Method and apparatus for collaborative updating of knowledge and strategy based on detection feedback

Abstract

The application discloses a knowledge and strategy cooperative updating method and device based on detection feedback. The method comprises the following steps: acquiring scene information of a current detection task; calling a target detection strategy from a detection strategy library according to the scene information to perform detection, so as to obtain a target detection result and feedback data of the target detection result; updating an abnormal knowledge graph according to the target detection result and the feedback data, and evaluating and updating the target detection strategy in the detection strategy library based on the updated abnormal knowledge graph.

Description

Technical Field

[0001] This application relates to the field of detection strategy optimization technology, and more specifically, to a method and apparatus for collaborative updating of knowledge and strategy based on detection feedback. Background Technology

[0002] In industrial online and offline inspection scenarios for film materials such as transparent films, optical films, composite films, and polarizers, long-term system operation continuously generates a large number of anomaly handling records. This data contains real-world anomaly mechanisms, effective testing experience, and failure case information from the production line, serving as a crucial knowledge resource for supporting continuous optimization of the inspection system. How to efficiently utilize this data to optimize inspection capabilities has become a common need within the industry.

[0003] Existing detection systems are mostly designed to complete basic detection tasks. They typically use separate modules to manage the detection process, algorithm parameters, and business logs, with various types of information stored and used in a discrete manner. The configuration and invocation of detection strategies rely on preset parameters and human experience, and the recording of anomalies is mainly based on raw results. This makes it difficult to transform a large amount of detection experience into reusable resources, and the value of data cannot be fully explored. Long-term operation is prone to high false detection and false negative rates, making it difficult to meet the requirements of high-precision and highly adaptable detection. Summary of the Invention

[0004] The main purpose of this application is to provide a method and apparatus for collaborative updating of knowledge and strategies based on detection feedback, in order to solve the problems that a large amount of detection experience is difficult to transform into reusable resources, the value of data cannot be fully explored, and the long-term operation is prone to high false detection and false negative rates, making it difficult to meet the requirements of high precision and high adaptability detection. It can achieve the effects of structured accumulation of detection experience, intelligent optimization of strategies, and continuous learning of the system.

[0005] To achieve the above objectives, a first aspect of this application proposes a method for collaborative updating of knowledge and strategies based on detection feedback, comprising: acquiring scene information of the current detection task; calling a target detection strategy from a detection strategy library to perform detection based on the scene information, obtaining target detection results and feedback data on the target detection results; updating an anomaly knowledge graph based on the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph.

[0006] According to the knowledge and strategy collaborative update method based on detection feedback provided in this application, the step of updating the anomaly knowledge graph based on the target detection result and the feedback data includes: extracting the anomaly type, anomaly location, and anomaly feature data from the target detection result, as well as the correction conclusion and handling association information from the feedback data; and integrating the anomaly type, anomaly location, anomaly feature data, correction conclusion, and handling association information as newly added or updated nodes and edge relationships into the anomaly knowledge graph to form an updated anomaly knowledge graph.

[0007] According to the knowledge and strategy collaborative update method based on detection feedback provided in this application, the evaluation and update of target detection strategies in the detection strategy library based on the updated anomaly knowledge graph includes: determining the anomaly feature matching degree and historical execution effect correlation data corresponding to the current detection scenario based on the updated anomaly knowledge graph; quantitatively evaluating the detection accuracy, false positive rate, and false negative rate of the target detection strategy based on the anomaly feature matching degree and the historical execution effect correlation data to obtain a quantitative evaluation result; increasing the priority of the target detection strategy when the quantitative evaluation result reaches a preset optimization threshold; and decreasing the priority of the target detection strategy or marking the target detection strategy as unoptimized when the quantitative evaluation result does not reach the preset optimization threshold.

