Transformer intelligent fault diagnosis and early warning method and system

Abstract

The application relates to the technical field of transformer operation and maintenance, in particular to a transformer intelligent fault diagnosis and early warning method and system. The method is applied to an operation and maintenance system and comprises the following steps: when an abnormal event is triggered, whether the abnormal event is a risk event is determined according to the type and the abnormal amplitude of an abnormal parameter in the abnormal event; in the case that the abnormal event is a risk event, a first operation and maintenance task is generated according to the type and the abnormal amplitude of the abnormal parameter, and a shutdown instruction for the transformer is executed; in the case that the abnormal event is a non-risk event, each estimated fault is determined according to the abnormal parameter, and concurrent abnormal data corresponding to each estimated fault is determined; each state parameter is sampled at a target frequency, and after each sampling, each state parameter is compared with the concurrent abnormal data corresponding to each estimated fault, a target fault is determined, and a second operation and maintenance task is generated according to the target fault. The scheme of the application can reduce the probability of generating invalid operation and maintenance tasks.

Description

Technical Field

[0001] This application relates to the technical field of transformer operation and maintenance, and in particular to a method and system for intelligent fault diagnosis and early warning of transformers. Background Technology

[0002] Transformers are key equipment in the operation of power systems, playing an irreplaceable role in long-distance power transmission, regional power distribution, and maintaining stable power supply.

[0003] In related technologies, the strategy of operation and maintenance system for transformer operation and maintenance management is as follows: by setting up multiple sensors inside the transformer to detect multiple parameters, when any parameter is abnormal, the operation and maintenance system predicts multiple possible causes of failure based on the abnormal parameter and the transformer's historical fault information; then, the abnormal parameter and multiple causes of failure generate operation and maintenance tasks and dispatch them to operation and maintenance personnel for execution.

[0004] However, fluctuations in power transmission and / or load can cause temporary anomalies in relevant parameters. Therefore, this operation and maintenance strategy in related technologies can sometimes lead to the generation of invalid operation and maintenance tasks, which consumes operation and maintenance resources. Furthermore, transformer operation and maintenance usually require power outages, so invalid operation and maintenance tasks can also affect the stability of power supply.

[0005] How to reduce the generation of invalid operation and maintenance tasks is an urgent problem to be solved. Summary of the Invention

[0006] Therefore, it is necessary to provide a method and system for intelligent fault diagnosis and early warning of transformers that can reduce the generation of ineffective operation and maintenance tasks, in order to address the above-mentioned technical problems.

[0007] Firstly, this application provides a method for intelligent fault diagnosis and early warning of transformers, applied to an operation and maintenance system, the method comprising: When an abnormal event is triggered, it is determined whether the abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event; the risk event is an event that affects the healthy operation of the transformer. In the event that the abnormal event is a risk event, a first maintenance task is generated based on the type of the abnormal parameter and the magnitude of the abnormality, and a shutdown command for the transformer is executed. In the case that the abnormal event is a non-risk event, each predicted fault is determined according to the abnormal parameters, and the concurrent abnormal data corresponding to each predicted fault is determined. Each state parameter is sampled at a target frequency, and each state parameter is compared with the concurrent anomaly data corresponding to each predicted fault each time it is sampled. The target fault is determined based on the comparison result. The target frequency is greater than the reference frequency for sampling each state parameter when no anomaly event occurs. A second maintenance task is generated based on the target fault.

[0008] In one embodiment, determining whether an abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event includes: Query the abnormal parameters and the abnormal magnitude from the risk label table; If the query result indicates that the type of the abnormal parameter belongs to the marked risk type, and / or the abnormal magnitude is greater than the risk magnitude corresponding to the type of the abnormal parameter, then the abnormal event is determined to be the risk event.

[0009] In one embodiment, generating the first maintenance task based on the abnormal parameter, the type of the abnormal parameter, and the magnitude of the abnormality includes: Obtain historical fault data corresponding to each fault type, including historical abnormal parameters, historical abnormal magnitude, and number of faults. Based on the type and magnitude of the abnormal parameter, a match is made in the historical fault data corresponding to each fault type, and the fault type that matches the type and magnitude of the abnormal parameter is taken as the predicted fault. The estimated faults are queued according to the number of faults to obtain a first queue, and the first maintenance task is generated according to the first queue.

[0010] In one embodiment, sampling each state parameter at a target frequency includes: The target frequency is determined based on the type of the abnormal parameter and / or the magnitude of the abnormality; Each state parameter is sampled at the target frequency.

