Multi-time sequence anomaly diagnosis method, device and storage medium

Abstract

The application discloses a multi-time sequence abnormality diagnosis method, equipment and a storage medium, and belongs to the technical field of industrial monitoring. The multi-time sequence abnormality diagnosis method comprises the following steps: a monitoring device is controlled to monitor an industrial operation device to obtain multi-source time sequence signals, a current fluctuation propagation path and an abnormal sequence group are determined according to the multi-source time sequence signals and a device characteristic fluctuation conduction map, and an abnormality diagnosis result is determined according to the current fluctuation propagation path, the abnormal sequence group and a preset discrimination index, thereby achieving the beneficial effect of improving the accuracy of device abnormality identification.

Description

Technical Field

[0001] This invention relates to the field of industrial monitoring technology, and in particular to a multi-time-series anomaly diagnosis method, device and storage medium. Background Technology

[0002] In the field of industrial automation, sensors and equipment scattered throughout a factory converge in a central control room, where all operational data converges. The control room needs to monitor and analyze the status of the equipment within the factory. Currently, the conventional approach is to set thresholds for equipment operating parameters to determine whether the equipment is operating normally or abnormally. However, this method relies solely on whether instantaneous values ​​exceed the thresholds, meaning even a random fluctuation in a sensor can trigger an alarm. In practice, to reduce false alarms, engineers often have to relax the thresholds, which reduces the system's sensitivity. Consequently, equipment anomalies cannot be accurately and promptly identified.

[0003] The above content is only used to help understand the technical solution of the present invention and does not represent an admission that the above content is prior art. Summary of the Invention

[0004] The main objective of this invention is to provide a multi-temporal anomaly diagnosis method, device, and storage medium, aiming to improve the accuracy of device anomaly identification. To achieve the above objective, this invention provides a multi-temporal anomaly diagnosis method, which includes the following steps: The control and monitoring equipment monitors industrial operating equipment and obtains multi-source time-series signals; The current wave propagation path and abnormal sequence group are determined based on the multi-source time-series signal and the device characteristic wave propagation diagram. The abnormal diagnosis result is determined based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

[0005] Optionally, before the step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram, the method further includes: The historical fluctuation characteristics of each of the industrial operating devices are determined based on historical multi-source time-series signals. The direction and time delay data of fluctuation changes among the industrial operating equipment are determined based on the historical fluctuation characteristics of each of the industrial operating equipment. The device characteristic fluctuation transmission diagram is constructed based on the device connection relationship, the fluctuation change direction, and the time delay data.

[0006] Optionally, the multi-source time-series signal includes: equipment status data for each industrial operating device, and the step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the equipment characteristic fluctuation transmission diagram includes: Determine the fluctuation characteristics and corresponding fluctuation times of each piece of industrial operating equipment based on the equipment status data of each piece of equipment. The real-time device status fluctuation relationship is determined based on the fluctuation characteristics of each device and the corresponding fluctuation time, and the current fluctuation propagation path is determined based on the real-time device status fluctuation relationship; When any of the industrial operating devices has an abnormal fluctuation characteristic, the abnormal sequence group is determined based on the current fluctuation propagation path and the abnormal characteristic.

[0007] Optionally, the step of determining the abnormal sequence group based on the current wave propagation path and the abnormal characteristics includes: The abnormal triggering device and the abnormal triggering time are determined based on the abnormal characteristics. The set of affected devices is determined based on the device characteristic fluctuation transmission diagram and the abnormal triggering devices; The timing relationship is determined based on the fluctuation time of each device in the affected device set and the anomaly trigger time; The abnormal sequence group is determined based on the temporal relationship.

[0008] Optionally, the step of determining the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index includes: A path consistency index is determined based on the current fluctuation propagation path and the set of historical regular paths. Determine the path embedding deviation index based on the current wave propagation path and the path embedding benchmark; The abnormal sequence matching index is determined based on the abnormal sequence group and the frequent sequence pattern library. The anomaly diagnosis result is determined based on the path consistency index, the path embedding deviation index, and the anomaly sequence matching index.

