Cloud native space asset state identification and fault location method and device based on deep learning, computer equipment, storage medium and computer program product

Abstract

The invention relates to a cloud native space asset state identification and fault location method and device based on deep learning, computer equipment, a storage medium and a computer program product. The method comprises the following steps: acquiring an evaluation index set of space assets and a plurality of initial historical nodes; normal historical nodes and fault historical nodes are screened out based on the evaluation index set, and an evaluation weight node set is generated according to the evaluation index set and the fault historical nodes; monitoring is executed according to preset first monitoring times and a first monitoring time period, a plurality of first monitoring node sets are obtained and integrated into a first comprehensive node, a first comprehensive evaluation value is calculated in combination with the evaluation weight node set, and the asset state is confirmed; and under the condition that the asset state is a fault state, performing dense monitoring according to a second monitoring frequency greater than the first monitoring frequency, identifying an abnormal node according to the reference node set, and determining an actual fault point by using the abnormal node and a pre-constructed positioning method. By adopting the method, the complex fault mode caused by multi-index linkage can be accurately identified.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a cloud-native space asset state recognition and fault positioning method and device based on deep learning, computer equipment, storage medium and computer program product. BACKGROUND

[0002] With the development of modern information infrastructure technology, space assets have become the key carrier to support network service operation, and the related state monitoring and fault diagnosis technology has gradually become the core technology direction to ensure the reliability of information infrastructure. Such technology needs to capture space asset operation data in real time, identify abnormal states and locate faults in a timely manner to avoid network service interruption, data loss and other problems caused by equipment failure. Its core feature is to adapt to the complex operating environment of high load and multi-parameter linkage of space assets, while meeting the real-time and accuracy requirements of fault diagnosis.

[0003] In traditional technology, the fault diagnosis of space assets mainly relies on the combination of manual inspection and static threshold alarm: on the one hand, through regular on-site inspection or remote viewing of device operation logs by operation and maintenance personnel, it is determined whether the asset has problems such as hardware aging and connection abnormalities; on the other hand, a fixed index threshold is preset, and an alarm is triggered when a device exceeds the threshold, so as to determine that the asset is in a fault or abnormal state.

[0004] However, the current traditional processing method has obvious technical defects: first, the alarm mechanism based on fixed threshold lacks flexibility and cannot adapt to the dynamic changes of business load and network environment of space assets, which easily leads to frequent false alarms or missed alarms; second, the traditional method usually only focuses on a single operating index, ignoring the correlation between multiple indexes, making it difficult to accurately identify complex fault patterns caused by multi-index linkage. SUMMARY

[0005] Therefore, it is necessary to provide a cloud-native space asset state recognition and fault positioning method, device, computer equipment, storage medium and computer program product based on deep learning, which can accurately identify complex fault patterns caused by multi-index linkage.

[0006] In a first aspect, the present application provides a cloud-native space asset state recognition and fault positioning method based on deep learning, which comprises:

[0007] obtaining an evaluation index set of a space asset and a plurality of initial historical nodes, the initial historical nodes comprising a plurality of single-index nodes recording index names and index values;

[0008] screening normal historical nodes and fault historical nodes based on the set of evaluation indexes, constructing a set of reference nodes in combination with the set of evaluation indexes, the normal historical nodes and a deep learning method, and generating a set of evaluation weight nodes according to the set of evaluation indexes and the fault historical nodes;

[0009] performing monitoring according to a preset first monitoring frequency and a first monitoring period, obtaining a plurality of first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive evaluation value in combination with the set of evaluation weight nodes and confirming the asset state;

[0010] in the case where the asset state is a fault state, performing intensive monitoring according to a second monitoring frequency greater than the first monitoring frequency, identifying abnormal nodes by comparing with the set of reference nodes, and confirming an actual fault point by using the abnormal nodes and a pre-constructed positioning method.

[0011] In some embodiments of the method, the set of evaluation indexes includes at least one evaluation index of operating parameters and loss data, and each evaluation index corresponds to a unique identifier for distinguishing different state dimensions.

[0012] In some embodiments of the method, the deep learning method includes a pre-trained neural network model, and the set of reference nodes includes a plurality of reference nodes for recording corresponding evaluation indexes and respective evaluation index reference intervals.