[0011] In one embodiment, determining each predicted fault based on the anomaly parameters, and determining the concurrent anomaly parameters corresponding to each predicted fault, includes: A fault analysis template is constructed based on the type of the abnormal parameter, the magnitude of the abnormality, and the structural data of the transformer; the structural data includes the connection relationship, positional relationship, and functional dependency relationship of each component. The fault analysis template is processed using a fault analysis model to obtain the predicted faults; the fault analysis model is constructed based on a large language model. From the historical fault data, the concurrent anomaly data corresponding to each predicted fault is determined. The concurrent anomaly data includes the type of anomaly parameter and the order in which they appear.

[0012] In one embodiment, comparing each of the state parameters with the concurrent anomaly data corresponding to each of the estimated faults, and determining the target fault based on the comparison result, includes: After each sampling, newly added abnormal parameters are determined from each of the aforementioned state parameters; The type and order of occurrence of each of the newly added abnormal parameters are compared with the concurrent abnormal data corresponding to each of the predicted faults to obtain a comparison result; the comparison result characterizes the target fault.

[0013] In one embodiment, the method further includes: Obtain the standard parameter ranges corresponding to the current detection cycle, and sample each state parameter at the reference frequency within the current cycle; Each of the stated state parameters is compared with each of the stated standard parameter ranges corresponding to the current detection cycle. If an abnormal parameter is found, the abnormal event is triggered.

[0014] Secondly, this application also provides a fault diagnosis and early warning device for use in an operation and maintenance system. The device includes an event monitoring module, a first operation and maintenance module, a fault prediction module, a comparison module, and a second operation and maintenance module, wherein: The event monitoring module is used to determine whether an abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event when an abnormal event is triggered; the risk event is an event that affects the healthy operation of the transformer. The first operation and maintenance module is used to generate a first operation and maintenance task and execute a shutdown command for the transformer when the abnormal event is a risk event, based on the type of the abnormal parameter and the magnitude of the abnormality. The fault prediction module is used to determine each predicted fault based on the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each predicted fault. The comparison module is used to sample each state parameter at a target frequency, and each time each state parameter is sampled, compare each state parameter with the concurrent abnormal data corresponding to each predicted fault, and determine the target fault based on the comparison result; wherein, the target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs; The second maintenance module is used to generate a second maintenance task based on the target fault.

[0015] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the transformer intelligent fault diagnosis and early warning method as described in any one of the first aspects above.

[0016] Fourthly, this application also provides an operation and maintenance system for performing the steps of the transformer intelligent fault diagnosis and early warning method as described in any one of the first aspects above; the system includes an anomaly monitoring unit, a risk operation and maintenance unit, and a non-risk operation and maintenance unit, wherein: The anomaly monitoring unit is used to determine whether an anomaly is a risk event based on the type and magnitude of the abnormal parameters in the anomaly event when an anomaly event is triggered; the risk event is an event that affects the healthy operation of the transformer. The risk operation and maintenance unit is used to generate a first operation and maintenance task and execute a shutdown command for the transformer when the abnormal event is a risk event, based on the type of the abnormal parameter and the magnitude of the abnormality. The non-risk operation and maintenance unit is used to determine each estimated fault according to the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each estimated fault. The non-risk operation and maintenance unit is further configured to sample each state parameter at a target frequency, and each time each state parameter is sampled, compare each state parameter with the concurrent abnormal data corresponding to each predicted fault, and determine the target fault based on the comparison result; wherein, the target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs; The non-risk operation and maintenance unit is also used to generate a second operation and maintenance task based on the target fault.

[0017] The aforementioned intelligent fault diagnosis and early warning method and system for transformers, when an abnormal event is triggered by abnormal parameters, further determines whether the abnormal event is a risk event. If the abnormal event is a non-risk event that does not affect the healthy operation of the transformer, the transformer continues to operate and acquires various state parameters at a higher sampling frequency, thereby facilitating a more accurate diagnosis of the transformer's health status. Simultaneously, based on the abnormal parameters, each predicted fault is determined and concurrent abnormal data of each predicted fault is acquired. Each time a state parameter is acquired, it is compared with the concurrent abnormal data to determine the matching target fault, and then a second maintenance task is generated based on the target fault. When no new abnormal parameters are added within a preset period, it indicates that there is no fault, that is, the parameter abnormality is a temporary fluctuation. In this way, the situation where a maintenance task is generated for every abnormal event can be avoided, thereby reducing the probability of generating invalid maintenance tasks. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is an application environment diagram of the transformer intelligent fault diagnosis and early warning method in one embodiment; Figure 2 This is a flowchart illustrating a transformer intelligent fault diagnosis and early warning method in one embodiment; Figure 3 This is a flowchart illustrating the process of generating the first maintenance task in one embodiment; Figure 4 This is a flowchart illustrating the process of determining concurrency exception parameters in one embodiment; Figure 5 This is a structural block diagram of a fault diagnosis and early warning device in one embodiment; Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0021] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0022] The intelligent fault diagnosis and early warning method for transformers provided in this application can be applied to, for example... Figure 1 The application environment shown is as follows. The transformer's internal components are equipped with sensors to detect various state parameters. The transformer also has a communication module that communicates with the sensors to collect the state parameters and upload them to a computer. The computer contains an operation and maintenance system to process the state parameters uploaded by the communication module.