[0009] Optionally, the step of determining the anomaly diagnosis result based on the path consistency index, the path embedding deviation index, and the anomaly sequence matching index includes: A comprehensive score is determined based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index. The abnormal diagnosis result is determined based on the comprehensive score and the preset threshold.

[0010] Optionally, the step of determining the comprehensive score based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index includes: The first weighting coefficient is determined based on the path consistency index; The second weighting coefficient is determined based on the path embedding deviation index; The third weighting coefficient is determined based on the abnormal sequence matching index; The comprehensive score is obtained by weighting and summing the path consistency index, the path embedding deviation index, the abnormal sequence matching index, the first weight coefficient, the second weight coefficient, and the third weight coefficient.

[0011] Furthermore, to achieve the above objectives, the present invention also provides a multi-temporal anomaly diagnostic device, the multi-temporal anomaly diagnostic device comprising: The acquisition module is used to control the monitoring equipment to monitor industrial operating equipment and obtain multi-source time-series signals; The analysis module is used to determine the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram; The diagnostic module is used to determine the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

[0012] In addition, to achieve the above objectives, the present invention also provides a multi-timing anomaly diagnostic device, the multi-timing anomaly diagnostic device comprising: a memory, a processor, and a multi-timing anomaly diagnostic program stored in the memory and executable on the processor, the multi-timing anomaly diagnostic program being configured to implement the steps of the multi-timing anomaly diagnostic method described above.

[0013] In addition, to achieve the above objectives, the present invention also provides a storage medium storing a multi-timing anomaly diagnostic program, wherein when the multi-timing anomaly diagnostic program is executed by a processor, it implements the steps of the multi-timing anomaly diagnostic method described above.

[0014] This invention proposes a multi-time-series anomaly diagnosis method. This method monitors industrial operating equipment using control monitoring equipment to obtain multi-source time-series signals. Based on the multi-source time-series signals and the equipment characteristic fluctuation transmission diagram, the current fluctuation propagation path and anomaly sequence group are determined. Based on the current fluctuation propagation path, the anomaly sequence group, and preset discrimination indicators, the anomaly diagnosis result is determined. Compared with traditional anomaly identification methods that set thresholds, this method can jointly determine anomalies through multiple anomaly sequence groups without relaxing the monitoring threshold, thereby improving the accuracy of identifying abnormal equipment. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of the structure of a multi-time-series anomaly diagnosis device for the hardware operating environment involved in the embodiments of the present invention; Figure 2 This is a flowchart illustrating the first embodiment of the multi-temporal anomaly diagnosis method of the present invention; Figure 3This is a flowchart illustrating the second embodiment of the multi-temporal anomaly diagnosis method of the present invention; Figure 4 This is a flowchart illustrating the third embodiment of the multi-temporal anomaly diagnosis method of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0016] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0017] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a multi-time-series anomaly diagnosis device for the hardware operating environment involved in the embodiments of the present invention.

[0018] like Figure 1 As shown, the multi-time-series anomaly diagnostic device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, an interaction device 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The interaction device 1003 may include a display screen or an input unit such as a keyboard. Optionally, the interaction device 1003 may also be connected to the communication bus via a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0019] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on multi-time-series anomaly diagnostic devices and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0020] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a data storage module, a network communication module, a user interface module, and a multi-timing anomaly diagnostic program.

[0021] exist Figure 1In the multi-time-series anomaly diagnostic device shown, the network interface 1004 is mainly used for data communication with other devices; the interactive device 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the multi-time-series anomaly diagnostic device of the present invention can be set in the multi-time-series anomaly diagnostic device, and the multi-time-series anomaly diagnostic device calls the multi-time-series anomaly diagnostic program stored in the memory 1005 through the processor 1001 and executes the multi-time-series anomaly diagnostic method provided in the embodiment of the present invention.

