. For anomalies that cannot be identified by intelligent analysis, engineers will conduct concentrated efforts to address them. In the alarm parameter details display module 126, they can view the original test data, telemetry data, test logs, and images, and feed the processing results and rectification measures back to the large model to continuously iterate the model's specialized inference capabilities. The review and submission to the system will be completed in the alarm processing module 127. In the log push module 108, the push cycle for anomaly processing of different aircraft models can be set, which can be set according to the batch production cycle of different models, such as half a day, daily, weekly, etc.

[0043] The real-time alarm monitoring module 104 is a visual interface used to display real-time alarm status. The final assembly plant workshop display module 118 can display the digital monitoring of the final assembly plant workshop and the aircraft final assembly work being carried out in each workshop; the statistical chart data display module 117 displays various statistical data charts, such as total production batches, product quantity, anomaly rate, anomaly handling time, etc.; the alarm statistics history module 119 displays the historical anomaly handling status of different models of batch-produced aircraft; the alarm log quick display module 116 displays the production calendar, which can view the daily data push volume and the number of anomalies, etc.

[0044] An online alarm intelligent analysis system for the mass production and final assembly process of an aircraft includes a data import module 101, a parameter rule configuration module 102, an alarm algorithm management module 103, a real-time alarm monitoring module 104, a model and factory configuration module 105, an abnormal alarm intelligent analysis module 106, an alarm query and processing module 107, and a log push module 108.

[0045] The data import module 101 is used to import all data files generated during the mass production process of the aircraft, including various test report data, raw telemetry data files collected during testing, corresponding telemetry database files, as well as test logs and on-site production image data; the parameter rule configuration module 102 is used to configure alarm rules corresponding to various parameters in the test reports, which can configure the same rule in batches or support setting stricter alarm rules for critical parameters individually; the alarm algorithm management module 103 is used to call the algorithm module and edit and process parameter alarm rules based on the selected module; the real-time alarm monitoring module 104 is used to... The system displays real-time alarm information with a visual interface. The model and factory configuration module 105 is used to configure each final assembly plant and its corresponding workshop, enabling a more intuitive display of the batch production process of different aircraft models. The abnormal alarm intelligent analysis module 106 is used to automatically perform intelligent analysis on the report data of abnormal products that trigger rules in the online alarm system. The alarm query and processing module 107 is used to query historical alarm data, auxiliary decision reports, and processing status. The log push module 108 is used to push abnormal alarm information and intelligent analysis auxiliary decision reports to alarm processing engineers at fixed intervals.

[0046] The data import module 101 includes a test data import module 109, a telemetry data import module 110, and a test log image information import module 111. This module is used to import all data files generated during the mass production process of the aircraft, including various test report data, raw telemetry data files collected during the test, corresponding telemetry database files, test logs, and on-site production image data. Among them, the various parameters of the test report data reflect the actual status of the aircraft during the production process and are the key data monitored by this system. Online alarms are carried out through this data and corresponding rules. The raw telemetry data files collected during the test are parsed through the telemetry database file to obtain readable time-series telemetry data, which, together with the test logs and on-site production image data, are used for intelligent analysis after abnormal alarms.

[0047] The parameter rule configuration module 102 includes a parameter rule setting preview module 112 and a parameter rule batch configuration module 113, which are used to configure alarm rules corresponding to various parameters in the test report. The parameter rule setting preview module 112 can view the preview of the alarm triggered by the parameter after setting the rule. The parameter rule batch configuration module 113 is used to batch configure multiple parameters to apply the same rule, and also supports setting stricter alarm rules for critical parameters separately.

[0048] The alarm algorithm management module 103 includes a control type management module 114 and an alarm condition configuration module 115, which are used to call the algorithm module and edit and process the parameter alarm rules based on the selected module, such as adjusting the control line algorithm, and can also add alarm rules that are more in line with the characteristics of the aircraft for mass production.

[0049] The real-time alarm monitoring module 104 includes an alarm log quick display module 116, a statistical chart data display module 117, a final assembly plant workshop display module 118, and an alarm statistics history module 119, which are used to display real-time alarm status and provide a visual interface. This interface can display various statistical data charts, monitoring history of different models of batch-produced aircraft, and digital monitoring of the final assembly plant workshop, etc.