[0013] In some embodiments of the method, the monitoring according to a preset first monitoring frequency and a first monitoring period, obtaining a plurality of first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive evaluation value in combination with the set of evaluation weight nodes and confirming the asset state, includes:

[0014] calculating the mean and peak values of each evaluation index in a single monitoring period, and recording the mean and peak values as first comprehensive index values in the corresponding first comprehensive node;

[0015] weighting the first comprehensive index values according to the evaluation weights corresponding to each evaluation index to obtain a first comprehensive evaluation value;

[0016] in the case where the first comprehensive evaluation value is within a preset normal threshold range, determining that the asset state is a normal state; otherwise, determining that the asset state is a fault state.

[0017] In some embodiments of the method, in the case where the asset state is a fault state, performing intensive monitoring according to a second monitoring frequency greater than the first monitoring frequency, identifying abnormal nodes by comparing with the set of reference nodes, and confirming an actual fault point by using the abnormal nodes and a pre-constructed positioning method, includes:

[0018] In the case of identifying the abnormal node, the index data obtained by the intensive monitoring is compared with the evaluation index reference interval corresponding to the reference node set, and the monitoring node whose index data exceeds the evaluation index reference interval is selected as the abnormal node.

[0019] In some embodiments of the method, the pre-constructed positioning method includes at least one of a fault tree analysis method and a neural network positioning model, and the actual fault point is confirmed by using the abnormal node and the pre-constructed positioning method, including:

[0020] The abnormal node is matched with the fault historical node, and the candidate fault point is obtained by sorting the matching degree. The candidate fault point is verified by hardware checking, configuration checking and network testing, and the candidate fault point is dynamically updated until the actual fault point is confirmed.

[0021] According to a second aspect of the embodiments of the present disclosure, a cloud native space asset state identification and fault positioning device based on deep learning is provided. The device includes:

[0022] The first module is configured to obtain an evaluation index set of a space asset and a plurality of initial historical nodes, and the initial historical nodes include a plurality of single-index nodes recording index names and index values.

[0023] The second module is configured to filter out normal historical nodes and fault historical nodes based on the evaluation index set, construct a reference node set in combination with the evaluation index set, the normal historical nodes and a deep learning method, and generate an evaluation weight node set according to the evaluation index set and the fault historical nodes.

[0024] The third module is configured to perform monitoring according to a preset first monitoring number and a first monitoring period, obtain a plurality of first monitoring node sets and integrate them into a first comprehensive node, calculate a first comprehensive evaluation value in combination with the evaluation weight node set, and confirm the asset state.

[0025] The fourth module is configured to, in the case that the asset state is a fault state, perform intensive monitoring according to a second monitoring number greater than the first monitoring number, identify an abnormal node by comparing with the reference node set, and confirm an actual fault point by using the abnormal node and a pre-constructed positioning method.

[0026] According to a third aspect of the embodiments of the present disclosure, a computer device is provided. The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0027] The evaluation index set of the space asset and the plurality of initial historical nodes are obtained, and the initial historical nodes include a plurality of single-index nodes recording index names and index values.

[0028] screening normal historical nodes and fault historical nodes based on the evaluation index set, constructing a reference node set in combination of the evaluation index set, the normal historical nodes and a deep learning method, and generating an evaluation weight node set according to the evaluation index set and the fault historical nodes;

[0029] performing monitoring according to a preset first monitoring number and a first monitoring period, obtaining a plurality of first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive evaluation value in combination of the evaluation weight node set and confirming the asset state;

[0030] in a case where the asset state is a fault state, performing intensive monitoring according to a second monitoring number greater than the first monitoring number, identifying abnormal nodes by comparing with the reference node set, and confirming an actual fault point by using the abnormal nodes and a pre-constructed positioning method.

[0031] According to a fourth aspect of the embodiments of the present disclosure, a computer readable storage medium is provided. The computer readable storage medium has a computer program stored thereon, and the computer program is executed by a processor to implement the following steps:

[0032] obtaining an evaluation index set and a plurality of initial historical nodes of a spatial asset, the initial historical nodes including a plurality of single-index nodes recording index names and index values;

[0033] screening normal historical nodes and fault historical nodes based on the evaluation index set, constructing a reference node set in combination of the evaluation index set, the normal historical nodes and a deep learning method, and generating an evaluation weight node set according to the evaluation index set and the fault historical nodes;

[0034] performing monitoring according to a preset first monitoring number and a first monitoring period, obtaining a plurality of first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive evaluation value in combination of the evaluation weight node set and confirming the asset state;

[0035] in a case where the asset state is a fault state, performing intensive monitoring according to a second monitoring number greater than the first monitoring number, identifying abnormal nodes by comparing with the reference node set, and confirming an actual fault point by using the abnormal nodes and a pre-constructed positioning method.