[0022] This invention provides a method for diagnosing multi-temporal anomalies, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of a multi-temporal anomaly diagnosis method according to the present invention.

[0023] In this embodiment, the multi-temporal anomaly diagnosis method includes: Step S1: Control the monitoring equipment to monitor the industrial operating equipment and obtain multi-source time-series signals; In this embodiment, the industrial operating equipment is not limited, and different equipment can be used for different production scenarios. Furthermore, there are multiple pieces of industrial operating equipment, and the monitoring equipment consists of sensors, specifically vibration sensors, temperature sensors, current sensors, and pressure transmitters. This allows for the acquisition of complete data on the production process of each piece of equipment. It should be noted that in the field of industrial automation, this data is typically collected in a central control room, where administrators directly monitor and control the equipment. Commonly, in industrial production processes involving motors, physical quantities such as bearing vibration intensity, winding temperature rise, motor current, and pipeline pressure can be collected separately. Since there are multiple industrial operating devices distributed across different process stages, the collected multi-source time-series signals constitute a monitoring data stream covering the entire production process, including the evolution of the operating status of individual devices and implicitly containing process correlations and energy transfer information between devices.

[0024] Step S2: Determine the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram; In this embodiment, the fluctuations of various features in the multi-source time-series signal are extracted, such as temperature fluctuation data of the heating equipment and vibration fluctuation data of the fan. The direction of wave propagation is determined by comparing the frequency, phase difference, and amplitude of the temperature and vibration fluctuation data. Based on the direction of wave propagation, the corresponding current wave propagation path is determined in the device feature wave propagation diagram. It should be noted that since the fluctuations of various features in the multi-source time-series signal are extracted, the fluctuation data is not limited to the temperature fluctuation data of the heating equipment and the vibration fluctuation data of the fan; it can also include other types of fluctuation data. Furthermore, it should be noted that not all fluctuations are abnormal. When some features are identified as having a high degree of deviation, fluctuation data of some related devices within their neighborhood time can be obtained, thereby forming the abnormal sequence group.

[0025] Step S3: Determine the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

[0026] In this embodiment, anomaly diagnosis can optionally be performed using the current fluctuation propagation path, anomaly sequence groups, and preset discrimination indicators. The current fluctuation propagation path provides the spatial diffusion pattern of the anomaly in the device network, helping to locate the device and the affected downstream links; the anomaly sequence group gathers multi-dimensional feature deviation evidence of the root cause device and related devices, obtaining a spatiotemporally correlated chain of anomaly events. The preset discrimination indicators may include at least one of the following: fluctuation amplitude threshold, propagation speed anomaly degree, and multi-channel synchronization deviation degree. The system compares the statistical characteristics of the anomaly sequence group with historical indicators before the current moment to determine the anomaly type, such as mechanical failure, electrical short circuit, process blockage, or external disturbance, and traces the anomaly starting point according to the fluctuation propagation path. The final anomaly diagnosis result includes the anomaly level, fault type, list of involved devices, and suggested handling measures, realizing a closed-loop output from multi-source time-series monitoring to interpretable root cause diagnosis, providing accurate decision-making basis for the operation and maintenance personnel in the central control room.

[0027] In this embodiment, industrial operating equipment is monitored by control monitoring equipment to obtain multi-source time-series signals. The current fluctuation propagation path and abnormal sequence group are determined based on the multi-source time-series signals and the equipment characteristic fluctuation transmission diagram. The abnormal diagnosis result is determined based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index. Compared with the traditional abnormal identification method that sets a threshold, this method can jointly determine the abnormality through multiple abnormal sequence groups without relaxing the monitoring threshold, thereby improving the accuracy of identifying abnormal equipment.