[0023] The computer equipment can be a personal PC, a dedicated data processing device, or one or more servers, etc.

[0024] In one exemplary embodiment, such as Figure 2 As shown, a method for intelligent fault diagnosis and early warning of transformers is provided, which is applied to... Figure 1 Taking the operation and maintenance system in the example, which is executed by computer equipment, as an example, the steps 10-50 are included: Step 10: When an abnormal event is triggered, determine whether the abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event; a risk event is an event that affects the healthy operation of the transformer.

[0025] In this embodiment of the application, a natural day is divided into multiple consecutive detection cycles in advance. The length of each detection cycle can be preset by the operation and maintenance personnel, for example, it can be 4 hours or 6 hours. In this embodiment of the application, no specific setting is made.

[0026] The operation and maintenance system collects various status parameters at a fixed reference frequency within each detection cycle; the reference frequency can be 10 times / minute or 20 times / minute. In this embodiment, the specific value of the reference frequency is not specifically limited.

[0027] Furthermore, the types of state parameters include, but are not limited to, the following major items: Power (active power + reactive power + apparent power): Analyzes power transmission efficiency and evaluates transformer internal losses and operating economy; Water content in oil: indicates whether the insulating oil is damp; Transformer body vibration (vibration amplitude + spectrum characteristics): monitoring whether the mechanical structure of the core and windings is stable; Internal oil pressure: Abnormal internal arc discharge can cause a sudden increase in pressure. Fuel tank and radiator temperature: characterizes the heat dissipation capacity.

[0028] Specifically, because the load pressure varies across different testing cycles, the standard parameter ranges used to determine abnormal parameters differ for each testing cycle. For example, during nighttime when electricity consumption and load pressure are higher, the active power of the transformer and the oil temperature of the insulating oil in the transformer should correspond to a larger parameter range. Standard parameter ranges for various parameters are pre-set for each testing cycle. For each testing cycle, the computer equipment acquires the standard parameter ranges corresponding to the current testing cycle and samples each state parameter at a reference frequency within the current cycle.

[0029] Furthermore, each time a state parameter is collected, it is compared with the standard parameter range corresponding to the current detection cycle based on the parameter type. An abnormal event is triggered when an abnormal parameter is found. An abnormal parameter is one that is not within the standard parameter range corresponding to its parameter type. The abnormal magnitude is the difference between the state parameter and the maximum / minimum value of the standard parameter range; the abnormal magnitude can be positive or negative. When the state parameter is greater than the maximum value of the standard parameter range, the abnormal magnitude is positive, equal to the state parameter minus the maximum value of the standard parameter range; when the state parameter is less than the minimum value of the standard parameter range, the abnormal magnitude is negative, equal to the state parameter minus the minimum value of the standard parameter range.

[0030] An abnormal event is triggered when an abnormal parameter is identified. Abnormal events are categorized as risk events and non-risk events. Risk events are those that affect the healthy operation of the transformer; these must be addressed immediately. For example, an abnormal temperature of the transformer's insulating oil with a significant abnormal amplitude. Non-risk events, on the other hand, are those that do not affect the healthy operation of the transformer and do not require immediate attention. For example, a high vibration frequency in the transformer (potentially caused by severe weather) may also be a risk factor.

[0031] Step 20: If the abnormal event is a risk event, generate the first maintenance task based on the type and magnitude of the abnormal parameters and execute the shutdown command for the transformer.

[0032] In this embodiment of the application, when an abnormal event is determined to be a risk event, various predicted faults are identified based on the type and magnitude of the abnormal parameters, and a first maintenance task is generated based on each predicted fault. Furthermore, the first maintenance task is dispatched to instruct the corresponding maintenance personnel to execute it. Simultaneously, to avoid the risk of abnormal transformer operation, a shutdown command is issued to the transformer to instruct it to stop power distribution, thereby facilitating troubleshooting and repair by maintenance personnel.

