Fault diagnosis method and device based on combination of diagnosis logic and fault tree

Abstract

The invention provides a fault diagnosis method and a fault diagnosis device based on combination of diagnosis logic and a fault tree, and the method comprises the steps: obtaining a fault event of a transformer substation, taking the fault event as a top event to construct a fault tree, and enabling the fault event to be an event that the transformer substation triggers a fault alarm; a plurality of cut sets are determined based on the fault tree, the cut sets are sets of bottom events of the fault tree, and all bottom events of the cut sets cause top events to occur; determining the occurrence probability of each cut set according to historical fault reasons of the fault event, wherein the historical fault reasons are in one-to-one correspondence with the cut sets; checking whether bottom events of the cut sets occur or not in sequence according to the occurrence probabilities of the cut sets from large to small until the cut sets in which all the bottom events occur are obtained; and determining the cut sets where all the bottom events occur as the fault causes of the fault events, thereby solving the problem of low working efficiency of maintainers caused by the fact that equipment which may have faults needs to be checked one by one when the intelligent substation secondary system generates fault alarm in the prior art.

Description

Technical Field

[0001] The present application relates to the field of substation fault diagnosis, and in particular, to a fault diagnosis method, a diagnostic device and a computer program product based on the combination of diagnostic logic and a fault tree. Background Art

[0002] The existing smart substation fault diagnosis technology mainly relies on the analysis of fault phenomena, but this method has obvious limitations. First, the networking characteristics of smart substations lead to a large number of information types and quantities, and the data quality may not meet the requirements. For example, there are problems such as data loss and inaccurate data collection, which affect the accuracy of fault diagnosis. Secondly, the existing fault diagnosis model is too simple and cannot effectively cope with complex actual scenarios, resulting in inaccurate diagnostic results in complex scenarios. In addition, the accuracy of fault diagnosis needs to be improved, and the fault diagnosis data is not comprehensive enough, resulting in low reliability of fault diagnosis results.

[0003] In a smart substation, when a secondary system fails, a large amount of fault data information will be generated. The operation and maintenance personnel need to face a large amount of fault data messages with a lot of interference information. If the maintenance personnel check them one by one, the workload will be greatly increased, and the large amount of fault messages will affect the judgment of the fault maintenance personnel. At the same time, when a fault occurs, the staff needs to receive multiple information at the same time, which requires a high level of experience and comprehensive analysis ability of the operation and maintenance personnel. It is difficult to quickly locate the fault, which undoubtedly affects the efficiency of maintenance and greatly reduces the work efficiency of the maintenance personnel. Summary of the invention

[0004] The main purpose of the present application is to provide a fault diagnosis method, a diagnostic device and a computer program product based on the combination of diagnostic logic and a fault tree, so as to at least solve the problem in the prior art that when a fault alarm occurs in the secondary system of a smart substation, it is necessary to check the possibly faulty devices one by one, resulting in low work efficiency of the staff.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a fault diagnosis method based on the combination of diagnostic logic and fault tree is provided, including: obtaining a fault event of a substation, and constructing a fault tree with the fault event as a top event, wherein the fault event is an event that triggers a fault alarm of the substation; determining multiple cut sets based on the fault tree, wherein the cut set is a collection of bottom events of the fault tree, and the occurrence of all the bottom events of the cut set causes the occurrence of the top event; determining the occurrence probability of each of the cut sets according to the historical fault causes of the fault event, wherein the historical fault causes correspond to the cut sets one by one; checking whether the bottom events of the cut sets occur in sequence according to the order of the occurrence probability of each of the cut sets from large to small, until the cut set in which all the bottom events occur is obtained; and determining the cut set in which all the bottom events occur as the fault cause of the fault event.

[0006] Optionally, a fault tree is constructed using the fault event as a top event, including: determining the cause that may cause the top event to occur, and obtaining at least one intermediate event; a determination step, determining the cause that may cause the intermediate event to occur, and obtaining a next-level event; using the next-level event as an intermediate event, repeating the determination step at least once until there is no next-level event for the intermediate event, and determining the bottom event of the fault tree for the intermediate event that does not have a next-level event.

[0007] Optionally, multiple cut sets are determined based on the fault tree, including: when a target bottom event occurs and causes the top event to occur, the target bottom event is used as a cut set, and the target bottom event is any one of the bottom events; when multiple target bottom events occur and cause the top event to occur, multiple target bottom events are used as a cut set.

[0008] Optionally, after determining a plurality of cut sets based on the fault tree, the method includes: if any one of the cut sets is a subset of another cut set, deleting the other cut set, so that any one of the cut sets is not a subset of other cut sets.

[0009] Optionally, determining the occurrence probability of each of the cut sets according to the historical fault causes of the fault event includes: obtaining the number of times the fault event occurs due to the historical fault causes in multiple historical time periods to obtain multiple times, and the duration of all the historical time periods is the same; determining a range according to the maximum and minimum values ​​of the multiple times, and evenly dividing the range into multiple equidistant intervals, the intervals corresponding to fuzzy sets one-to-one, and the fuzzy sets including "very small", "small", "medium", "large" and "very large"; randomly selecting a predetermined number of the times and classifying them into corresponding intervals, and counting the number of the times in each interval, the fuzzy set corresponding to the number one-to-one; constructing a membership function for each of the fuzzy sets; determining the cut set corresponding to each of the fuzzy sets according to the membership function of each of the fuzzy sets , the cut set represents all the members in the fuzzy set that are greater than or equal to The set of elements of is the lower limit of the cut set, is the upper limit of the cut set; the average fuzzy number of each cut set is calculated according to the number corresponding to each fuzzy set, and the membership function corresponding to the average fuzzy number is generated to obtain the average membership function; the occurrence probability of the cut set corresponding to the historical fault cause is determined according to the average membership function.

[0010] Optionally, constructing a membership function for each of the fuzzy sets respectively includes: constructing a target membership function, wherein the target membership function is: , , , Satisfy 0≤ < < ≤1, represents the probability that the historical fault cause leads to the fault event, a, b and c are all constants; construct the membership function corresponding to the “very small” ; Construct the membership function corresponding to the "small" ; Construct the membership function corresponding to the "medium" ; Construct the membership function corresponding to the "big" ; Construct the membership function corresponding to the "very large" .