[0050] The model and factory configuration module 105 includes the model configuration module 120 and the factory workshop configuration module 121, which are used to configure each final assembly plant and the corresponding workshop, and can more intuitively display the batch production process flow of different models of aircraft products.

[0051] The anomaly alarm intelligent analysis module 106 includes a telemetry data intelligent parsing module 122, a test log image reasoning module 123, and a decision report auxiliary generation module 124. It is used to automatically perform intelligent analysis on the report data of abnormal products that trigger rules in the online alarm system, reason from unstructured test logs to determine whether there are any violations of test procedures that led to the anomaly; use deep learning algorithms to diagnose the performance anomalies of aircraft components from time-series telemetry data to determine whether the product itself is abnormal; and identify images of the product's final assembly site to analyze whether there are any abnormal personnel operations or material misuses that do not conform to normal production processes.

[0052] The alarm query and processing module 107 includes an alarm history query module 125, an alarm parameter details display module 126, and an alarm processing module 127, which are used to query historical alarm data, auxiliary decision-making reports, and processing status. Engineers can review abnormal data and corresponding intelligent analysis reports in this module to confirm the alarm processing results.

[0053] The data import module 101 includes a test data import module 109, a telemetry data import module 110, and a test log image information import module 111.

[0054] This invention relates to the field of aircraft testing technology, and more particularly to an online alarm intelligent analysis system for the mass production and final assembly process of aircraft. The system includes a data import module, a parameter rule configuration module, an alarm algorithm management module, a real-time alarm monitoring module, a model and factory configuration module, an abnormal alarm intelligent analysis module, an alarm query and processing module, and a log push module. The data import module imports all data files generated during the mass production process, including various test report data, raw telemetry data files collected during testing, and corresponding telemetry database files. The parameter rule configuration module configures alarm rules corresponding to various parameters in the test reports, allowing for batch configuration of the same rule or the establishment of stricter alarm rules for critical parameters. The alarm algorithm management module calls the algorithm module and edits and processes parameter alarm rules based on the selected module. The real-time alarm monitoring module displays the actual alarm data. The system provides real-time alarm information with a visual interface. The model and factory configuration module configures each assembly plant and its corresponding workshop, providing a more intuitive view of the batch production process for different aircraft models. The intelligent alarm analysis module automatically analyzes report data from abnormal products triggered by rules in the online alarm system. It infers from unstructured test logs, diagnoses time-series telemetry data using deep learning algorithms, identifies anomalies in the image analysis of the product's assembly site that do not conform to the normal production process, and generates an intelligent analysis-assisted decision-making report. The alarm query and processing module queries historical alarm data, assisted decision-making reports, and processing status. Engineers can review abnormal data and corresponding intelligent analysis reports in this module to confirm alarm processing results. The log push module pushes abnormal alarm information and intelligent analysis-assisted decision-making reports to alarm processing engineers at fixed intervals. This invention, based on the combination of aircraft testing characteristics and a general artificial intelligence model, enables rapid identification of anomalies during aircraft final assembly. Through AI large-scale model inference and aircraft component performance anomaly diagnosis algorithms, it performs multi-dimensional heterogeneous data fusion diagnosis on anomaly alarm products and rapidly generates auxiliary decision-making reports, obtaining reliable and effective improvement suggestions. This solves the problems of untimely online quality control and heavy reliance on designers in the handling of anomalies during existing aircraft mass production and final assembly processes. Those skilled in the art can understand this embodiment as a more specific description of Embodiment 1.