[0033] Among them, a pre-trained fault prediction model can be used to determine each predicted fault based on the type and magnitude of abnormal parameters, and each predicted fault can also be determined by condition matching. Furthermore, each predicted fault is sorted in order of probability from high to low to form a first queue, and then a first maintenance task is constructed based on the first queue to indicate the priority of maintenance personnel to carry out inspection and troubleshooting.

[0034] Step 30: If the abnormal event is a non-risk event, determine each predicted fault based on the abnormal parameters, and determine the concurrent abnormal data corresponding to each predicted fault.

[0035] Step 40: Sample each state parameter at the target frequency, and compare each state parameter with the concurrent anomaly data corresponding to each predicted fault each time it is sampled. Based on the comparison results, determine the target fault. The target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs.

[0036] In the embodiments of this application, when the abnormal event is a non-risk event, since the abnormal event does not affect the healthy operation of the transformer, the transformer can continue to operate to obtain more status data to help determine the actual target fault.

[0037] Each predicted fault is identified based on abnormal parameters, and the corresponding concurrent abnormal data for each predicted fault is determined. The concurrent abnormal data includes the type of abnormal parameters and their order of occurrence. This concurrent abnormal data is determined based on historical fault data of the transformer.

[0038] During the continuous operation of the transformer, status data is continuously collected at a higher target frequency. Each time status data is collected, sequential abnormal parameters are re-determined, and it is determined whether any new abnormal parameters are present. If no new abnormal parameters are detected within a preset period, it indicates that there is no fault, meaning the parameter abnormalities are temporary fluctuations. If new abnormal parameters are present, each abnormal parameter is compared with the concurrent abnormal data corresponding to each predicted fault, and the target fault is determined based on the comparison results. The preset period is shorter than the detection period; its specific duration is not specifically set in this embodiment. Furthermore, the preset duration can also be determined based on the type of abnormal parameter.

[0039] Step 50: Generate a second maintenance task based on the target fault.

[0040] In this embodiment of the application, when a target fault is determined, the maintenance process information corresponding to the target fault is obtained, and a second maintenance task is constructed based on the maintenance process information and assigned to the corresponding maintenance personnel for execution.

[0041] In the aforementioned intelligent fault diagnosis and early warning method for transformers, when an abnormal event is triggered by abnormal parameters, it is further determined whether the abnormal event is a risk event. If the abnormal event is a non-risk event that does not affect the healthy operation of the transformer, the transformer continues to operate and acquires various state parameters at a higher sampling frequency, thereby facilitating a more accurate diagnosis of the transformer's health status. Simultaneously, based on the abnormal parameters, each predicted fault is determined and concurrent abnormal data of each predicted fault is acquired. Each time a state parameter is collected, the state parameter is compared with the concurrent abnormal data to determine the matching target fault, and then a second maintenance task is generated based on the target fault. When no new abnormal parameters are added within a preset period, it indicates that there is no fault, that is, the parameter abnormality is a temporary fluctuation. In this way, the situation where a maintenance task is generated for every abnormal event can be avoided, thereby reducing the probability of generating invalid maintenance tasks.

[0042] In one embodiment, step 10, the process of determining whether an abnormal event is a risk event, may specifically include: querying the abnormal parameters and abnormal magnitude from the risk label table; if the query result indicates that the type of the abnormal parameter belongs to the labeled risk type, and / or the abnormal magnitude is greater than the risk magnitude corresponding to the type of the abnormal parameter, then the abnormal event is determined to be a risk event.

[0043] Specifically, a risk labeling table is pre-set, which includes multiple risk conditions. Among them, the risk conditions may only include the risk type that triggers the risk event. That is to say, as long as the abnormal parameter belongs to the risk type, the risk event it triggers is a risk event.

[0044] Risk conditions can also include only the risk magnitude corresponding to the parameter type that triggers the risk event. In other words, regardless of whether the abnormal parameter type belongs to a risk type, as long as the abnormal magnitude corresponding to the abnormal parameter exceeds the risk magnitude of that parameter type, the triggered abnormal event is a risk event.

[0045] Risk conditions can also include both the risk type that triggers the risk event and the risk magnitude corresponding to the parameter type that triggers the risk event. In other words, only if the abnormal parameter belongs to a risk type and the abnormal magnitude exceeds the corresponding risk magnitude will the triggered abnormal event be a risk event.

[0046] In one embodiment, such as Figure 3 As shown, in step 20, the first maintenance task is generated based on the abnormal parameters, the type of abnormal parameters, and the magnitude of the abnormality, including steps 21-23, wherein: Step 21: Obtain historical fault data corresponding to each fault type. Historical fault data includes historical abnormal parameters, historical abnormal magnitude, and number of faults. Step 22: Based on the type and magnitude of the abnormal parameters, match them in the historical fault data corresponding to each fault type, and take the fault type that matches the type and magnitude of the abnormal parameters as the predicted fault. Step 23: Queue the estimated faults according to the number of faults to obtain the first queue, and generate the first maintenance task according to the first queue.