[0011] Optionally, the average fuzzy number of each cut-off set is calculated according to the number corresponding to each fuzzy set, and a membership function corresponding to the average fuzzy number is generated to obtain the average membership function, including: normalizing the number corresponding to each fuzzy set to obtain the weight of each fuzzy set; performing weighted averaging on the lower limits of the cut-off sets of each fuzzy set according to the weight of each fuzzy set to obtain an average lower limit; performing weighted averaging on the upper limits of the cut-off sets of each fuzzy set according to the weight of each fuzzy set to obtain an average upper limit; using the average lower limit as the lower limit of the average fuzzy number, and using the average upper limit as the upper limit of the average fuzzy number, to obtain the average fuzzy number. ; Constructing the average membership function corresponding to the average fuzzy number according to the average fuzzy number and the target membership function , where d, e, f and g are all constants.

[0012] Optionally, determining the occurrence probability of the cut set corresponding to the historical fault cause according to the average membership function includes: constructing a maximum fuzzy set membership function and a minimum fuzzy set membership function, wherein the maximum fuzzy set membership function is , the minimum fuzzy set membership function is ;pass The right fuzzy probability of the average fuzzy number is calculated, where represents the supremum, represents the average membership function, is the maximum fuzzy set membership function; The left fuzzy probability of the average fuzzy number is calculated, where represents the supremum, represents the average membership function, is the minimum fuzzy set membership function; The occurrence probability of the cut set corresponding to the historical fault cause is calculated.

[0013] According to another aspect of the present application, a fault diagnosis device based on a combination of diagnostic logic and a fault tree is provided, comprising: a first acquisition unit, used to acquire a fault event of a substation, and construct a fault tree using the fault event as a top event, wherein the fault event is an event that triggers a fault alarm of the substation; a first determination unit, used to determine a plurality of cut sets based on the fault tree, wherein the cut set is a collection of bottom events of the fault tree, and the occurrence of all the bottom events of the cut set causes the occurrence of the top event; a second determination unit, used to determine the occurrence probability of each of the cut sets according to the historical fault causes of the fault event, wherein the historical fault causes correspond one-to-one to the cut sets; a first control unit, used to check in order from large to small the occurrence probability of each of the cut sets whether the bottom events of the cut set occur, until the cut set in which all the bottom events occur is obtained; a third determination unit, used to determine the cut set in which all the bottom events occur as the fault cause of the fault event.

[0014] According to another aspect of the present application, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, any one of the methods described above is implemented.

[0015] Applying the technical solution of the present application, in the above-mentioned fault diagnosis method based on the combination of diagnostic logic and fault tree, it includes: obtaining the fault event of the substation, and constructing a fault tree with the above-mentioned fault event as the top event, wherein the above-mentioned fault event is the event that triggers the fault alarm of the above-mentioned substation; determining multiple cut sets based on the above-mentioned fault tree, wherein the above-mentioned cut set is a collection of the bottom events of the above-mentioned fault tree, and the occurrence of all the above-mentioned bottom events in the above-mentioned cut set leads to the occurrence of the above-mentioned top event; determining the occurrence probability of each of the above-mentioned cut sets according to the historical fault causes of the above-mentioned fault events, and the above-mentioned historical fault causes correspond to the above-mentioned cut sets one by one; checking whether the above-mentioned bottom events of the above-mentioned cut sets occur in turn according to the order of the occurrence probability of each of the above-mentioned cut sets from large to small, until the above-mentioned cut set in which all the above-mentioned bottom events occur is obtained; and determining the above-mentioned cut set in which all the above-mentioned bottom events occur as the fault cause of the above-mentioned fault event. The present application constructs a fault tree model of the secondary system of an intelligent substation through fault events, uses the fault tree to determine the cut set, uses the historical fault causes of the fault event as a reference, and combines the membership function to calculate the probability of occurrence of the cut set. In the fault troubleshooting process, maintenance personnel only need to check whether the equipment related to the bottom event of the fault tree has a fault according to the probability of occurrence of the cut set from large to small, without having to check the faulty equipment one by one, thereby improving the work efficiency of maintenance personnel and solving the problem in the prior art that when a fault alarm occurs in the secondary system of the intelligent substation, it is necessary to check the equipment that may have a fault one by one, resulting in low work efficiency of the staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1A hardware structure block diagram of a mobile terminal according to a fault diagnosis method based on a combination of diagnostic logic and a fault tree provided in an embodiment of the present application is shown;

[0017] Figure 2 A schematic flow chart of a fault diagnosis method based on the combination of diagnostic logic and fault tree provided according to an embodiment of the present application is shown;

[0018] Figure 3 A fault tree event and its symbol diagram of a fault diagnosis method based on a combination of diagnostic logic and a fault tree provided in accordance with an embodiment of the present application are shown;

[0019] Figure 4 A fault tree model diagram of a fault diagnosis method based on a combination of diagnostic logic and a fault tree is shown according to an embodiment of the present application;

[0020] Figure 5 A membership function diagram corresponding to each fuzzy set of a fault diagnosis method based on the combination of diagnostic logic and fault tree provided in an embodiment of the present application is shown;

[0021] Figure 6 A fuzzy probability diagram of an example of a fault diagnosis method based on a combination of diagnostic logic and a fault tree provided according to an embodiment of the present application is shown;

[0022] Figure 7 A structural block diagram of a fault diagnosis device based on a combination of diagnostic logic and a fault tree is shown according to an embodiment of the present application;

[0023] The above drawings include the following reference numerals:

[0024] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0025] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0026] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0027] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0028] For the convenience of description, some nouns or terms involved in the embodiments of the present application are explained below:

[0029] Bottom event: Located at the bottom of the fault tree, it will only cause other events to occur. It cannot continue to develop downward in the fault tree model, so it is only an input event of the logic gate.

[0030] Top event: It is at the top level of the fault tree. It is the analysis target of the fault tree and the result event of concern in the fault tree analysis. Therefore, it is only the output event of the logic gate.

[0031] Intermediate event: It is located between the top event and the bottom event of the fault tree and serves as an intermediate event of input and output. It can be used as an input event and output event of a logic gate.