[0008] According to the knowledge and strategy collaborative update method based on detection feedback provided in this application, after updating the anomaly knowledge graph according to the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph, the method further includes: recording the version update process of the anomaly knowledge graph and the detection strategy library; if the comprehensive detection index of the new version is lower than the preset standard, then performing a version rollback operation; if the comprehensive detection index of the new version is higher than or equal to the preset standard, then setting the new version as the default running version.

[0009] According to the knowledge and strategy collaborative update method based on detection feedback provided in this application, the feedback data includes at least one of: manual review conclusion, process treatment result, and misjudgment mark.

[0010] According to the knowledge and strategy collaborative update method based on detection feedback provided in this application, the scenario information includes at least one of the following: detection object type, detection accuracy requirements, detection environment parameters, and production batch information.

[0011] This application also provides a knowledge and strategy collaborative update device based on detection feedback, comprising the following modules: an acquisition module and a processing module; the acquisition module is used to acquire scene information of the current detection task; the processing module is used to call target detection strategies from the detection strategy library according to the scene information to perform detection, obtain target detection results and feedback data on the target detection results; update the anomaly knowledge graph according to the target detection results and the feedback data, and evaluate and update the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph.

[0012] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the knowledge and strategy collaborative update method based on detection feedback as described above.

[0013] This application also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the knowledge and policy collaborative update method based on detection feedback as described above.

[0014] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the knowledge and strategy collaborative update method based on detection feedback as described above.

[0015] The technical solutions provided by the embodiments of this application can include the following beneficial effects: Firstly, the scenario information of the current detection task is obtained, and the target detection strategy is called from the detection strategy library based on the scenario information to execute the detection. This allows for precise matching of detection strategies with actual working conditions, overcoming the drawbacks of traditional solutions that rely on fixed preset parameters and manual experience to configure strategies, effectively improving the scenario adaptability of strategy calls. Secondly, by using the target detection results and feedback data generated by the detection to update the anomaly knowledge graph, fragmented information can be transformed into structured knowledge, achieving systematic accumulation of detection experience and anomaly characteristics, fully exploring the potential value of data, and solving the problems of scattered information storage and difficulty in reusing effective experience in the prior art. Thirdly, by relying on the updated anomaly knowledge graph to evaluate and iteratively update the target detection strategies in the detection strategy library, and continuously optimizing the detection strategies based on the knowledge system formed by real detection, the probability of false detection and missed detection during long-term system operation can be significantly reduced, continuously improving the accuracy of membrane material detection and adaptability to complex working conditions. In summary, this solution achieves structured management of testing data, reusable testing experience, and dynamic autonomous optimization of testing strategies. It overcomes the shortcomings of traditional membrane material testing systems in terms of continuous optimization capabilities and can fully meet the application requirements of high precision and high adaptability in online and offline testing scenarios for industrial membrane materials. Attached Figure Description

[0016] The accompanying drawings, which form part of this application, are used to provide a further understanding of the application and to make other features, objects, and advantages of the application more apparent. The illustrative embodiments and descriptions of this application are used to explain the application and do not constitute an undue limitation of the application. In the drawings: Figure 1 A flowchart illustrating the knowledge and strategy collaborative update method based on detection feedback provided in this application; Figure 2 This is a schematic diagram of the structure of the knowledge and strategy collaborative update device based on detection feedback provided in this application; Figure 3 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0017] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0018] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0019] In this application, the terms "upper," "lower," "left," "right," "front," "rear," "top," "bottom," "inner," "outer," "middle," "vertical," "horizontal," "lateral," and "longitudinal" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are primarily for the purpose of better describing this application and its embodiments, and are not intended to limit the indicated device, element, or component to having a specific orientation, or to be constructed and operated in a specific orientation.

[0020] Furthermore, in addition to indicating location or positional relationship, some of the aforementioned terms may also have other meanings. For example, the term "above" may also be used in some cases to indicate a certain dependency or connection relationship. Those skilled in the art can understand the specific meaning of these terms in this application based on the specific circumstances.