[0047] Specifically, for each fault type, historical operation and maintenance data corresponding to that fault type are statistically analyzed to obtain the corresponding historical anomaly parameters, the average historical anomaly magnitude for each historical anomaly parameter, and the number of faults. Based on the type and magnitude of the anomaly parameters, a match is made in the historical fault data corresponding to each fault type, and the fault types that match the type and magnitude of the anomaly parameters are used as predicted faults. At least one predicted fault is included.

[0048] When multiple faults are predicted, they are queued in descending order of their historical occurrence counts to form a first queue. The order within this first queue represents the probability of a predicted fault actually occurring. In other words, the more historical occurrences a predicted fault has, the higher its probability of occurrence. The first maintenance task generated from this first queue guides maintenance personnel in prioritizing troubleshooting and repair, thus facilitating rapid fault location.

[0049] In one embodiment, step 30 is preceded by determining a target frequency based on the type and / or magnitude of the anomalous parameters.

[0050] Specifically, a first correspondence between parameter types and target frequencies is pre-defined, as well as a second correspondence between the abnormal amplitude and frequency for each parameter type. In actual implementation, the corresponding target frequency can be determined based on the type of the abnormal parameter and the first correspondence or the abnormal amplitude and the second correspondence. Alternatively, a first target frequency can be determined simultaneously based on the type of the abnormal parameter and the first correspondence, and a second target frequency can be determined based on the abnormal amplitude and the second correspondence, with the larger of the first and second target frequencies being determined as the target frequency. Further, in step 30, each state parameter is sampled using the target frequency.

[0051] In one embodiment, such as Figure 4 As shown, in step 30, each predicted fault is determined based on the abnormal parameters, and the corresponding concurrent abnormal parameters for each predicted fault are determined. Specifically, this may include steps 31-33, wherein: Step 31: Construct a fault analysis template based on the types and magnitudes of abnormal parameters and the structural data of the transformer; Step 32: Process the fault analysis template using the fault analysis model to obtain each predicted fault; the fault analysis model is built based on a large language model.

[0052] Specifically, the structural data includes the connection relationships, positional relationships, and functional dependencies of each component. The structural data is stored in XML or JSON format and is obtained by analyzing the structural link diagram of the transformer using a multimodal large model. Abnormal parameters, their types, magnitudes, and structural data are populated into the analysis template to construct a fault analysis template. Pre-set task prompts in the analysis template instruct the fault analysis model to analyze the abnormal parameters, their types, and magnitudes based on the structural data to determine possible predicted faults.

[0053] Step 33: From the historical fault data, determine the concurrent anomaly data corresponding to each predicted fault. The concurrent anomaly data includes the type of anomaly parameter and the order in which they appear.

[0054] Specifically, historical fault data of the transformer is acquired, and concurrent abnormal data is determined from the historical fault data. The concurrent abnormal data includes historical abnormal parameters that appear in all historical fault data. Furthermore, based on the occurrence time of each historical abnormal parameter in each historical fault data, the order of occurrence of each abnormal parameter is statistically analyzed, and finally, the concurrent abnormal data corresponding to each fault type is obtained.

[0055] Furthermore, in step 40, after each sampling, newly added abnormal parameters are determined from each state parameter; the type and occurrence order of each newly added abnormal parameter are compared with the concurrent abnormal data corresponding to each predicted fault to obtain the comparison result; the comparison result characterizes the target fault.

[0056] Specifically, after each sampling, a comparison is made with the standard parameter ranges corresponding to the current detection cycle to determine whether any new abnormal parameters have been added from the sampled state parameters. The type of the new abnormal parameter differs from the abnormal parameter that triggers an abnormal event. If a new abnormal parameter exists, it is determined whether it triggers a risk event. If a risk event is triggered, the transformer is instructed to shut down, and a first maintenance task is generated based on the abnormal parameters, the new abnormal parameter, and their respective abnormal magnitudes. If the new abnormal parameter is determined not to trigger a risk event, the transformer continues to run, and state parameters are continuously collected to determine if a new abnormal parameter has been added.

[0057] The order in which each newly added abnormal parameter appears is recorded. Then, based on the abnormal parameters, their types, and their order of appearance, the data is compared with the concurrent abnormal data corresponding to each predicted fault to obtain a comparison result. Each sampling corresponds to one comparison result. If at least three abnormal parameters match the concurrent abnormal data of one of the fault types in terms of type and order of appearance, then that fault type is determined as the target fault type.