[0028] Furthermore, based on the first embodiment, a second embodiment of the multi-temporal anomaly diagnosis method of the present invention is proposed. In this embodiment, reference is made to... Figure 3 Before the step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram, the method further includes: Step S201: Determine the historical fluctuation characteristics of each of the industrial operating devices based on historical multi-source time-series signals; In this embodiment, the historical multi-source time-series signal refers to data collected before the current moment. The specific data type is not limited, but the historical multi-source time-series signal and the multi-source time-series signal must have the same data type. Specifically, for each industrial operating device, the system performs channel-by-channel analysis of the historical data to extract historical fluctuation characteristics that characterize the inherent fluctuation patterns under normal operating conditions. These characteristics include time-domain statistics such as baseline mean, standard deviation, and peak-to-peak value; frequency-domain dominant frequency components and energy concentration; and time-series patterns such as periodic fluctuation amplitude and autocorrelation decay characteristics. These historical fluctuation characteristics constitute a digital fingerprint baseline for the healthy operation of the device, used to define the fluctuation boundaries under normal operating conditions and provide a reference system for subsequent identification of abnormal fluctuations deviating from the baseline.

[0029] Step S202: Determine the direction of fluctuation and time delay data between the industrial operating equipment based on the historical fluctuation characteristics of each of the industrial operating equipment; Furthermore, after obtaining the historical fluctuation characteristics, a time-series correlation algorithm based on cross-correlation analysis is used to analyze any two industrial operating devices with process connections to determine the direction of fluctuation change and time delay data. The direction of fluctuation change represents the spatial orientation of abnormal energy or disturbance transmission between devices, such as from an upstream heating device to a downstream delivery pump; the time delay data quantifies the lag time required for the fluctuation to propagate from the source device to the target device, reflecting the volumetric inertia of the process pipeline or the rigid delay of mechanical transmission. Typically, the historical fluctuation characteristics are divided into different time intervals, and the time delay data between each industrial operating device within each time interval is calculated. The probability of fluctuation change and the average fluctuation delay are statistically analyzed. The direction of fluctuation change is then determined based on the probability of fluctuation change; specifically, when the probability of fluctuation change is higher than a preset probability threshold, a corresponding direction of fluctuation change is determined. The time delay data is then determined based on the average fluctuation delay.

[0030] Step S203: Construct the device characteristic fluctuation transmission diagram based on the device connection relationship, the fluctuation change direction, and the time delay data.

[0031] In this embodiment, the equipment characteristic fluctuation transmission diagram uses each industrial operating device as a vertex. When two devices are physically connected by a process, and the fluctuation change probability calculated in step S202 is higher than a preset threshold, a directed edge is established. The direction of the edge strictly corresponds to the fluctuation change direction, representing the spatial direction of the disturbance transmission between devices. Optionally, the weight of the edge is assigned by the average fluctuation delay, which is used to quantify the typical lag time required for the fluctuation to propagate from the source device to the target device. Preferably, the fluctuation change probability can be used as the confidence level or transmission gain coefficient of the edge.

[0032] In this embodiment, the graph structure formed based on the above method not only retains the physical adjacency relationship between devices, but also embeds the temporal characteristics and statistical significance of dynamic fluctuation propagation, becoming a carrier for describing how anomalies propagate and spread in the device network. Subsequently, during real-time monitoring, the anomaly propagation path can be quickly matched, realizing the change from single-point alarm to networked anomaly identification.

[0033] Furthermore, based on the first and second embodiments, a third embodiment of the multi-temporal anomaly diagnosis method of the present invention is proposed. In this embodiment, reference is made to... Figure 4 The multi-source time-series signal includes: equipment status data for each industrial operating device; the step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the equipment characteristic fluctuation transmission diagram includes: Step S21: Determine the fluctuation characteristics and corresponding fluctuation times of each piece of industrial operating equipment based on the equipment status data of each piece of equipment. Specifically, for each device, its fluctuation characteristics are extracted using sliding window statistics or wavelet transform, including fluctuation amplitude, frequency components, and abrupt change slope, and the corresponding fluctuation moments where the fluctuation characteristics significantly deviate from the historical baseline are marked. This process realizes the transformation from continuous monitoring data to discrete fluctuation events, providing timestamped feature nodes for subsequent identification of cross-device correlations.