[0036] According to a fifth aspect of the embodiments of the present disclosure, a computer program product is provided. The computer program product includes a computer program, and the computer program is executed by a processor to implement the following steps:

[0037] obtaining an evaluation index set and a plurality of initial historical nodes of a spatial asset, the initial historical nodes including a plurality of single-index nodes recording index names and index values;

[0038] screening normal historical nodes and fault historical nodes based on the evaluation index set, constructing a reference node set in combination with the evaluation index set, the normal historical nodes and a deep learning method, and generating an evaluation weight node set according to the evaluation index set and the fault historical nodes;

[0039] performing monitoring according to a preset first monitoring number and a first monitoring period, obtaining a plurality of first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive evaluation value in combination with the evaluation weight node set and confirming the asset state;

[0040] In the case where the asset state is a fault state, performing intensive monitoring according to a second monitoring number greater than the first monitoring number, identifying abnormal nodes by comparing with the reference node set, and confirming an actual fault point by using the abnormal nodes and a pre-constructed positioning method.

[0041] The cloud native space asset state identification and fault positioning scheme based on deep learning provided by the embodiments of the present application breaks the limitations of traditional methods relying on fixed thresholds and single indicators through systematic data analysis and hierarchical monitoring strategies. With the help of deep learning to construct the reference node set and the evaluation weight node set, adaptive adaptation to dynamically changing business loads and network environments is achieved, effectively reducing false positives and false negatives. At the same time, through multi-index comprehensive evaluation and fault positioning process, complex fault modes caused by multi-index linkage can be accurately identified, significantly improving the accuracy of space asset state identification and the precision of fault positioning, providing reliable support for asset maintenance.

[0042] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0043] The accompanying drawings incorporated into the specification and forming a part thereof illustrate embodiments consistent with the present disclosure and together with the description serve to explain the principles of the present disclosure, and do not constitute an undue limitation on the present disclosure.

[0044] Figure 1 A flowchart of a cloud native space asset state identification and fault positioning method based on deep learning according to an exemplary embodiment is shown;

[0045] Figure 2 A flowchart of an asset state confirmation step according to an exemplary embodiment is shown;

[0046] Figure 3 A block diagram of a cloud native space asset state identification and fault positioning apparatus based on deep learning according to an exemplary embodiment is shown;

[0047] Figure 4 An internal structure diagram of a computer device according to an exemplary embodiment is shown. DETAILED DESCRIPTION

[0048] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0049] It should be noted that the terms "first", "second", and the like in the description and claims of the present disclosure and the above-described drawings are used to distinguish similar objects, and do not necessarily have to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Rather, they are merely examples of devices and methods consistent with some aspects of the present disclosure. The terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, product or equipment including a series of elements includes not only those elements, but also other elements not explicitly listed or further includes elements inherent to such process, method, product or equipment. Without more limitations, it does not exclude the presence of other same or equivalent elements in the process, method, product or equipment including the described elements. For example, if the first, second, etc. terms are used to represent names, they do not mean any particular order.

[0050] In some embodiments of the present disclosure, a cloud-native space asset state recognition and fault location method based on deep learning is provided, as shown in Figure 1 The method comprises the following steps:

[0051] S20, obtaining an evaluation index set of a space asset and a plurality of initial historical nodes, the initial historical nodes comprising a plurality of single-index nodes recording index names and index values.

[0052] Space assets generally refer to various network devices, servers, storage devices and other hardware resources supporting the operation of modern information infrastructure, as well as software systems and network components matched therewith, which are the core carriers to ensure the continuous and stable operation of network services. The evaluation index set is a key data set reflecting the running state of the space asset, covering various key information in the asset operation process, which can fully depict the working state of the asset.

[0053] The initial historical node is a basic unit for recording past operation data of the space asset, and each initial historical node contains a plurality of single-index nodes. The single-index node is the smallest data unit for recording specific information of a single evaluation index, and completely retains the index name and corresponding numerical value to provide original data support for subsequent analysis.