[0058] In the scheme of this application embodiment, when there are abnormal parameters that do not affect the healthy operation of the transformer, the transformer is kept running continuously, and various state parameters are collected at a higher sampling frequency. First, multiple predicted faults are determined based on the initially determined abnormal parameters, and concurrent abnormal data corresponding to each predicted fault are identified. New abnormal parameters are determined based on the data obtained from continuous sampling, and then compared sequentially with the concurrent abnormal data corresponding to each fault type according to the type and order of occurrence of the new abnormal parameters. That is, each time a new abnormal parameter appears, a portion of non-fault types that do not meet the criteria can be eliminated. After multiple rounds of comparison and elimination, the fault types with the same abnormal parameter type and the same order of occurrence are finally identified as the target types, thereby greatly improving the accuracy of the determined target fault types.

[0059] In summary, the intelligent transformer fault diagnosis and early warning method provided in this application has the following beneficial effects: By distinguishing between risk events and non-risk events, the system immediately shuts down and generates a first maintenance task only during risk events, while maintaining continuous operation of the transformer and sampling status parameters at a target frequency higher than the reference frequency during non-risk events. This effectively avoids the direct triggering of invalid maintenance tasks by temporary parameter anomalies caused by power fluctuations or load fluctuations, thereby reducing the probability of generating invalid maintenance tasks and reducing the waste of maintenance resources. Simultaneously, by avoiding unnecessary power outages for maintenance, the stability of power supply is improved. For non-risk events, multiple predicted faults are determined based on the type and magnitude of the abnormal parameters, and concurrent abnormal data corresponding to each predicted fault is obtained. New abnormal parameters are obtained through continuous high-frequency sampling, and the type and order of occurrence of the new abnormal parameters are compared sequentially with each concurrent abnormal data. Each new abnormal parameter eliminates some inconsistent fault types. After multiple rounds of comparison and elimination, the target fault is finally accurately determined, significantly improving the accuracy of fault diagnosis. In addition, the target frequency can be dynamically adjusted according to the type and magnitude of abnormal parameters, which can flexibly adapt to abnormal situations of different severity. Based on the number of historical failures, the estimated failures are queued and the first maintenance task is generated, which can guide maintenance personnel to troubleshoot and repair according to priority, further improving maintenance efficiency.

[0060] It should be understood that although the steps in the flowcharts of the above embodiments are shown 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 steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0061] Based on the same inventive concept, this application also provides a fault diagnosis and early warning device for implementing the above-mentioned intelligent fault diagnosis and early warning method for transformers. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more fault diagnosis and early warning device embodiments provided below can be found in the limitations of the intelligent fault diagnosis and early warning method for transformers described above, and will not be repeated here.

[0062] In one exemplary embodiment, such as Figure 5 As shown, a fault diagnosis and early warning device 500 is provided, including an event monitoring module 501, a first maintenance module 502, a fault prediction module 503, a comparison module 504, and a second maintenance module 505, wherein: The event monitoring module 501 is used to determine whether an abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event when an abnormal event is triggered; a risk event is an event that affects the healthy operation of the transformer. The first maintenance module 502 is used to generate a first maintenance task and execute a shutdown command for the transformer when the abnormal event is a risk event, based on the type and magnitude of the abnormal parameters. The fault prediction module 503 is used to determine each predicted fault based on the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each predicted fault. The comparison module 504 is used to sample each state parameter at a target frequency, and compare each state parameter with the concurrent abnormal data corresponding to each predicted fault each time it is sampled, and determine the target fault based on the comparison result; wherein, the target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs; The second maintenance module 505 is used to generate a second maintenance task based on the target fault.

[0063] In one embodiment, the event monitoring module 501 is specifically used for: Query the abnormal parameters and abnormal magnitudes from the risk label table; If the query result indicates that the type of the abnormal parameter belongs to the marked risk type, and / or the abnormal magnitude is greater than the risk magnitude corresponding to the type of the abnormal parameter, then the abnormal event is determined to be a risk event.

[0064] In one embodiment, the first operation and maintenance module is specifically used for: Obtain historical fault data corresponding to each fault type. The historical fault data includes historical abnormal parameters, historical abnormal magnitude, and number of faults. Based on the type and magnitude of the abnormal parameters, a match is made in the historical fault data corresponding to each fault type, and the fault type that matches the type and magnitude of the abnormal parameters is taken as the predicted fault. The predicted faults are queued according to the number of faults to obtain the first queue, and the first maintenance task is generated based on the first queue.