[0032] OR gate: When any of the sub-events occurs, the parent event will occur.

[0033] AND gate: The parent event will only occur when all child events occur at the same time.

[0034] Transfer symbol: Since the fault tree may be too large and events often do not appear in the same location, triangle symbols and straight lines are used to express the transition in and out of a certain place. The transition in and out symbols need to be used at the same time.

[0035] Supremum: It is defined as a real number set A. If there is a real number M such that no number in A exceeds M, then M is called an upper bound of A. If there is a smallest upper bound among all those upper bounds, it is called the supremum of A.

[0036] As introduced in the background technology, in the prior art, when a secondary system fails, a large amount of fault data information will be generated. The operation and maintenance personnel need to face a large amount of fault data messages with a lot of interference information. If the maintenance personnel check them one by one, the workload will be greatly increased. In order to solve the problem that when a fault alarm occurs in the secondary system of the smart substation in the prior art, it is necessary to check the possibly faulty equipment one by one, resulting in low work efficiency of the staff, an embodiment of the present application provides a fault diagnosis method based on the combination of diagnostic logic and fault tree.

[0037] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0038] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 1 is a hardware structure block diagram of a mobile terminal of a fault diagnosis method based on a combination of diagnostic logic and fault tree according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0039] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the display method of device information in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. The transmission device 106 is used to receive or send data via a network. The above-mentioned specific examples of the network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0040] In this embodiment, a fault diagnosis method based on a combination of diagnostic logic and a fault tree is provided, which runs on a mobile terminal, a computer terminal or a similar computing device. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0041] Figure 2 FIG. 1 is a flow chart of a fault diagnosis method based on a combination of diagnostic logic and a fault tree according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0042] Step S201, obtaining a fault event of a substation, and constructing a fault tree using the fault event as a top event, wherein the fault event is an event that triggers a fault alarm of the substation;

[0043] Specifically, the starting point for constructing a fault tree is to determine the specific fault events that trigger the alarm in the substation, and regard these events as the highest-level events of the fault tree, that is, these events are regarded as the top events of the fault tree. In this way, the various causes that may lead to the occurrence of the top events can be systematically traced to form a complete fault tree. The defined fault tree not only helps to quickly identify and locate faults, but also improves the understanding and analysis of the causes of faults, thereby optimizing the fault maintenance strategy of the substation.

[0044] Step S202, determining a plurality of cut sets based on the fault tree, wherein the cut set is a set of bottom events of the fault tree, and the occurrence of all the bottom events of the cut set leads to the occurrence of the top event;

[0045] Specifically, using the constructed fault tree, multiple sets of bottom events that lead to top events can be identified. These sets are called cut sets. Each cut set contains a specific set of bottom events. When these bottom events occur at the same time, the top event will inevitably occur. This method can reveal a variety of potential fault events that lead to top events, thereby providing important information for fault troubleshooting. At the same time, the analysis of cut sets helps maintenance personnel understand the complexity of the system more quickly, understand the various causes of fault events, and improve the efficiency of fault handling.

[0046] Step S203, determining the occurrence probability of each of the above cut sets according to the historical fault causes of the above fault events, wherein the above historical fault causes correspond to the above cut sets one by one;

[0047] Specifically, by analyzing the data of historical fault causes of fault events that lead to the generation of alarm information, the occurrence probability of each cut set is obtained, that is, the probability of a set of cut sets that lead to the occurrence of the top event of the fault tree is obtained, so as to determine the specific fault cause through the obtained probabilities of each cut set, which is convenient for subsequent troubleshooting. This probability reflects the possibility of each cut set triggering the top event based on past experience, which can ensure that the fault cause leading to the fault event is obtained more scientifically, better arrange maintenance work, reduce unnecessary maintenance costs, and ensure that the problem can be quickly located and solved when a real fault occurs.

[0048] Step S204, checking whether the above-mentioned bottom events of the above-mentioned cut sets have occurred in order of the occurrence probability of each of the above-mentioned cut sets from large to small, until the above-mentioned cut sets in which all the above-mentioned bottom events have occurred are obtained;

[0049] Specifically, according to the probability of occurrence of each cut set, a probability order from high to low can be determined, and then the bottom events corresponding to each cut set are checked one by one in this order to see if they occur. This process is continued until a cut set is found in which all bottom events occur. In this way, the specific cause of the fault event is determined, thereby accurately locating the fault. By giving priority to checking the events corresponding to the cut sets that are more likely to occur, unnecessary inspection work can be reduced and the cause of the fault can be determined more quickly.

[0050] Step S205: determining the cut set where all the bottom events occur as the fault cause of the fault event.

[0051] Specifically, by checking one by one whether the bottom event corresponding to each of the above cut sets has occurred, the cut sets in which all the bottom events have occurred can be confirmed as the specific cause of the fault event. In this way, the fault handling process can be made more targeted and the processing efficiency of fault handling can be further improved.

[0052] Through this embodiment, in the above-mentioned fault diagnosis method based on the combination of diagnostic logic and fault tree, a fault tree model of the secondary system of the smart substation is constructed through fault events, the fault tree is used to determine the cut set, the historical fault causes of the fault event are used as a reference, and the membership function is combined to calculate the occurrence probability of the cut set. In the fault troubleshooting process, the maintenance personnel only need to check whether the equipment related to the bottom event of the fault tree has a fault according to the occurrence probability of the cut set from large to small, without checking the faulty equipment one by one, which improves the work efficiency of the maintenance personnel and solves the problem in the prior art that when a fault alarm occurs in the secondary system of the smart substation, it is necessary to check the equipment that may have a fault one by one, resulting in low work efficiency of the staff.

[0053] The top event, intermediate event, bottom event, AND gate, OR gate, and transfer symbol mentioned in the above terms are shown in the figure below: Figure 3 shown.