[0034] Step S22: Determine the real-time device status fluctuation relationship based on the fluctuation characteristics and corresponding fluctuation time of each device, and determine the current fluctuation propagation path based on the real-time device status fluctuation relationship; By comparing the chronological order and time interval of fluctuations in adjacent devices, and combining the edge directions and weight constraints defined in the device characteristic fluctuation transmission graph constructed in the second embodiment, significant correlation pairs conforming to the graph structure are selected to form real-time device state fluctuation relationships. Furthermore, this relationship is matched with the transmission graph to determine the actual flow path of abnormal energy in the device network, i.e., the current fluctuation propagation path. This path dynamically reflects the diffusion trajectory of the anomaly in the physical production line, differing from a static graph structure and possessing real-time adaptability to operating conditions.

[0035] Step S23: When the fluctuation characteristics of any of the industrial operating devices are abnormal, the abnormal sequence group is determined based on the current fluctuation propagation path and the abnormal characteristics.

[0036] In this embodiment, the degree of abnormality of the fluctuation characteristics of each device is assessed. When the fluctuation characteristic of any device exceeds a preset abnormality threshold, an abnormal sequence group construction process is triggered. First, using the abnormal characteristic as a seed, the root cause device is traced forward along the current fluctuation propagation path determined in step S22, and the affected devices are identified backward. Devices with significant temporal correlation and feature deviations are clustered into an abnormal sequence group. It should be noted that a data anomaly of a device may be caused by the fluctuation of the previous node. Therefore, it is necessary to obtain a complete abnormal sequence group, which may include some data with normal indicators.

[0037] Furthermore, the step of determining the abnormal sequence group based on the current wave propagation path and the abnormal characteristics includes: The abnormal triggering device and the abnormal triggering time are determined based on the abnormal characteristics. The set of affected devices is determined based on the device characteristic fluctuation transmission diagram and the abnormal triggering devices; The timing relationship is determined based on the fluctuation time of each device in the affected device set and the anomaly trigger time; The abnormal sequence group is determined based on the temporal relationship.

[0038] In this embodiment, for each device in the affected device set, the fluctuation time marked in step S21 needs to be extracted, and the time difference between this time and the anomaly trigger time needs to be calculated. By comparing the weights of the corresponding edges in the device characteristic fluctuation transmission graph with the time delay data, the temporal consistency of the fluctuation transmission is verified, and a subset of devices that meet the transmission time delay constraints is selected to establish a temporal relationship. This relationship quantifies the propagation speed and order of the anomaly among the devices, and is used to distinguish between normal independent fluctuations and anomaly transmission responses. Based on the verified temporal relationship, the anomaly triggering device and the selected affected devices are sorted according to the fluctuation occurrence time to construct an anomaly sequence group. This group presents the complete process of the anomaly spreading from the root cause device to related devices in the form of a time series.

[0039] Furthermore, based on any of the above embodiments, a fourth embodiment of the multi-temporal anomaly diagnosis method of the present invention is proposed. In this embodiment, the step of determining the anomaly diagnosis result based on the current fluctuation propagation path, the anomaly sequence group, and the preset discrimination index includes: A path consistency index is determined based on the current fluctuation propagation path and the set of historical regular paths. Determine the path embedding deviation index based on the current wave propagation path and the path embedding benchmark; The abnormal sequence matching index is determined based on the abnormal sequence group and the frequent sequence pattern library. The anomaly diagnosis result is determined based on the path consistency index, the path embedding deviation index, and the anomaly sequence matching index.