[0054] S22. Based on the evaluation index set, normal historical nodes and faulty historical nodes are selected. A reference node set is constructed by combining the evaluation index set, normal historical nodes and deep learning methods. An evaluation weight node set is generated based on the evaluation index set and faulty historical nodes.

[0055] Normal historical nodes are selected from the initial historical nodes and reflect that the space assets are in normal operating condition. Faulty historical nodes are node data that reflect that the assets have malfunctioned or are in abnormal operating condition. Together, they constitute the historical data benchmark.

[0056] The reference node set is a standard dataset constructed based on normal historical nodes and deep learning methods, providing a clear basis for determining whether current asset indicators are normal. The evaluation weight node set is a weighted set that reflects the degree of impact of different evaluation indicators on the fault, used to quantify the importance of each indicator in the status assessment.

[0057] The evaluation weight node set contains multiple evaluation weight nodes, each including an evaluation weight and an evaluation indicator. Each evaluation weight node corresponds one-to-one with an evaluation indicator. Weights are dynamically allocated based on the actual anomalies of each evaluation indicator in historical failure nodes. The number of anomalies and the number of healthy nodes for each evaluation indicator are incorporated into the evaluation weight calculation. This allows the evaluation weight to more accurately reflect the actual importance of the corresponding evaluation indicator in server status evaluation. A flexible evaluation weight acquisition scheme is used to determine if there are nodes with non-zero anomalies, adjusting the weight calculation method accordingly. Evaluation indicators with zero anomalies are assigned the minimum weight, ensuring that all evaluation indicators have a certain influence. This method of obtaining evaluation weights through statistical analysis of historical failure nodes accurately reflects the historical contribution of each evaluation indicator in status evaluation, highlighting the impact of key evaluation indicators on status evaluation, and further improving the accuracy and efficiency of status evaluation.

[0058] S24. Perform monitoring according to the preset first monitoring number and first monitoring period, obtain multiple first monitoring node sets and integrate them into a first comprehensive node, calculate the first comprehensive evaluation value by combining the evaluation weight node set and confirm the asset status.

[0059] The first monitoring node set is the set of real-time data collected during the initial monitoring process. The first comprehensive node is a comprehensive data unit formed by integrating and processing the monitoring data within a single monitoring cycle. The first comprehensive evaluation value is the core value used to determine the asset status, calculated by combining evaluation weights.

[0060] Each first comprehensive node includes a first comprehensive index value and an evaluation index. By adaptively selecting a dual-mode calculation strategy of fine-tuning the stable trend or weighted fusion of fluctuating data, the accuracy of the first comprehensive index value is improved. An incremental weighting mechanism is adopted to give higher weight to the recent first monitoring node, so that the generated first comprehensive node reflects the current status of the system in real time and improves the accuracy of status assessment.

[0061] S26. When the asset is in a fault state, perform intensive monitoring with a second monitoring number greater than the first monitoring number, identify abnormal nodes by comparing with the reference node set, and confirm the actual fault point by using the abnormal nodes and the pre-built positioning method.

[0062] An abnormal node is a monitoring node whose indicator data exceeds the reference standard during intensive monitoring under fault conditions. The actual fault point is the specific component, software module, or configuration problem that caused the asset to fail.

[0063] In some embodiments of this disclosure, a systematic data analysis and hierarchical monitoring strategy breaks through the limitations of traditional methods that rely on fixed thresholds and single indicators. By leveraging deep learning to construct a reference node set and an evaluation weight node set, adaptive adaptation to dynamically changing service loads and network environments is achieved, effectively reducing false alarms and missed alarms. Simultaneously, through multi-indicator comprehensive evaluation and fault location processes, complex fault modes caused by the linkage of multiple indicators can be accurately identified, significantly improving the accuracy of spatial asset status identification and the precision of fault location, providing reliable support for asset maintenance.

[0064] In some embodiments of this disclosure, the set of evaluation metrics includes at least one evaluation metric from operating parameters and loss data, and each evaluation metric corresponds to a unique identifier, which is used to distinguish different state dimensions.