[0021] Furthermore, the terms "installation," "setup," "equipped with," "connection," "linked," and "socketing" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or an internal connection between two devices, components, or parts. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0022] This application describes some exemplary embodiments for illustrative purposes. It should be understood that this application may be implemented in other ways not specifically shown in the accompanying drawings.

[0023] like Figure 1 As shown, this application provides a method for collaborative updating of knowledge and policies based on detection feedback. This method can be applied to a device for collaborative updating of knowledge and policies based on detection feedback. The method may include steps S101-S103: S101. The knowledge and strategy collaborative update device based on detection feedback obtains the scene information of the current detection task.

[0024] The aforementioned scenario information includes at least one of the following: the type of object to be detected, the accuracy requirements for detection, the environmental parameters for detection, and the production batch information, which is used to characterize the working conditions and requirements of the current detection task.

[0025] Specifically, in industrial online and offline testing scenarios for film materials such as transparent films, optical films, composite films, and polarizers, the knowledge and strategy collaborative update device based on testing feedback can collect the working condition information corresponding to the current testing task in real time, and extract scenario information such as the type of testing object, testing accuracy requirements, testing environment parameters, and production batch information from it, continuously acquiring and recording the real-time status of scenario information.

[0026] S102. The knowledge and strategy collaborative update device based on detection feedback calls the target detection strategy from the detection strategy library according to the scene information to perform detection, and obtains the target detection result and feedback data on the target detection result.

[0027] The aforementioned feedback data includes at least one of the following: manual review conclusions, process handling results, and misjudgment markers, used to characterize the correctness of the test results and the handling status.

[0028] Specifically, the knowledge and strategy collaborative update device based on detection feedback can compare the real-time acquired scene information with the pre-stored strategy matching conditions in the detection strategy library, select the detection strategy that matches the current scene information as the target detection strategy and call it to execute, complete the defect detection process of the membrane material to be tested, and output the target detection result automatically judged by the machine; at the same time, it acquires at least one of the following feedback data for the target detection result: manual review conclusion, process treatment result, and misjudgment mark, forming feedback information corresponding to the current detection process.

[0029] S103. The knowledge and strategy collaborative update device based on detection feedback updates the anomaly knowledge graph according to the target detection results and the feedback data, and evaluates and updates the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph.

[0030] Optionally, updating the anomaly knowledge graph based on the target detection results and the feedback data includes: extracting the anomaly type, anomaly location, and anomaly feature data from the target detection results, as well as the correction conclusion and handling association information from the feedback data; and integrating the anomaly type, anomaly location, anomaly feature data, correction conclusion, and handling association information as newly added or updated nodes and edge relationships into the anomaly knowledge graph to form an updated anomaly knowledge graph.

[0031] Specifically, the knowledge and strategy collaborative update device based on detection feedback can analyze and extract features from the target detection results, separating structured detection information such as anomaly type, anomaly location, and anomaly feature data. At the same time, it extracts feedback data to obtain manually given correction conclusions and corresponding handling information related to the production line. The extracted information is mapped to entity nodes and associated edges in the anomaly knowledge graph. Existing identical nodes and edge relationships in the graph are covered and updated, and missing information is added and supplemented to complete the fusion and storage of information, forming a more comprehensive and more accurate updated anomaly knowledge graph.

[0032] Optionally, the anomaly knowledge graph node may include at least one or more of the following: sample type, anomaly type, stratum label, manifest evidence, collection strategy, fusion strategy, human conclusion, and process result.

[0033] The above solution can transform discrete, unstructured detection results and feedback information into a standardized, reusable knowledge structure, enabling the systematic accumulation of detection experience and anomaly mechanisms, fully exploring the value of data throughout the entire industrial testing process, and solving the problems of information dispersion and the inability to inherit and reuse experience in existing technologies.