[0054] In order to more completely construct a fault tree, in an optional implementation, a fault event of the substation is obtained, and the fault event is used as a top event to construct a fault tree. The method includes:

[0055] Step S2011, determining the possible cause of the top event, and obtaining at least one intermediate event;

[0056] Specifically, in the process of establishing a fault tree, by tracing the top event, we can determine at least one intermediate event, which may trigger the occurrence of the top event. This process involves starting from the top, tracing back the various branches of the fault tree, and determining the intermediate links that lead to the top event. By monitoring and analyzing the intermediate events, the fault tree structure can be determined more clearly, which helps the subsequent process of establishing the fault tree.

[0057] Step S2012, a determination step, determines the possible cause of the above intermediate event, and obtains the next level event;

[0058] Specifically, in the process of establishing a fault tree, after determining the intermediate events, we trace back to the events at the next level. This step involves starting from the intermediate events, further exploring and analyzing the causes of these events, thereby digging deep into each level of the fault tree.

[0059] Step S2013, taking the next level event as the intermediate event, repeating the above determination step at least once, until the intermediate event has no next level event, and determining the bottom event of the fault tree with the intermediate event that has no next level event.

[0060] Specifically, the next-level event analyzed is regarded as a new intermediate event, and the confirmation steps are repeated until it is impossible to trace back to a deeper event. In this process, the intermediate event that cannot be further decomposed is defined as the bottom event of the fault tree. The bottom event has two states. If the bottom event is 1, it means that the bottom event occurs. If the bottom event is 0, the bottom event does not occur. Through this iterative process, the depth of the fault tree is continuously expanded until it reaches its bottom layer, completing the establishment of the entire fault tree. This method not only helps us deeply understand the root cause of the fault, but also provides a detailed logical path for the precise location of the fault.

[0061] It can be understood that the steps to build a fault tree are:

[0062] Be familiar with the system. Before establishing a fault tree, you must first be very familiar with the system to be established, understand the system mechanism, operation mode, fault logic, functional structure, etc., and collect relevant experience, operation, maintenance and other information to prepare for building a fault tree;

[0063] Determine the top event. Take the currently issued alarm information as the top event. Each fault tree has only one top event. By searching for information and analyzing all possible fault modes, determine the top event. The top event must be able to further decompose the cause of the top event.

[0064] Construct a fault tree, starting from the top event, analyze the direct cause of the top event as the intermediate event, and then deduce the cause of the upper level event step by step until the bottom event is derived. Connect each event through logical symbols to form a fault tree.

[0065] In order to determine the cut sets of the fault tree, in an optional implementation manner, multiple cut sets are determined based on the fault tree, and the method includes:

[0066] Step S2021, when a target bottom event occurs and causes the top event to occur, the target bottom event is used as the cut set, and the target bottom event is any one of the bottom events;

[0067] Specifically, for any single bottom event, when it is determined that the occurrence of a single bottom event is sufficient to trigger the occurrence of a top event, the bottom event is treated as a cut set. This processing method helps to quickly identify single fault points that may cause the occurrence of the top event, thereby speeding up fault processing.

[0068] Step S2022: when the occurrence of multiple target bottom events leads to the occurrence of the top event, the multiple target bottom events are used as a cut set.

[0069] Specifically, the occurrence of a top event may not be triggered by a single bottom event, but may be caused by a combination of several bottom events. When multiple bottom events work together to cause the top event to occur, these bottom events as a whole can be regarded as a cut set.

[0070] In order to determine the minimum cut set, in an optional implementation manner, after determining multiple cut sets based on the fault tree, the method includes:

[0071] Step S301: if any of the above cut sets is a subset of another cut set, delete the other cut set so that any of the above cut sets is not a subset of the other cut sets.

[0072] Specifically, if it is found that in a group of cut sets, a cut set is a subset of another cut set, then the other subset should be deleted. In this way, it is ensured that there is no situation where a cut set is a subset of another cut set among all cut sets, that is, each cut set is independent, which simplifies the set of cut sets and eliminates redundant information, thereby making the subsequent fault diagnosis process more efficient.

[0073] It can be understood that the minimum cut set is the set in which the top event will not occur when any bottom event is removed from the cut set. The cut sets of the same fault tree may contain each other, while the minimum cut set is a set of bottom events that can only be contained and cannot contain other cut sets.

[0074] In one embodiment, according to Figure 4 The fault tree model shown introduces the steps to find the minimum cut set:

[0075] S1: Below the top event is an OR gate, so its three sub-events M1, M2, and M3 are written in three rows, and each sub-event leads to a cut set;

[0076] S2: Events M1 and M2 are under OR gates, so their sub-events M4, X1, X2, and X3 are written in 4 rows. M3 is under AND gates, so M5 and M6 are written in 1 row.

[0077] S3: Below event M4 is an OR gate, and its sub-events X6 and X8 are written in two lines;

[0078] S4: Below event M5 is an OR gate, and its two sub-events are written in two lines X4M6 and X5M6;

[0079] S5: Below event M6 is an OR gate. Replace M6 with its two sub-events X6 and X7 to get X4X6, X4X7, X5X6, and X5X7 written in 4 rows;

[0080] S6: Finally, 9 cut sets can be obtained, namely {X6}, {X8}, {X1}, {X2}, {X3}, {X4, X6}, {X4, X7}, {X5, X6}, {X5, X7};

[0081] S7: Through Boolean algebra operations, we can simplify and obtain: X6+X4X6=X6, X6+X5X6=X6, and finally we can get 7 minimum cut sets: {X6}, {X8}, {X1}, {X2}, {X3}, {X4, X7}, {X5, X7}.

[0082] In order to determine the occurrence probability of each of the above cut sets, in an optional implementation manner, the occurrence probability of each of the above cut sets is determined according to the historical fault causes of the above fault events. The above method includes:

[0083] Step S2031, obtaining the number of times the fault event occurred due to the historical fault cause in multiple historical time periods, and obtaining multiple numbers, wherein the duration of all the historical time periods is the same;

[0084] Specifically, in multiple time periods of equal length, the number of all fault causes that cause fault events is collected and counted. These fault causes correspond one-to-one to the bottom events of the fault tree. By collecting data, a set of data is obtained, which includes the number of each fault cause that causes the fault event in each time period.