[0065] In some implementation methods, the operational dimensions of the space assets are first comprehensively reviewed, and assessment indicators reflecting their core status are collected to form an assessment indicator set, ensuring that the indicators cover key aspects such as hardware operation, software performance, and data transmission. Simultaneously, historical data from the asset's past operations is retrieved and organized into multiple initial historical nodes. Each initial historical node contains multiple single-indicator nodes, and each single-indicator node fully records the indicator name and corresponding indicator value, ensuring that the data is complete and error-free, laying a reliable foundation for subsequent analysis.

[0066] In some embodiments of this disclosure, the evaluation index set explicitly includes key indicators such as operating parameters and wear data, and each evaluation index is assigned a unique identifier, making the distinction between different state dimensions clearer. This design ensures the comprehensiveness and uniqueness of the evaluation indexes, avoids index confusion and omissions, and makes subsequent data analysis and state assessment more targeted and accurate, providing a standardized and unified data foundation for the entire state identification and fault location process.

[0067] In some embodiments of this disclosure, the deep learning method includes a pre-trained neural network model, the reference node set includes multiple reference nodes, and the reference nodes are used to record the corresponding evaluation index and the corresponding evaluation index reference interval.

[0068] In some implementations, based on the acquired set of evaluation indicators, initial historical nodes are analyzed and screened one by one to distinguish between normal historical nodes and faulty historical nodes. Combining the set of evaluation indicators and normal historical nodes, deep learning methods such as pre-trained neural network models are used for data analysis and calculation to determine the normal fluctuation range of each evaluation indicator, constructing a reference node set containing the evaluation indicators and their corresponding reference intervals. Simultaneously, based on the set of evaluation indicators and faulty historical nodes, the influence of different evaluation indicators during the fault occurrence process is analyzed, the abnormal frequency and influence range of each indicator are statistically analyzed, and an evaluation weight node set containing the evaluation indicators and their corresponding weights is generated to ensure that the weight allocation is consistent with the actual importance of the indicators.

[0069] In some embodiments of this disclosure, a reference node set is constructed using deep learning methods such as pre-trained neural network models, making the reference ranges of evaluation indicators in the reference node set more closely match the actual operating patterns of assets. The reference node set clearly records the evaluation indicators and their corresponding reference ranges, providing a precise and scientific basis for judging indicator anomalies, effectively improving the accuracy of anomaly identification, and laying a reliable standard foundation for subsequent status confirmation and fault location.

[0070] In some embodiments of this disclosure, reference is made to Figure 2 S24 includes:

[0071] S242. Calculate the mean and peak values ​​of each evaluation indicator within a single monitoring cycle, and record the mean and peak values ​​as the first comprehensive indicator values ​​at the corresponding first comprehensive node.

[0072] S244. The first comprehensive indicator value is calculated by weighting the first comprehensive indicator value according to the evaluation weights corresponding to each evaluation indicator.

[0073] S246. If the first comprehensive evaluation value is within the preset normal threshold range, the asset status is determined to be normal; otherwise, the asset status is determined to be faulty.

[0074] In some implementations, based on the daily operational patterns and monitoring needs of the space assets, a reasonable number of initial monitoring sessions and a specific monitoring period are set. The assets are monitored initially according to plan, and complete data for each assessment indicator is collected during each monitoring session, forming multiple sets of initial monitoring nodes. The indicator data in each initial monitoring node set are integrated and processed, and the mean and peak values ​​of each assessment indicator within a single monitoring period are calculated. These combined data are recorded as the initial comprehensive indicator value in the corresponding initial comprehensive node. Combining the weight allocation of the assessment weight node set, the initial comprehensive indicator values ​​are weighted and calculated to obtain the initial comprehensive assessment value. The initial comprehensive assessment value is compared with a preset normal threshold range to determine whether the asset status is normal or faulty.

[0075] In some embodiments of this disclosure, a first comprehensive node is constructed by calculating the mean and peak values ​​of the evaluation indicators, comprehensively reflecting the overall performance of the indicators within a single monitoring period. A weighted calculation based on evaluation weights yields the first comprehensive evaluation value, making the status judgment more reflective of the differences in importance among the indicators. This hierarchical calculation and judgment method makes the determination of asset status more scientific and accurate, effectively avoiding the one-sidedness of single-indicator judgments and improving the reliability of status identification.