[0034] Optionally, the step of evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph includes: determining the anomaly feature matching degree and historical execution effect association data corresponding to the current detection scenario based on the updated anomaly knowledge graph; quantitatively evaluating the detection accuracy, false positive rate, and false negative rate of the target detection strategy based on the anomaly feature matching degree and the historical execution effect association data to obtain a quantitative evaluation result; increasing the priority of the target detection strategy when the quantitative evaluation result reaches a preset optimization threshold; and decreasing the priority of the target detection strategy or marking the target detection strategy as unoptimized when the quantitative evaluation result does not reach the preset optimization threshold.

[0035] Specifically, the knowledge and strategy collaborative update device based on detection feedback can perform matching calculations on abnormal features in the current detection scenario according to the updated abnormal knowledge graph to obtain the abnormal feature matching degree, and retrieve the historical execution effect correlation data of the target detection strategy in the same or similar scenarios; using the abnormal feature matching degree and historical execution effect correlation data as the evaluation basis, the detection accuracy, false positive rate, and false negative rate of the actual operation of the target detection strategy are statistically and quantitatively calculated to obtain the quantitative evaluation result; the quantitative evaluation result is compared with the preset optimization threshold. If the quantitative evaluation result reaches the preset optimization threshold, the calling priority of the target detection strategy in the detection strategy library is increased; if the quantitative evaluation result does not reach the preset optimization threshold, the calling priority of the target detection strategy is reduced, or the target detection strategy is directly marked as to be optimized and synchronized to the detection strategy library.

[0036] For example, in the scenario of detecting scratches on the surface of a polarizer, the collaborative update device can match the current abnormal features with historical scratch features with a matching degree of 96% through the updated anomaly knowledge graph, and retrieve the historical detection accuracy of the target detection strategy as 98%, false alarm rate as 1.2%, and false negative rate as 0.8%. If the calculated quantitative evaluation result reaches the preset optimization threshold, the priority of the target detection strategy is upgraded from medium to high. If the detection accuracy of a target detection strategy is only 82% and the false alarm rate is 8.5% in the same scenario, and the evaluation result does not reach the preset optimization threshold, its priority is reduced from high to low and marked as a state to be optimized.

[0037] The above solution enables closed-loop quantitative evaluation and intelligent iteration of detection strategies by relying on structured and continuously updated anomaly knowledge. This breaks away from the limitations of traditional detection systems that rely on manual experience to adjust strategies, ensuring that strategy priorities are consistent with actual detection results. This effectively reduces the probability of false detections and missed detections, and continuously improves the detection accuracy and scenario adaptability of membrane material detection systems.

[0038] It should be noted that the updated anomaly knowledge graph and detection strategy library can provide standardized and structured knowledge support for intelligent agents, which can be used by intelligent agents to retrieve, match and reason in key stages such as anomaly hypothesis generation, data acquisition path planning and detection result interpretation, thereby improving the accuracy, adaptability and decision interpretability of the entire intelligent detection process.

[0039] Optionally, after updating the anomaly knowledge graph based on the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph, the knowledge and strategy collaborative update device based on detection feedback can record the version update process of the anomaly knowledge graph and the detection strategy library; if the comprehensive detection index of the new version is lower than the preset standard, a version rollback operation is performed; if the comprehensive detection index of the new version is higher than or equal to the preset standard, the new version is set as the default running version.

[0040] Specifically, after completing the update of the anomaly knowledge graph and the detection strategy library, the knowledge and strategy collaborative update device based on detection feedback can assign a unique version number to the updated content, record the update time, update content, parameter comparison before and after the update, and related detection task information, forming a complete version record and archiving it. Then, the collaborative update device can perform online operation monitoring on the updated new version, calculate comprehensive detection indicators and compare them with preset standards. If the comprehensive detection indicators of the new version are lower than the preset standards, it is determined that the update effect has deteriorated, and an automatic version rollback operation is performed to restore the anomaly knowledge graph and the detection strategy library to the most recent historical version that is running normally. If the comprehensive detection indicators of the new version are higher than or equal to the preset standards, it is determined that the update is effective, and the new version is set as the system default running version.