[0085] Step S2032, determining a range according to the maximum and minimum values ​​of the multiple times, and evenly dividing the range into a plurality of equally spaced intervals, wherein the intervals correspond to fuzzy sets one by one, and the fuzzy sets include "very small", "small", "medium", "large" and "very large";

[0086] Specifically, after obtaining multiple times of fault events caused by historical causes of faults in each historical time period, the maximum and minimum times are determined based on the multiple times obtained, a numerical interval covering this range is established, and this interval is evenly divided into several equally spaced segments, each segment corresponding to a fuzzy set, which are "very small", "small", "medium", "large" and "very large". Through this method, the specific number of occurrences can be converted into a qualitative description of the fuzzy set, making the evaluation of the frequency of occurrence of fault causes more intuitive and flexible, and can be more easily integrated into the fault prediction model to improve the explanatory power and prediction accuracy of the model.

[0087] Step S2033, randomly selecting a predetermined number of the times to be classified into the corresponding intervals, and counting the number of the times in each interval, wherein the fuzzy sets correspond to the numbers one by one;

[0088] Specifically, a certain number of data are randomly selected from the number of occurrences of fault causes recorded in each historical time period, and these data are classified into corresponding fuzzy set categories according to the numerical range they fall in, such as "very small", "small", "medium", "large" and "very large", and the number of times included in each fuzzy set category is counted, so as to establish a corresponding relationship between the fuzzy set and the number count, providing a quantitative basis for the fuzzy processing of the number of occurrences of fault causes.

[0089] Step S2034, constructing a membership function for each of the above fuzzy sets;

[0090] Specifically, a corresponding membership function is created for each defined fuzzy set. These functions are used to describe the correlation between the number of occurrences of each fault cause in each time period and the fuzzy set. After establishing the membership function, the fuzzy relationship between the number of occurrences of the fault cause and each fuzzy set can be expressed more accurately.

[0091] Step S2035, determining the cut-off set corresponding to each of the fuzzy sets according to the membership function of each of the fuzzy sets , the above cut set represents all the members in the above fuzzy set with a membership greater than or equal to The set of elements of is the lower limit of the above cut set, is the upper limit of the above cut set;

[0092] Specifically, based on the membership function of each fuzzy set, a cut set is defined for the membership function corresponding to each fuzzy set: The cut set is all elements whose membership values ​​are greater than or equal to a certain threshold. Through the concept of cut set, the elements in the fuzzy set can be further quantified, and the fuzzy membership values ​​can be converted into a specific set range.

[0093] Step S2036, calculating the average fuzzy number of each of the above cut sets according to the above numbers corresponding to each of the above fuzzy sets, and generating a membership function corresponding to the above average fuzzy number to obtain an average membership function;

[0094] Specifically, based on the statistical data of the number corresponding to each fuzzy set, the weighted average calculation of the cut sets corresponding to each fuzzy set is performed to obtain the average fuzzy number. The corresponding average membership function can be obtained by analogy with the target membership function according to the average fuzzy number. Through this method, the specific data of each fuzzy set can be converted into a unified metric, making the statistical analysis of the number of faults more reasonable.

[0095] Step S2037: determining the occurrence probability of the cut set corresponding to the historical fault cause according to the average membership function.

[0096] Specifically, according to the calculated average membership function, it is converted into the probability value of a specific cut set to evaluate the probability of occurrence of the historical fault cause corresponding to each cut set. In this way, the membership of the fuzzy set can be converted into a specific probability value, providing a quantitative measure for the possibility of the occurrence of the fault cause.

[0097] In order to determine the membership function corresponding to each fuzzy set, in an optional implementation, a membership function is constructed for each of the above fuzzy sets respectively. The above method includes:

[0098] Step S20341, constructing a target membership function, the target membership function is: , , , Satisfy 0≤ < < ≤1, represents the probability that the above historical fault causes lead to the above fault events, where a, b and c are all constants;

[0099] Specifically, we define a target membership function, , , are the parameters of the target membership function, which determine the shape of the membership function and ensure that the output of the function is within the range of valid probabilities. Through this function, we can quantify the fuzzy set. By adjusting the parameters of the membership function, customized analysis can be performed for different fault environments or conditions.

[0100] Construct the membership function corresponding to the above “very small” ;

[0101] Specifically, the fuzzy set is constructed as a "very small" membership function , membership function The parameters are =0, =0.25.

[0102] Construct the above membership function corresponding to the above “small” ;

[0103] Specifically, construct the membership function of the fuzzy set as "small" , membership function The parameters are =0, =0.25, =0.5.

[0104] Construct the above membership function corresponding to the above "medium" ;

[0105] Specifically, the membership function of the fuzzy set is constructed as "medium" , membership function The parameters are =0.25, =0.5, =0.75.

[0106] Construct the above membership function corresponding to the above “big” ;

[0107] Specifically, construct the membership function of the fuzzy set as "large" , membership function The parameters are =0.5, =0.75, =1.

[0108] Construct the membership function corresponding to the above “very large” .

[0109] Specifically, the fuzzy set is constructed as a membership function of "very large" , membership function The parameters are =0.75, =1.

[0110] In one embodiment, if Figure 5 As shown in the figure, taking the triangle membership function image as an example, the membership function image corresponding to each fuzzy set is constructed. Figure 5 The membership function images of the above fuzzy sets of "very small", "small", "medium", "large" and "very large" are respectively shown.

[0111] In order to obtain the average membership function, in an optional implementation, the average fuzzy number of each of the above-mentioned cut sets is calculated according to the above-mentioned number corresponding to each of the above-mentioned fuzzy sets, and the membership function corresponding to the above-mentioned average fuzzy number is generated to obtain the average membership function. The above-mentioned method includes:

[0112] Step S20361, normalizing the numbers corresponding to the fuzzy sets to obtain the weights of the fuzzy sets;

[0113] Specifically, the statistical data of the number corresponding to each fuzzy set is standardized so that these numbers are converted into corresponding weight ratios. These weights represent the proportion of each fuzzy set in the total, reflecting the importance of each fuzzy set relative to the total data.