[0076] In some embodiments of this disclosure, S26 includes:

[0077] When abnormal nodes are identified, the indicator data obtained from intensive monitoring is compared with the reference range of the evaluation indicators corresponding to the reference node set, and the monitoring nodes whose indicator data exceeds the reference range of the evaluation indicators are selected as abnormal nodes.

[0078] In some implementations, if an asset is determined to be in a fault state, a more intensive second monitoring process is immediately initiated, with more monitoring sessions than the first, to more accurately capture indicator fluctuations. The indicator data obtained from this intensive monitoring is compared one by one with the corresponding evaluation indicator reference ranges in the reference node set. Monitoring nodes exceeding the reference range are identified as anomalous nodes. Using pre-built localization methods such as fault tree analysis and neural network localization models, anomalous nodes are matched with historical fault nodes, and candidate fault points are obtained by ranking them according to their matching degree. These candidate fault points are verified one by one through hardware checks, configuration verification, and network testing. The candidate fault point list is dynamically updated based on the verification results until the actual fault point is accurately confirmed.

[0079] In some embodiments of this disclosure, more intensive secondary monitoring is initiated under fault conditions, enabling more detailed capture of indicator fluctuations. By comparing and filtering abnormal nodes with the evaluation indicator reference range of the reference node set, the accuracy of abnormal node identification is ensured, providing an accurate direction for subsequent fault location, reducing ineffective troubleshooting work, and improving fault handling efficiency.

[0080] In some embodiments of this disclosure, the pre-built localization method includes at least one of fault tree analysis and neural network localization model, and S26 includes:

[0081] The abnormal nodes are matched with the historical fault nodes and sorted by matching degree to obtain candidate fault points. The candidate fault points are verified by hardware inspection, configuration verification and network testing. The candidate fault points are dynamically updated until the actual fault point is confirmed.

[0082] In some embodiments of this disclosure, pre-constructed fault tree analysis, neural network localization models, and other localization methods provide diverse and efficient technical support for fault point identification. By matching and sorting abnormal nodes with historical fault nodes and verifying candidate fault points from multiple dimensions, rapid identification and accurate confirmation of fault points are achieved, avoiding blind investigation, significantly shortening fault location time, improving fault repair efficiency, and reducing the impact of asset failures on business operations.

[0083] This disclosure presents several cloud-native spatial asset status identification and fault location methods based on deep learning. Through systematic data analysis and hierarchical monitoring strategies, these methods overcome the limitations of traditional approaches that rely on fixed thresholds and single indicators. By leveraging deep learning to construct reference node sets and evaluation weight node sets, they achieve adaptive adaptation to dynamically changing business loads and network environments, effectively reducing false positives and false negatives. Furthermore, through multi-indicator comprehensive evaluation and fault location processes, they can accurately identify complex fault modes caused by the interaction of multiple indicators, significantly improving the accuracy of spatial asset status identification and the precision of fault location, providing reliable support for asset maintenance.

[0084] It is understood that the various embodiments of the methods described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. Related details can be found in the descriptions of other method embodiments.

[0085] It should be understood that although the steps in the flowcharts shown in the accompanying drawings are displayed sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some of the steps in the accompanying drawings may include multiple steps or stages, which are not necessarily completed at the same time, but may be executed at different times, and the execution order of these steps or stages is not necessarily sequential, but may be performed alternately or in turn with other steps or at least a portion of the steps or stages of other steps.

[0086] Based on the description of the above-described embodiments of the deep learning-based cloud-native spatial asset status identification and fault location method, this disclosure also provides a deep learning-based cloud-native spatial asset status identification and fault location device for implementing the aforementioned deep learning-based cloud-native spatial asset status identification and fault location method. The device may include a system (including a distributed system), software (application), module, component, controller, server, terminal, etc., using the method described in the embodiments of this specification, combined with necessary implementation hardware. Based on the same innovative concept, the devices in one or more embodiments provided in this disclosure are as described in the following embodiments. Since the implementation schemes and methods for solving the problem by the devices are similar, the implementation of specific devices in the embodiments of this specification can refer to the implementation of the aforementioned method, and repeated details will not be repeated. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0087] Figure 3 This is a schematic block diagram illustrating a cloud-native spatial asset status identification and fault location device based on deep learning, according to an exemplary embodiment. The device can be the aforementioned terminal, a server, or a module, component, device, control unit, etc., integrated into the terminal. For details, please refer to... Figure 3 The device 100 may include: a first module 120, a second module 140, a third module 160, and a fourth module 180. The first module 120 is used to acquire an evaluation index set for spatial assets and multiple initial historical nodes, wherein the initial historical nodes include multiple single-index nodes recording index names and values. The second module 140 is used to filter out normal historical nodes and faulty historical nodes based on the evaluation index set, construct a reference node set by combining the evaluation index set, normal historical nodes, and deep learning methods, and generate an evaluation weight node set based on the evaluation index set and faulty historical nodes. The third module 160 is used to perform monitoring according to a preset first monitoring number and a first monitoring period, acquire multiple first monitoring node sets and integrate them into a first comprehensive node, calculate a first comprehensive evaluation value by combining the evaluation weight node set, and confirm the asset status. The fourth module 180 is used to perform intensive monitoring at a second monitoring number greater than the first monitoring number when the asset status is faulty, identify abnormal nodes by comparing with the reference node set, and confirm the actual fault point using the abnormal nodes and a pre-built positioning method.