[0041] For example, in the transparent film defect detection scenario, after the collaborative update device completes the map and strategy update, it generates version V1.2. The online monitoring statistics show that the new version has a detection accuracy of 92% and a false negative rate of 2.1%, which is lower than the preset standard of 95% accuracy and 1% false negative rate. Therefore, it will automatically roll back to the normally operating version V1.1. If the new version has a detection accuracy of 96% and a false negative rate of 0.9%, which meets the preset standard, then V1.2 will be set as the default running version.

[0042] The above solution enables the full update process of the anomaly knowledge graph and detection strategy base to be recorded, compared, and traceable, avoiding performance degradation due to update failure, quickly restoring stable operation when update degradation occurs, ensuring the continuous and reliable operation of the detection system, and improving overall operational security and system robustness.

[0043] It should be noted that for detection strategies that conflict in the detection strategy library, the collaborative update device can retain multiple parallel versions and independently record and distinguish the applicable scenario information, applicable conditions, anomaly types and operating effects of each version strategy. This clearly defines the applicable boundaries of different version strategies, avoids mutual interference between different strategies, and ensures that the agent can match suitable and effective strategy entries under different detection conditions, thereby improving the detection stability and strategy availability in complex scenarios.

[0044] In this embodiment, since the scenario information of the current detection task is first obtained, and the target detection strategy is called from the detection strategy library to execute the detection based on the scenario information, the detection strategy can be accurately matched with the actual working conditions. This avoids the drawbacks of traditional solutions that rely on fixed preset parameters and manual experience to configure strategies, and effectively improves the scenario adaptability of strategy calls. Since the target detection results and feedback data generated by the detection are used to update the anomaly knowledge graph, fragmented information can be transformed into structured knowledge, realizing the systematic accumulation of detection experience and anomaly characteristics, fully exploring the potential value of data, and solving the problems of scattered information storage and difficulty in reusing effective experience in the prior art. Furthermore, since the target detection strategy in the detection strategy library is evaluated and iteratively updated based on the updated anomaly knowledge graph, and the detection strategy is continuously optimized based on the knowledge system formed by real detection, the probability of false detection and missed detection during long-term system operation can be significantly reduced, and the detection accuracy of membrane materials and adaptability to complex working conditions can be continuously improved. In summary, this solution achieves structured management of testing data, reusable testing experience, and dynamic autonomous optimization of testing strategies. It overcomes the shortcomings of traditional membrane material testing systems in terms of continuous optimization capabilities and can fully meet the application requirements of high precision and high adaptability in online and offline testing scenarios for industrial membrane materials.

[0045] The foregoing mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, it includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0046] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0047] It should be noted that the apparatus in the embodiments of this application includes a virtual apparatus and a physical apparatus. The virtual apparatus can be a knowledge and strategy collaborative update apparatus based on detection feedback, and the physical apparatus can include electronic devices, computer storage media, and computer program products.

[0048] The knowledge and policy collaborative update method based on detection feedback provided in this application can be executed by a knowledge and policy collaborative update device based on detection feedback, or a control module within that device for knowledge and policy collaborative update based on detection feedback. This application uses the execution of the knowledge and policy collaborative update method based on detection feedback by a knowledge and policy collaborative update device based on detection feedback as an example to illustrate the knowledge and policy collaborative update device based on detection feedback provided in this application.

[0049] It should be noted that, according to the above method examples, the knowledge and strategy collaborative update device based on detection feedback can be divided into functional modules. For example, each function can be divided into its own functional modules, or two or more functions can be integrated into one processing module. The integrated modules can be implemented in hardware or as software functional modules. Optionally, the module division in this embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0050] like Figure 2 As shown in the figure, this application embodiment provides a knowledge and strategy collaborative update device 200 based on detection feedback. This knowledge and strategy collaborative update device 200 based on detection feedback includes: an acquisition module 201 and a processing module 202. The acquisition module 201 is used to acquire scene information of the current detection task; the processing module 202 is used to call target detection strategies from a detection strategy library according to the scene information to perform detection, obtain target detection results and feedback data on the target detection results; update an anomaly knowledge graph according to the target detection results and the feedback data, and evaluate and update the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph.