[0114] Step S20362: perform weighted averaging on the lower limits of the cut sets of the fuzzy sets according to the weights of the fuzzy sets to obtain an average lower limit; perform weighted averaging on the upper limits of the cut sets of the fuzzy sets according to the weights of the fuzzy sets to obtain an average upper limit; use the average lower limit as the lower limit of the average fuzzy number; use the average upper limit as the upper limit of the average fuzzy number to obtain the average fuzzy number. ;

[0115] Specifically, according to the weight of each fuzzy set, the weighted average calculation is performed on the lower limit and upper limit of the cut set of each fuzzy set to obtain an average lower limit value and an average upper limit value respectively. Then, the calculated average lower limit value and average upper limit value are defined as the corresponding limits of the average fuzzy number, and the average fuzzy number is obtained. , thus forming a specific numerical range of the average fuzzy number. This method ensures that the impact of different fuzzy sets on the final average fuzzy number matches its importance in the overall data through weight adjustment, providing a more balanced and representative estimate.

[0116] Step S20363: construct an average membership function corresponding to the average fuzzy number according to the average fuzzy number and the target membership function. , where d, e, f and g are all constants.

[0117] Specifically, according to the calculated average fuzzy number, an inverse truncation process is performed on it. By analogy with the target membership function, the average membership function corresponding to the average fuzzy number can be obtained. In this way, we can map the numerical range of the average fuzzy number to the membership function, thereby obtaining a function that describes its degree of membership.

[0118] It can be understood that the anti-truncation set is a concept in fuzzy mathematics, which is used to extract accurate information from a fuzzy set. In a fuzzy set, an element has a membership from 0 to 1, which indicates the degree to which the element belongs to the set. The anti-truncation set processing refers to extracting all elements whose membership is greater than or equal to a specific membership threshold from the fuzzy set to form an accurate set.

[0119] In order to obtain the occurrence probability of the cut set corresponding to the historical fault cause, in an optional implementation manner, the occurrence probability of the cut set corresponding to the historical fault cause is determined according to the average membership function. The method includes:

[0120] Step S20371, constructing a maximum fuzzy set membership function and a minimum fuzzy set membership function, the maximum fuzzy set membership function is , the above minimum fuzzy set membership function is ;

[0121] Specifically, we define two membership functions, namely the maximum fuzzy set membership function and the minimum fuzzy set membership function. The maximum fuzzy set membership function represents the maximum possible degree to which an element belongs to the fuzzy set, while the minimum fuzzy set membership function represents the minimum possible degree to which an element belongs to the fuzzy set. These two membership functions reflect the performance of the relationship between the element and the fuzzy set under the most optimistic and pessimistic circumstances, respectively.

[0122] Step S20372, by The right fuzzy probability of the above average fuzzy number is calculated, where represents the supremum, represents the above average membership function, is the membership function of the above maximum fuzzy set;

[0123] Specifically, in order to determine the right fuzzy probability of the average fuzzy number, we need to calculate the intersection of the average membership function and the maximum fuzzy set membership function, and find the supremum value from them. The supremum value of this intersection is defined as the right fuzzy probability.

[0124] Step S20373, by The left fuzzy probability of the above average fuzzy number is calculated, where represents the supremum, represents the above average membership function, is the minimum fuzzy set membership function mentioned above;

[0125] Specifically, in order to determine the left fuzzy probability of the average fuzzy number, we need to calculate the intersection of the average membership function and the minimum fuzzy set membership function, and find the supremum value from them. The supremum value of this intersection is defined as the left fuzzy probability.

[0126] Step S20374, by The occurrence probability of the above cut set corresponding to the above historical failure causes is calculated.

[0127] Specifically, using the preset formula , we can calculate the probability of occurrence of the cut sets corresponding to the historical fault causes. By combining the membership function and fuzzy set theory, we can convert the membership of the fuzzy set into a specific probability value. Then we can determine which fault causes have the highest probability of occurrence based on the probability of each cut set, so as to conduct fault troubleshooting.

[0128] In one embodiment, the process of calculating the average membership function and determining the occurrence probability of the cut set corresponding to the historical fault cause according to the average membership function are explained by selecting the number of occurrences of fault events caused by historical fault causes within a predetermined number of 10 historical time periods.

[0129] The statistical results of the number of failure events caused by historical fault causes in the above 10 historical time periods are as follows: the number belonging to the fuzzy set "very small" is 1, the number belonging to the fuzzy set "small" is 3, the number belonging to the fuzzy set "medium" is 4, the number belonging to the fuzzy set "large" is 2, and the number belonging to the fuzzy set "very large" is 0.

[0130] The cut-off set of the fuzzy set membership function is obtained. According to the above fuzzy set membership function, we can find another , calculated ,Will As the lower limit of the cut set of the membership function of the fuzzy set being "medium"; , calculated ,Will As the upper limit of the cut set of the membership function when the fuzzy set is "medium", the cut set of the membership function when the fuzzy set is "medium" is ;

[0131] Similarly, the cut set of the membership function with a fuzzy set of "very small" is ;

[0132] The cut set of the membership function when the fuzzy set is "small" is ;

[0133] The cut set of the membership function when the fuzzy set is "large" is ;

[0134] The cut-off set of the membership function when the fuzzy set is "very large" is .

[0135] According to different fuzzy sets Cut set, according to the weight of each fuzzy set, the lower limit of the cut set of each fuzzy set is weighted averaged to obtain the average lower limit, according to the weight of each fuzzy set, the upper limit of the cut set of each fuzzy set is weighted averaged to obtain the average upper limit, the average lower limit is used as the lower limit of the average fuzzy number, and the average upper limit is used as the upper limit of the average fuzzy number to calculate the average fuzzy number .

[0136] Average fuzzy number Perform reverse truncation processing, and ,get ,Other ,get , and compare it with the above target membership function to get the average membership function .

[0137] like Figure 6 As shown, according to Calculate the right fuzzy probability of the above average fuzzy number according to Calculate the left fuzzy probability of the above average fuzzy number, call the corresponding function in Matlab to calculate, and get =0.651, =0.541.

[0138] according to Calculated =0.445, The value of is taken as the occurrence probability of the cut set corresponding to the historical fault cause.

[0139] In another embodiment, the above embodiment is described by taking a 500kV substation as an example.