[0040] In this embodiment, the path consistency index compares the current fluctuation propagation path with a set of historical normal paths, calculating the node overlap rate or edge matching degree to reflect the degree of conformity between the abnormal propagation topology and known failure modes. The path embedding deviation index uses graph embedding technology to map the current path to a low-dimensional vector space, calculates its distance from the path embedding benchmark under normal operating conditions, and quantifies the degree of structural semantic deviation. The abnormal sequence matching index performs a time-series comparison of abnormal sequence groups with a frequent sequence pattern library to assess the similarity with historical typical failure sequences. These three indices characterize anomalies from the perspectives of topological structure, geometric embedding, and temporal evolution, respectively, and together constitute a complementary set of discriminative evidence, providing multi-dimensional feature support for fusion decision-making.

[0041] Furthermore, the step of determining the abnormal diagnosis result based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index includes: A comprehensive score is determined based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index. The abnormal diagnosis result is determined based on the comprehensive score and the preset threshold.

[0042] In this embodiment, three indicators are normalized and weighted to generate a comprehensive score, using a single scalar to uniformly quantify the overall significance of the anomaly. The comprehensive score is then compared to a preset threshold, which is offline tuned based on the field's false alarm tolerance, missed alarm risk, and equipment criticality level. Critical equipment uses a lower threshold to improve sensitivity. When the score exceeds the threshold, the system outputs an anomaly diagnosis result including the anomaly level, fault type confidence level, and root cause location; otherwise, it is judged as normal fluctuation, suppressing invalid alarms. This mechanism achieves a balance between diagnostic sensitivity and specificity, ensuring that the anomaly diagnosis results are both timely and reliable.

[0043] Furthermore, the step of determining the comprehensive score based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index includes: The first weighting coefficient is determined based on the path consistency index; The second weighting coefficient is determined based on the path embedding deviation index; The third weighting coefficient is determined based on the abnormal sequence matching index; The comprehensive score is obtained by weighting and summing the path consistency index, the path embedding deviation index, the abnormal sequence matching index, the first weight coefficient, the second weight coefficient, and the third weight coefficient.

[0044] In this embodiment, weight coefficients are specifically assigned to the three indicators: the first weight coefficient corresponds to the path consistency indicator, representing the importance of topological matching evidence; the second weight coefficient corresponds to the path embedding deviation indicator, representing the weight of structural deviation evidence; and the third weight coefficient corresponds to the abnormal sequence matching indicator, representing the contribution of time-series pattern evidence. The weights can be dynamically adjusted according to equipment type or process stage; for example, the first weight coefficient can be increased in a stable production line. Finally, the system multiplies each of the three indicators by its corresponding weight and then sums them to obtain a comprehensive score.

[0045] Furthermore, embodiments of the present invention also propose a multi-temporal anomaly diagnostic device, the multi-temporal anomaly diagnostic device comprising: The acquisition module is used to control the monitoring equipment to monitor industrial operating equipment and obtain multi-source time-series signals; The analysis module is used to determine the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram; The diagnostic module is used to determine the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

[0046] Furthermore, this invention also proposes a multi-timing anomaly diagnostic device, which includes: a memory, a processor, and a multi-timing anomaly diagnostic program stored in the memory and executable on the processor. The multi-timing anomaly diagnostic program is configured to implement the steps of the multi-timing anomaly diagnostic method described above.

[0047] Furthermore, embodiments of the present invention also propose a storage medium storing a multi-timing anomaly diagnostic program, wherein when the multi-timing anomaly diagnostic program is executed by a processor, it implements the steps of the multi-timing anomaly diagnostic method described above.

[0048] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0049] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0050] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0051] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A multi-temporal anomaly diagnosis method, characterized in that, The multi-temporal anomaly diagnosis method includes the following steps: The control and monitoring equipment monitors industrial operating equipment and obtains multi-source time-series signals; The current wave propagation path and abnormal sequence group are determined based on the multi-source time-series signal and the device characteristic wave propagation diagram. The abnormal diagnosis result is determined based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

2. The multi-temporal anomaly diagnosis method as described in claim 1, characterized in that, Before the step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram, the method further includes: The historical fluctuation characteristics of each of the industrial operating devices are determined based on historical multi-source time-series signals. The direction and time delay data of fluctuation changes among the industrial operating equipment are determined based on the historical fluctuation characteristics of each of the industrial operating equipment. The device characteristic fluctuation transmission diagram is constructed based on the device connection relationship, the fluctuation change direction, and the time delay data.