[0051] Optionally, updating the anomaly knowledge graph based on the target detection results and the feedback data includes: extracting the anomaly type, anomaly location, and anomaly feature data from the target detection results, as well as the correction conclusion and handling association information from the feedback data; and integrating the anomaly type, anomaly location, anomaly feature data, correction conclusion, and handling association information as newly added or updated nodes and edge relationships into the anomaly knowledge graph to form an updated anomaly knowledge graph.

[0052] Optionally, the step of evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph includes: determining the anomaly feature matching degree and historical execution effect association data corresponding to the current detection scenario based on the updated anomaly knowledge graph; quantitatively evaluating the detection accuracy, false positive rate, and false negative rate of the target detection strategy based on the anomaly feature matching degree and the historical execution effect association data to obtain a quantitative evaluation result; increasing the priority of the target detection strategy when the quantitative evaluation result reaches a preset optimization threshold; and decreasing the priority of the target detection strategy or marking the target detection strategy as unoptimized when the quantitative evaluation result does not reach the preset optimization threshold.

[0053] Optionally, after updating the anomaly knowledge graph based on the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph, the processing module 202 is used to record the version of the update process of the anomaly knowledge graph and the detection strategy library; if the comprehensive detection index of the new version is lower than the preset standard, a version rollback operation is performed; if the comprehensive detection index of the new version is higher than or equal to the preset standard, the new version is set as the default running version.

[0054] Optionally, the feedback data includes at least one of: manual review conclusions, process handling results, and misjudgment markers.

[0055] Optionally, the scenario information includes at least one of the following: the type of object to be detected, the required accuracy of the detection, the environmental parameters of the detection, and the production batch information.

[0056] In this embodiment, since the scenario information of the current detection task is first obtained, and the target detection strategy is called from the detection strategy library to execute the detection based on the scenario information, the detection strategy can be accurately matched with the actual working conditions. This avoids the drawbacks of traditional solutions that rely on fixed preset parameters and manual experience to configure strategies, and effectively improves the scenario adaptability of strategy calls. Since the target detection results and feedback data generated by the detection are used to update the anomaly knowledge graph, fragmented information can be transformed into structured knowledge, realizing the systematic accumulation of detection experience and anomaly characteristics, fully exploring the potential value of data, and solving the problems of scattered information storage and difficulty in reusing effective experience in the prior art. Furthermore, since the target detection strategy in the detection strategy library is evaluated and iteratively updated based on the updated anomaly knowledge graph, and the detection strategy is continuously optimized based on the knowledge system formed by real detection, the probability of false detection and missed detection during long-term system operation can be significantly reduced, and the detection accuracy of membrane materials and adaptability to complex working conditions can be continuously improved. In summary, this solution achieves structured management of testing data, reusable testing experience, and dynamic autonomous optimization of testing strategies. It overcomes the shortcomings of traditional membrane material testing systems in terms of continuous optimization capabilities and can fully meet the application requirements of high precision and high adaptability in online and offline testing scenarios for industrial membrane materials.

[0057] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include a processor 310, a communications interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute a knowledge and policy collaborative update method based on detection feedback. This method includes: acquiring scene information of the current detection task; calling a target detection policy from a detection policy library based on the scene information to perform detection, obtaining target detection results and feedback data on the target detection results; updating an anomaly knowledge graph based on the target detection results and the feedback data, and evaluating and updating the target detection policies in the detection policy library based on the updated anomaly knowledge graph.

[0058] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0059] On the other hand, this application also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the knowledge and strategy collaborative update method based on detection feedback provided by the above methods. The method includes: obtaining scene information of the current detection task; calling a target detection strategy from a detection strategy library according to the scene information to perform detection, obtaining target detection results and feedback data on the target detection results; updating an anomaly knowledge graph according to the target detection results and the feedback data, and evaluating and updating the target detection strategy in the detection strategy library based on the updated anomaly knowledge graph.