[0055] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0056] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. An online alarm intelligent analysis system for the mass production and final assembly process of an aircraft, characterized in that, include: Data import module (101), parameter rule configuration module (102), alarm algorithm management module (103), real-time alarm monitoring module (104), model and factory configuration module (105), abnormal alarm intelligent analysis module (106), alarm query and processing module (107) and log push module (108). The data import module (101) is used to import test report data, original telemetry data files, telemetry database files, test logs and on-site production image data generated during the mass production of aircraft; The parameter rule configuration module (102) is used to configure alarm rules for various parameters in the test report, and supports batch configuration and individual rule settings for critical parameters; the alarm algorithm management module (103) is used to call and edit the algorithm module of parameter alarm rules, and supports adjusting the control line algorithm and adding aircraft feature alarm rules. The real-time alarm monitoring module (104) is a visual interface used to display real-time alarm status, statistical data charts, monitoring history of different aircraft models, and digital monitoring of the final assembly plant workshop; the model and plant configuration module (105) is used to configure the final assembly plant and corresponding workshops, and to display the batch production process flow of different aircraft models; the abnormal alarm intelligent analysis module (106) is used to automatically perform intelligent analysis on abnormal product data triggered by rules, including test log reasoning, time-series telemetry data diagnosis, and on-site image recognition, and generate auxiliary decision reports; the alarm query and processing module (107) is used to query historical alarm data, auxiliary decision reports, and processing status, for engineers to review and confirm alarm processing results; the log push module (108) is used to periodically push abnormal alarm information and intelligent analysis auxiliary decision reports to alarm processing engineers.

2. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The data import module (101) includes a test data import module (109), a telemetry data import module (110), and a test log image information import module (111). The test data import module (109) is used to import various test report data; the telemetry data import module (110) is used to import the original telemetry data file and the telemetry database file; the test log image information import module (111) is used to import the test log and the on-site production image data.

3. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The parameter rule configuration module (102) includes a parameter rule setting preview module (112) and a parameter rule batch configuration module (113). The parameter rule setting preview module (112) is used to preview the alarm triggering situation after the parameter rule is set; the parameter rule batch configuration module (113) is used to batch configure multiple parameters under the same rule or set rules separately for key parameters.

4. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The alarm algorithm management module (103) includes a control type management module (114) and an alarm condition configuration module (115). The control type management module (114) is used to manage control line algorithm types; the alarm condition configuration module (115) is used to configure alarm condition rules.

5. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The real-time alarm monitoring module (104) includes an alarm log quick display module (116), a statistical chart data display module (117), a final assembly plant workshop display module (118), and an alarm statistics history module (119). The alarm log quick display module (116) is used to display the production calendar, daily data push volume and number of anomalies; the statistical chart data display module (117) is used to display the total production batches, product quantity, anomaly rate and anomaly handling time statistics; the final assembly plant workshop display module (118) is used to display the digital monitoring of the final assembly plant workshop and the aircraft final assembly work status of each workshop; the alarm statistics history module (119) is used to display the historical anomaly handling status of different models of batch-produced aircraft.

6. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The model and factory configuration module (105) includes a model configuration module (120) and a factory workshop configuration module (121). The model configuration module (120) is used to configure aircraft model information; the factory and workshop configuration module (121) is used to configure final assembly factory and workshop information.

7. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The abnormal alarm intelligent analysis module (106) includes a telemetry data intelligent parsing module (122), a test log image reasoning module (123), and a decision report auxiliary generation module (124). The telemetry data intelligent analysis module (122) is used to analyze the telemetry data of abnormal product testing process and perform intelligent diagnosis based on deep learning model; the test log image reasoning module (123) is used to perform large model reasoning through test log and on-site image to identify the cause of abnormality; the decision report auxiliary generation module (124) is used to generate a decision auxiliary report on rectification measures based on prior knowledge set.

8. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The alarm query and processing module (107) includes an alarm history query module (125), an alarm parameter details display module (126), and an alarm processing module (127). The alarm history query module (125) is used to query historical alarm information; the alarm parameter details display module (126) is used to display the original test data, telemetry data, test logs and images; the alarm processing module (127) is used for engineers to review and submit processing results.

9. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The AI ​​big model in the abnormal alarm intelligent analysis module (106) is a replaceable model. It is trained by importing historical aircraft production quality problem handling case sets and aircraft component performance abnormality test datasets, and adjusted to a more adaptable big model.

10. The intelligent online alarm analysis system for the mass production and final assembly process of aircraft according to claim 1, characterized in that, The time-series telemetry data aircraft component performance anomaly diagnosis algorithm in the anomaly alarm intelligent analysis module (106) classifies anomalies by processing the results and continuously iteratively adjusts the parameters of the deep learning model.

Images