[0088] In some embodiments of the device, the evaluation index set includes at least one evaluation index from operating parameters and loss data, and each evaluation index corresponds to a unique identifier, which is used to distinguish different state dimensions.

[0089] In some embodiments of the device, the deep learning method includes a pre-trained neural network model, the reference node set includes multiple reference nodes, and the reference nodes are used to record the corresponding evaluation index and the corresponding evaluation index reference interval.

[0090] In some embodiments of the device, the third module 160 is further used to calculate the mean and peak values ​​of each evaluation indicator within a single monitoring cycle, and record the mean and peak values ​​as the first comprehensive indicator values ​​at the corresponding first comprehensive node; perform weighted calculation on the first comprehensive indicator values ​​according to the evaluation weights corresponding to each evaluation indicator to obtain the first comprehensive evaluation value; if the first comprehensive evaluation value is within the preset normal threshold range, determine that the asset status is normal; otherwise, determine that the asset status is faulty.

[0091] In some embodiments of the device, the fourth module 180 is further configured to, when an abnormal node is identified, compare the indicator data obtained by dense monitoring with the evaluation indicator reference range corresponding to the reference node set, and screen out the monitoring nodes whose indicator data exceeds the evaluation indicator reference range as abnormal nodes.

[0092] In some embodiments of the device, the fourth module 180 is further configured to match abnormal nodes with fault history nodes and sort them by matching degree to obtain candidate fault points, verify the candidate fault points through hardware checks, configuration checks and network tests, and dynamically update the candidate fault points until the actual fault point is confirmed.

[0093] The modules in the aforementioned deep learning-based cloud-native spatial asset status identification and fault location device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0094] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a deep learning-based cloud-native spatial asset status identification and fault location method.

[0095] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0096] Based on the foregoing description of the relevant methods and apparatus embodiments, this disclosure also provides a computer device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the deep learning-based cloud-native spatial asset status identification and fault location method described in any embodiment of this specification.

[0097] Based on the foregoing description of the relevant methods and apparatus embodiments, this disclosure also provides a computer-readable storage medium that, when the instructions in the computer-readable storage medium are executed by the processor of a computer device, enables the computer device to implement the deep learning-based cloud-native spatial asset status identification and fault location method as described in any embodiment of this disclosure.

[0098] Based on the foregoing description of the relevant methods and apparatus embodiments, this disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the deep learning-based cloud-native spatial asset status identification and fault location method described in any embodiment of this specification.

[0099] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, hardware + program embodiments are relatively simple in description because they are fundamentally similar to method embodiments; relevant parts can be referred to the descriptions in the method embodiments.

[0100] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0101] It should be noted that the apparatus, computer equipment, storage medium, and computer program products described above may also include other implementation methods according to the description of the method embodiments. Specific implementation methods can be found in the description of the relevant method embodiments. Furthermore, new embodiments formed by combinations of features from various methods, apparatuses, devices, and server embodiments still fall within the scope of this disclosure and will not be elaborated upon here.

[0102] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, when implementing one or more of these specifications, the functions of each module can be implemented in the same or different software and / or hardware, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are merely illustrative. For example, the division of modules or units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling and communication connections between the devices or units shown or described can be implemented through direct and / or indirect coupling / connection, through standard or custom interfaces or protocols, and can be implemented electrically, mechanically, or in other forms.