[0060] In another aspect, this application also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a knowledge and strategy collaborative update method based on detection feedback provided by the above methods. The method includes: acquiring scene information of the current detection task; calling a target detection strategy from a detection strategy library according to the scene information to perform detection, obtaining target detection results and feedback data on the target detection results; updating an anomaly knowledge graph according to the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph.

[0061] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0062] Obviously, those skilled in the art should understand that the various units or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device, or fabricating them separately as individual integrated circuit modules, or fabricating multiple modules or steps into a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0063] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for collaborative updating of knowledge and strategies based on detection feedback, characterized in that, include: Obtain scene information for the current detection task; Based on the scene information, the target detection strategy is called from the detection strategy library to perform detection, and the target detection result and feedback data on the target detection result are obtained. The anomaly knowledge graph is updated based on the target detection results and the feedback data, and the target detection strategies in the detection strategy library are evaluated and updated based on the updated anomaly knowledge graph.

2. The knowledge and strategy collaborative update method based on detection feedback according to claim 1, characterized in that, The step of updating the anomaly knowledge graph based on the target detection results and the feedback data includes: Extract the anomaly type, anomaly location, and anomaly feature data from the target detection results, as well as the correction conclusions and handling-related information from the feedback data; The anomaly type, anomaly location, anomaly feature data, correction conclusion, and handling-related information are added or updated as nodes and edge relationships, and integrated into the anomaly knowledge graph to form an updated anomaly knowledge graph.

3. The knowledge and strategy collaborative update method based on detection feedback according to claim 1, characterized in that, The evaluation and updating of target detection strategies in the detection strategy library based on the updated anomaly knowledge graph includes: Based on the updated anomaly knowledge graph, determine the anomaly feature matching degree and historical execution effect association data corresponding to the current detection scenario; The detection accuracy, false positive rate, and false negative rate of the target detection strategy are quantitatively evaluated based on the abnormal feature matching degree and the historical execution effect correlation data to obtain the quantitative evaluation results. If the quantitative evaluation result reaches the preset optimization threshold, the priority of the target detection strategy is increased; if the quantitative evaluation result does not reach the preset optimization threshold, the priority of the target detection strategy is decreased, or the target detection strategy is marked as pending optimization.

4. The knowledge and strategy collaborative update method based on detection feedback according to claim 1, characterized in that, After updating the anomaly knowledge graph based on the target detection results and the feedback data, and evaluating and updating the target detection strategies in the detection strategy library based on the updated anomaly knowledge graph, the method further includes: The update process of the anomaly knowledge graph and the detection strategy library is recorded. If the overall test index of the new version is lower than the preset standard, a version rollback operation will be performed; if the overall test index of the new version is higher than or equal to the preset standard, the new version will be set as the default running version.

5. The knowledge and strategy collaborative update method based on detection feedback according to claim 1, characterized in that, The feedback data includes at least one of the following: manual review conclusions, process handling results, and misjudgment markers.

6. The knowledge and strategy collaborative update method based on detection feedback according to claim 1, characterized in that, The scenario information includes at least one of the following: the type of object to be detected, the required accuracy of the detection, the parameters of the detection environment, and the production batch information.

7. A knowledge and strategy collaborative update device based on detection feedback, characterized in that, include: Acquisition module and processing module; The acquisition module is used to acquire scene information of the current detection task; The processing module is used to call a target detection strategy from the detection strategy library according to the scene information to perform detection, and obtain the target detection result and feedback data on the target detection result; The anomaly knowledge graph is updated based on the target detection results and the feedback data, and the target detection strategies in the detection strategy library are evaluated and updated based on the updated anomaly knowledge graph.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the knowledge and policy collaborative update method based on detection feedback as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the knowledge and strategy collaborative update method based on detection feedback as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the knowledge and strategy collaborative update method based on detection feedback as described in any one of claims 1 to 6.

Images