[0065] In one embodiment, the fault diagnosis and early warning device further includes a target frequency determination module, specifically used for: Determine the target frequency based on the type and / or magnitude of the abnormal parameters; Each state parameter is sampled at the target frequency.

[0066] In one embodiment, the fault prediction module is specifically used for: A fault analysis template is constructed based on the types and magnitudes of abnormal parameters and the structural data of the transformer; the structural data includes the connection relationships, positional relationships, and functional dependencies of each component. The fault analysis model is used to process the fault analysis template to obtain various predicted faults; the fault analysis model is built based on a large language model. From the historical fault data, the concurrent anomaly data corresponding to each predicted fault is determined. The concurrent anomaly data includes the type of anomaly parameter and the order in which they appear.

[0067] In one embodiment, the comparison module is specifically used for: After each sampling, newly added abnormal parameters are identified from each state parameter. The type and order of occurrence of each newly added abnormal parameter are compared with the concurrent abnormal data corresponding to each predicted fault to obtain the comparison results; the comparison results characterize the target fault.

[0068] In one embodiment, the event monitoring module is specifically used for: Obtain the standard parameter ranges corresponding to the current detection cycle, and sample each state parameter at the reference frequency within the current cycle; Each status parameter is compared with the standard parameter range corresponding to the current detection cycle, and an abnormal event is triggered when an abnormal parameter is found.

[0069] Each module in the aforementioned fault diagnosis and early warning device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0070] In one exemplary embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. 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 and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, Near Field Communication (NFC), or other technologies. When the computer program is executed by the processor, it implements a method for intelligent fault diagnosis and early warning of transformers. The display unit is used to form a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device of the computer device can be a touch layer covering the display screen, or buttons, trackballs, or touchpads set on the casing of the computer device, or external keyboards, touchpads, or mice, etc.

[0071] Those skilled in the art will understand that Figure 6 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.

[0072] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program as the steps of any of the methods described in the above embodiments of the transformer intelligent fault diagnosis and early warning method.

[0073] In one exemplary embodiment, an operation and maintenance system is provided for performing the steps of any of the methods described in the above embodiments of the intelligent fault diagnosis and early warning method for transformers.

[0074] Specifically, the operation and maintenance system includes an anomaly monitoring unit, a risk-based operation and maintenance unit, and a non-risk-based operation and maintenance unit, among which: The anomaly monitoring unit is used to determine whether an anomaly is a risk event based on the type and magnitude of the abnormal parameters in the anomaly event when it is triggered; a risk event is an event that affects the healthy operation of the transformer. The risk operation and maintenance unit is used to generate the first operation and maintenance task and execute the shutdown command for the transformer when the abnormal event is a risk event, based on the type and magnitude of the abnormal parameters. The non-risk operation and maintenance unit is used to determine each estimated fault based on the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each estimated fault. The non-risk operation and maintenance unit is also used to sample each state parameter at a target frequency, and when each state parameter is sampled, it is compared with the concurrent abnormal data corresponding to each predicted fault. Based on the comparison results, the target fault is determined. The target frequency is greater than the base frequency for sampling each state parameter when no abnormal event occurs. The non-risk operation and maintenance unit is also used to generate a second operation and maintenance task based on the target fault.

[0075] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0076] Those skilled in the art will understand that all or part of the processes in the methods of 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 of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory 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, artificial intelligence (AI) processors, etc., and are not limited to these.

[0077] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0078] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for intelligent fault diagnosis and early warning of transformers, characterized in that, Applied to operation and maintenance systems, the method includes: When an abnormal event is triggered, it is determined whether the abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event; the risk event is an event that affects the healthy operation of the transformer. In the event that the abnormal event is a risk event, a first maintenance task is generated based on the type of the abnormal parameter and the magnitude of the abnormality, and a shutdown command for the transformer is executed. In the case that the abnormal event is a non-risk event, each predicted fault is determined according to the abnormal parameters, and the concurrent abnormal data corresponding to each predicted fault is determined. Each state parameter is sampled at a target frequency, and each state parameter is compared with the concurrent anomaly data corresponding to each predicted fault each time it is sampled. The target fault is determined based on the comparison result. The target frequency is greater than the reference frequency for sampling each state parameter when no anomaly event occurs. A second maintenance task is generated based on the target fault.

2. The method according to claim 1, characterized in that, The step of determining whether an abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event includes: Query the abnormal parameters and the abnormal magnitude from the risk label table; If the query result indicates that the type of the abnormal parameter belongs to the marked risk type, and / or the abnormal magnitude is greater than the risk magnitude corresponding to the type of the abnormal parameter, then the abnormal event is determined to be the risk event.