[0140] In this example, the line protection device of the Shenqiao substation 211 issued an alarm of "abnormal synchronous voltage". The possible causes of the alarm can be obtained through the established fault tree. At the same time, the line protection device of the Shenxi line 212 reported "TV disconnection". The alarm information object is also voltage. According to the established fault tree, the cause of the fault alarm can also be obtained, and the possible fault points associated with it have an intersection.

[0141] After eliminating the primary equipment failure, the "alarm" light of the 211 Shenqiao line protection device was on and could not be reset, and the device panel displayed "synchronous voltage abnormality, locked reclosing", and the "alarm" light of the 212 Shenxi line line protection device was on and could not be reset, and the device panel displayed "TV disconnection, locked backup protection". After on-site inspection, the maintenance personnel found that another set of related protection devices did not report "TV disconnection", eliminating the voltage transformer body failure, and the 211 Shenqiao line line protection merging unit sampling was normal, eliminating the merging unit failure.

[0142] According to the above investigation, the possible fault points associated with the 212 Shenxi Line are: 1. Secondary circuit fault from the TV secondary winding to the parallel panel, 2. Poor insulation of the busbar merging unit input plug-in, 3. Circuit breaker fault, 4. PT secondary voltage circuit break; the possible fault points associated with the 211 Shenqiao Line are: 1. Secondary circuit fault from the TV secondary winding to the parallel panel, 2. Poor insulation of the input plug-in of the line merging unit of the 211 Shenqiao Line, 3. Line protection device fault of the 211 Shenqiao Line, 4. Failure of the direct optical fiber link or optical port of the 211 Shenqiao Line.

[0143] Through the above-mentioned fault diagnosis method based on the combination of diagnostic logic and fault tree, the following fault probability set is obtained: h=[0.2269, 0.0982, 0.0327, 0.0982, 0.0982, 0.0327, 0.0982, 0.3150].

[0144] The first 7 columns represent the probability of possible fault points, i.e., the secondary circuit fault from TV secondary winding to parallel panel in column 1, poor insulation of busbar merging unit input plug-in in column 2, circuit breaker fault in column 3, disconnection of PT secondary voltage circuit in column 4, poor insulation of 211 Shenqiao line merging unit input plug-in in column 5, line protection device fault in column 211 Shenqiao line in column 6, and direct optical fiber link or optical port fault in column 211 Shenqiao line in column 7. The last column represents the probability of uncertainty, which is 0.3150.

[0145] It can be seen that the defect point is most likely the "secondary circuit from the TV secondary winding to the parallel screen", and should be given priority as a defect diagnosis item.

[0146] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0147] The embodiment of the present application also provides a kind of fault diagnosis device based on the combination of diagnostic logic and fault tree. It should be noted that the fault diagnosis device based on the combination of diagnostic logic and fault tree in the embodiment of the present application can be used to execute the fault diagnosis method based on the combination of diagnostic logic and fault tree provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions thereof will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware is also possible and conceived.

[0148] The following introduces a fault diagnosis device based on the combination of diagnostic logic and fault tree provided in an embodiment of the present application.

[0149] Figure 7 FIG. 1 is a schematic diagram of a fault diagnosis device based on a combination of diagnostic logic and a fault tree according to an embodiment of the present application. Figure 7 As shown, the device comprises:

[0150] A first acquisition unit 10 is used to acquire a fault event of the substation and use the fault event as a top event to construct a fault tree, wherein the fault event is an event that triggers a fault alarm of the substation;

[0151] A first determining unit 20 is configured to determine a plurality of cut sets based on the fault tree, wherein the cut set is a set of bottom events of the fault tree, and occurrence of all the bottom events of the cut set leads to occurrence of the top event;

[0152] A second determining unit 30 is used to determine the occurrence probability of each of the above cut sets according to the historical fault causes of the above fault events, and the above historical fault causes correspond to the above cut sets one by one;

[0153] A first control unit 40 is used to check whether the above-mentioned bottom events of the above-mentioned cut sets occur in order of the occurrence probability of each of the above-mentioned cut sets from large to small, until the above-mentioned cut sets in which all the above-mentioned bottom events occur are obtained;

[0154] The third determining unit 50 is configured to determine the cut set in which all the bottom events occur as the fault cause of the fault event.

[0155] Through this embodiment, a fault tree model of the secondary system of the smart substation is constructed through fault events, the cut set is determined by using the fault tree, the historical fault causes of the fault events are used as a reference, and the probability of occurrence of the cut set is calculated in combination with the membership function. In the fault troubleshooting process, the maintenance personnel only need to check whether the equipment related to the bottom event of the fault tree has a fault according to the probability of occurrence of the cut set from large to small, without checking the faulty equipment one by one, which improves the work efficiency of the maintenance personnel and solves the problem in the prior art that when a fault alarm occurs in the secondary system of the smart substation, it is necessary to check the equipment that may have a fault one by one, resulting in low work efficiency of the staff.

[0156] In order to more completely construct a fault tree, in an optional implementation, a fault event of the substation is obtained, and the fault tree is constructed by taking the fault event as a top event. The first acquisition unit includes:

[0157] A first acquisition module determines the cause that may cause the top event to occur, and obtains at least one intermediate event;

[0158] Specifically, in the process of establishing a fault tree, by tracing the top event, we can determine at least one intermediate event, which may trigger the occurrence of the top event. This process involves starting from the top, tracing back the various branches of the fault tree, and determining the intermediate links that lead to the top event. By monitoring and analyzing the intermediate events, the fault tree structure can be determined more clearly, which helps the subsequent process of establishing the fault tree.

[0159] The second acquisition module determines the steps and causes that may cause the above intermediate events to occur, and obtains the next level event;

[0160] Specifically, in the process of establishing a fault tree, after determining the intermediate events, we trace back to the events at the next level. This step involves starting from the intermediate events, further exploring and analyzing the causes of these events, thereby digging deep into each level of the fault tree.

[0161] The third acquisition module takes the next level event as an intermediate event, repeats the above determination step at least once, until the intermediate event does not have a next level event, and determines the bottom event of the fault tree with the intermediate event that does not have a next level event.