3. The multi-temporal anomaly diagnosis method as described in claim 1, characterized in that, The multi-source time-series signal includes: equipment status data for each industrial operating device. The step of determining the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the equipment characteristic fluctuation transmission diagram includes: Determine the fluctuation characteristics and corresponding fluctuation times of each piece of industrial operating equipment based on the equipment status data of each piece of equipment. The real-time device status fluctuation relationship is determined based on the fluctuation characteristics of each device and the corresponding fluctuation time, and the current fluctuation propagation path is determined based on the real-time device status fluctuation relationship; When any of the industrial operating devices has an abnormal fluctuation characteristic, the abnormal sequence group is determined based on the current fluctuation propagation path and the abnormal characteristic.

4. The multi-temporal anomaly diagnosis method as described in claim 3, characterized in that, The step of determining the abnormal sequence group based on the current wave propagation path and the abnormal characteristics includes: The abnormal triggering device and the abnormal triggering time are determined based on the abnormal characteristics. The set of affected devices is determined based on the device characteristic fluctuation transmission diagram and the abnormal triggering devices; The timing relationship is determined based on the fluctuation time of each device in the affected device set and the anomaly trigger time; The abnormal sequence group is determined based on the temporal relationship.

5. The multi-temporal anomaly diagnosis method according to any one of claims 1 to 4, characterized in that, The step of determining the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index includes: A path consistency index is determined based on the current fluctuation propagation path and the set of historical regular paths. Determine the path embedding deviation index based on the current wave propagation path and the path embedding benchmark; The abnormal sequence matching index is determined based on the abnormal sequence group and the frequent sequence pattern library. The anomaly diagnosis result is determined based on the path consistency index, the path embedding deviation index, and the anomaly sequence matching index.

6. The multi-temporal anomaly diagnosis method as described in claim 5, characterized in that, The step of determining the abnormal diagnosis result based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index includes: A comprehensive score is determined based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index. The abnormal diagnosis result is determined based on the comprehensive score and the preset threshold.

7. The multi-temporal anomaly diagnosis method as described in claim 6, characterized in that, The step of determining the comprehensive score based on the path consistency index, the path embedding deviation index, and the abnormal sequence matching index includes: The first weighting coefficient is determined based on the path consistency index; The second weighting coefficient is determined based on the path embedding deviation index; The third weighting coefficient is determined based on the abnormal sequence matching index; The comprehensive score is obtained by weighting and summing the path consistency index, the path embedding deviation index, the abnormal sequence matching index, the first weight coefficient, the second weight coefficient, and the third weight coefficient.

8. A multi-time-series anomaly diagnostic device, characterized in that, The multi-time-series anomaly diagnostic device includes: The acquisition module is used to control the monitoring equipment to monitor industrial operating equipment and obtain multi-source time-series signals; The analysis module is used to determine the current fluctuation propagation path and abnormal sequence group based on the multi-source time-series signal and the device characteristic fluctuation transmission diagram; The diagnostic module is used to determine the abnormal diagnosis result based on the current fluctuation propagation path, the abnormal sequence group, and the preset discrimination index.

9. A multi-temporal anomaly diagnostic device, characterized in that, The multi-time-series anomaly diagnostic device includes: a memory, a processor, and a multi-time-series anomaly diagnostic program stored in the memory and executable on the processor, wherein the multi-time-series anomaly diagnostic program is configured to implement the steps of the multi-time-series anomaly diagnostic method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a multi-timing anomaly diagnostic program, which, when executed by a processor, implements the steps of the multi-timing anomaly diagnostic method as described in any one of claims 1 to 7.

Images