[0103] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This disclosure is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0104] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A cloud-native spatial asset status identification and fault location method based on deep learning, characterized in that, The method includes: Acquire a set of evaluation indicators for spatial assets and multiple initial historical nodes, wherein the initial historical nodes include multiple single indicator nodes that record indicator names and indicator values; Based on the evaluation index set, normal historical nodes and faulty historical nodes are selected. A reference node set is constructed by combining the evaluation index set, normal historical nodes and deep learning methods. An evaluation weight node set is generated based on the evaluation index set and faulty historical nodes. Monitoring is performed according to the preset first monitoring number and first monitoring period, multiple first monitoring node sets are obtained and integrated into a first comprehensive node, and the first comprehensive assessment value is calculated by combining the assessment weight node set and the asset status is confirmed. When the asset is in a fault state, intensive monitoring is performed at a second number of monitoring times, which is greater than the first number of monitoring times. Abnormal nodes are identified by comparing with the reference node set, and the actual fault point is confirmed by using the abnormal nodes and the pre-built localization method.

2. The method according to claim 1, characterized in that, The evaluation index set includes at least one evaluation index from operating parameters and loss data. Each evaluation index corresponds to a unique identifier, which is used to distinguish different state dimensions.

3. The method according to claim 2, characterized in that, The deep learning method includes a pre-trained neural network model, and the reference node set includes multiple reference nodes, which are used to record the corresponding evaluation index and the corresponding evaluation index reference interval.

4. The method according to claim 1, characterized in that, The process of performing monitoring according to a preset first monitoring number and first monitoring period, acquiring multiple first monitoring node sets and integrating them into a first comprehensive node, calculating a first comprehensive assessment value by combining the assessment weight node set, and confirming the asset status includes: Calculate the mean and peak values ​​of each evaluation indicator within a single monitoring period, and record the mean and peak values ​​as the first comprehensive indicator values ​​at the corresponding first comprehensive node; The first comprehensive indicator value is obtained by weighting the values ​​of the first comprehensive indicator according to the evaluation weights corresponding to each evaluation indicator. If the first comprehensive assessment value is within the preset normal threshold range, the asset status is determined to be normal; otherwise, the asset status is determined to be faulty.

5. The method according to claim 1, characterized in that, When the asset is in a faulty state, intensive monitoring is performed at a second monitoring frequency greater than the first monitoring frequency. Abnormal nodes are identified by comparing with a reference node set, and the actual fault point is confirmed using the abnormal nodes and a pre-built localization method, including: When abnormal nodes are identified, the indicator data obtained from intensive monitoring is compared with the reference range of the evaluation indicators corresponding to the reference node set, and the monitoring nodes whose indicator data exceeds the reference range of the evaluation indicators are selected as abnormal nodes.

6. The method according to claim 1, characterized in that, The pre-built localization method includes at least one of fault tree analysis and neural network localization model. The step of using abnormal nodes and the pre-built localization method to identify the actual fault point includes: The abnormal nodes are matched with the historical fault nodes and sorted by matching degree to obtain candidate fault points. The candidate fault points are verified by hardware inspection, configuration verification and network testing. The candidate fault points are dynamically updated until the actual fault point is confirmed.

7. A cloud-native spatial asset status identification and fault location device based on deep learning, characterized in that, The device includes: The first module is used to acquire the set of evaluation indicators for spatial assets and multiple initial historical nodes, wherein the initial historical nodes include multiple single indicator nodes that record indicator names and indicator values. The second module is used to filter out normal historical nodes and faulty historical nodes based on the evaluation index set, construct a reference node set by combining the evaluation index set, normal historical nodes and deep learning methods, and generate an evaluation weight node set based on the evaluation index set and faulty historical nodes. The third module is used to perform monitoring according to the preset first monitoring number and first monitoring period, obtain multiple first monitoring node sets and integrate them into a first comprehensive node, calculate the first comprehensive evaluation value by combining the evaluation weight node set and confirm the asset status. The fourth module is used to perform intensive monitoring at a second number of monitoring times, greater than the first number of monitoring times, when the asset is in a fault state, identify abnormal nodes by comparing with the reference node set, and confirm the actual fault point by using the abnormal nodes and the pre-built location method.

8. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, It stores a computer program thereon, which, when executed by a processor, implements the steps of the method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 6.

Images