3. The method according to claim 1, characterized in that, The step of generating a first maintenance task based on the abnormal parameters, the type of the abnormal parameters, and the magnitude of the abnormality includes: Obtain historical fault data corresponding to each fault type, including historical abnormal parameters, historical abnormal magnitude, and number of faults. Based on the type and magnitude of the abnormal parameter, a match is made in the historical fault data corresponding to each fault type, and the fault type that matches the type and magnitude of the abnormal parameter is taken as the predicted fault. The estimated faults are queued according to the number of faults to obtain a first queue, and the first maintenance task is generated according to the first queue.

4. The method according to claim 1, characterized in that, The sampling of each state parameter at the target frequency includes: The target frequency is determined based on the type of the abnormal parameter and / or the magnitude of the abnormality; Each state parameter is sampled at the target frequency.

5. The method according to any one of claims 1 to 4, characterized in that, The step of determining each predicted fault based on the abnormal parameters, and determining the concurrent abnormal parameters corresponding to each predicted fault, includes: A fault analysis template is constructed based on the type of the abnormal parameter, the magnitude of the abnormality, and the structural data of the transformer; the structural data includes the connection relationship, positional relationship, and functional dependency relationship of each component. The fault analysis template is processed using a fault analysis model to obtain the predicted faults; the fault analysis model is constructed based on a large language model. From the historical fault data, the concurrent anomaly data corresponding to each predicted fault is determined. The concurrent anomaly data includes the type of anomaly parameter and the order in which they appear.

6. The method according to claim 5, characterized in that, The step of comparing each of the state parameters with the concurrent anomaly data corresponding to each of the predicted faults, and determining the target fault based on the comparison results, includes: After each sampling, newly added abnormal parameters are determined from each of the aforementioned state parameters; The type and order of occurrence of each of the newly added abnormal parameters are compared with the concurrent abnormal data corresponding to each of the predicted faults to obtain a comparison result; the comparison result characterizes the target fault.

7. The method according to claim 1 or 6, characterized in that, The method further includes: Obtain the standard parameter ranges corresponding to the current detection cycle, and sample each state parameter at the reference frequency within the current cycle; Each of the stated state parameters is compared with each of the stated standard parameter ranges corresponding to the current detection cycle. If an abnormal parameter is found, the abnormal event is triggered.

8. A fault diagnosis and early warning device, characterized in that, Applied to an operation and maintenance system, the device includes an event monitoring module, a first operation and maintenance module, a fault prediction module, a comparison module, and a second operation and maintenance module, wherein: The event monitoring module is used to determine whether an abnormal event is a risk event based on the type and magnitude of the abnormal parameters in the abnormal event when an abnormal event is triggered; the risk event is an event that affects the healthy operation of the transformer. The first operation and maintenance module is used to generate a first operation and maintenance task and execute a shutdown command for the transformer when the abnormal event is a risk event, based on the type of the abnormal parameter and the magnitude of the abnormality. The fault prediction module is used to determine each predicted fault based on the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each predicted fault. The comparison module is used to sample each state parameter at a target frequency, and each time each state parameter is sampled, compare each state parameter with the concurrent abnormal data corresponding to each predicted fault, and determine the target fault based on the comparison result; wherein, the target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs; The second maintenance module is used to generate a second maintenance task based on the target fault.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-7.

10. An operation and maintenance system, characterized in that, The system is used to perform the steps of the method as described in any one of claims 1-7; the system includes an anomaly monitoring unit, a risk operation and maintenance unit, and a non-risk operation and maintenance unit, wherein: The anomaly monitoring unit is used to determine whether an anomaly is a risk event based on the type and magnitude of the abnormal parameters in the anomaly event when an anomaly event is triggered; the risk event is an event that affects the healthy operation of the transformer. The risk operation and maintenance unit is used to generate a first operation and maintenance task and execute a shutdown command for the transformer when the abnormal event is a risk event, based on the type of the abnormal parameter and the magnitude of the abnormality. The non-risk operation and maintenance unit is used to determine each estimated fault according to the abnormal parameters when the abnormal event is a non-risk event, and to determine the concurrent abnormal data corresponding to each estimated fault. The non-risk operation and maintenance unit is further configured to sample each state parameter at a target frequency, and each time each state parameter is sampled, compare each state parameter with the concurrent abnormal data corresponding to each predicted fault, and determine the target fault based on the comparison result; wherein, the target frequency is greater than the reference frequency for sampling each state parameter when no abnormal event occurs; The non-risk operation and maintenance unit is also used to generate a second operation and maintenance task based on the target fault.

Images