[0162] Specifically, the next-level event analyzed is regarded as a new intermediate event, and the confirmation steps are repeated until it is impossible to trace back to a deeper event. In this process, the intermediate event that cannot be further decomposed is defined as the bottom event of the fault tree. The bottom event has two states. If the bottom event is 1, it means that the bottom event occurs. If the bottom event is 0, the bottom event does not occur. Through this iterative process, the depth of the fault tree is continuously expanded until it reaches its bottom layer, completing the establishment of the entire fault tree. This method not only helps us deeply understand the root cause of the fault, but also provides a detailed logical path for the precise location of the fault.

[0163] In order to determine the cut sets of the fault tree, in an optional implementation manner, multiple cut sets are determined based on the fault tree, and the first determining unit includes:

[0164] A first determination module, when a target bottom event occurs and causes the top event to occur, takes the target bottom event as the cut set, and the target bottom event is any one of the bottom events;

[0165] Specifically, for any single bottom event, when it is determined that the occurrence of a single bottom event is sufficient to trigger the occurrence of a top event, the bottom event is treated as a cut set. This processing method helps to quickly identify single fault points that may cause the occurrence of the top event, thereby speeding up fault processing.

[0166] The second determining module, in the case where the occurrence of the plurality of target bottom events leads to the occurrence of the top event, regards the plurality of target bottom events as a cut set.

[0167] Specifically, the occurrence of a top event may not be triggered by a single bottom event, but may be caused by a combination of several bottom events. When multiple bottom events work together to cause the top event to occur, these bottom events as a whole can be regarded as a cut set.

[0168] In order to determine the minimum cut set, in an optional implementation manner, after determining multiple cut sets based on the fault tree, the apparatus includes:

[0169] The fourth determining unit deletes the other cut set if any one of the cut sets is a subset of another cut set, so that any one of the cut sets is not a subset of the other cut sets.

[0170] Specifically, if it is found that in a group of cut sets, a cut set is a subset of another cut set, then the other subset should be deleted. In this way, it is ensured that there is no situation where a cut set is a subset of another cut set among all cut sets, that is, each cut set is independent, which simplifies the set of cut sets and eliminates redundant information, thereby making the subsequent fault diagnosis process more efficient.

[0171] In order to determine the occurrence probability of each of the above cut sets, in an optional implementation manner, the occurrence probability of each of the above cut sets is determined according to the historical fault cause of the above fault event, and the above second determination unit includes:

[0172] A third determination module obtains the number of times the fault event occurs due to the historical fault cause in multiple historical time periods, and obtains multiple times, where the duration of all the historical time periods is the same;

[0173] Specifically, in multiple time periods of equal length, the number of all fault causes that cause fault events is collected and counted. These fault causes correspond one-to-one to the bottom events of the fault tree. By collecting data, a set of data is obtained, which includes the number of each fault cause that causes the fault event in each time period.

[0174] a fourth determination module, determining a range according to the maximum and minimum values ​​of the plurality of times, and evenly dividing the range into a plurality of equally spaced intervals, wherein the intervals correspond one to one with fuzzy sets, and the fuzzy sets include “very small”, “small”, “medium”, “large”, and “very large”;

[0175] Specifically, after obtaining multiple times of fault events caused by historical causes of faults in each historical time period, the maximum and minimum times are determined based on the multiple times obtained, a numerical interval covering this range is established, and this interval is evenly divided into several equally spaced segments, each segment corresponding to a fuzzy set, which are "very small", "small", "medium", "large" and "very large". Through this method, the specific number of occurrences can be converted into a qualitative description of the fuzzy set, making the evaluation of the frequency of occurrence of fault causes more intuitive and flexible, and can be more easily integrated into the fault prediction model to improve the explanatory power and prediction accuracy of the model.

[0176] A fifth determination module randomly selects a predetermined number of the times to be classified into the corresponding intervals, and counts the number of the times in each interval, wherein the fuzzy sets correspond to the numbers one by one;

[0177] Specifically, a certain number of data are randomly selected from the number of occurrences of fault causes recorded in each historical time period, and these data are classified into corresponding fuzzy set categories according to the numerical range they fall in, such as "very small", "small", "medium", "large" and "very large", and the number of times included in each fuzzy set category is counted, so as to establish a corresponding relationship between the fuzzy set and the number count, providing a quantitative basis for the fuzzy processing of the number of occurrences of fault causes.

[0178] The sixth determination module constructs a membership function for each of the above fuzzy sets;

[0179] Specifically, a corresponding membership function is created for each defined fuzzy set. These functions are used to describe the correlation between the number of occurrences of each fault cause in each time period and the fuzzy set. After establishing the membership function, the fuzzy relationship between the number of occurrences of the fault cause and each fuzzy set can be expressed more accurately.

[0180] The seventh determination module determines the cut-off set corresponding to each of the fuzzy sets according to the membership function of each of the fuzzy sets. , the above cut set represents all the members in the above fuzzy set with a membership greater than or equal to The set of elements of is the lower limit of the above cut set, is the upper limit of the above cut set;

[0181] Specifically, based on the membership function of each fuzzy set, a cut set is defined for the membership function corresponding to each fuzzy set: The cut set is all elements whose membership values ​​are greater than or equal to a certain threshold. Through the concept of cut set, the elements in the fuzzy set can be further quantified, and the fuzzy membership values ​​can be converted into a specific set range.

[0182] An eighth determination module, calculating the average fuzzy number of each of the above cut sets according to the above numbers corresponding to each of the above fuzzy sets, and generating a membership function corresponding to the above average fuzzy number to obtain an average membership function;

[0183] Specifically, based on the statistical data of the number corresponding to each fuzzy set, the weighted average calculation of the cut sets corresponding to each fuzzy set is performed to obtain the average fuzzy number. The corresponding average membership function can be obtained by analogy with the target membership function according to the average fuzzy number. Through this method, the specific data of each fuzzy set can be converted into a unified metric, making the statistical analysis of the number of faults more reasonable.

[0184] A ninth determination module determines the occurrence probability of the cut set corresponding to the historical fault cause according to the average membership function.

[0185] Specifically, according to the calculated average membership function, it is converted into the probability value of a specific cut set to evaluate the probability of occurrence of the historical fault cause corresponding to each cut set. In this way, the membership of the fuzzy set can be converted into a specific probability value, providing a quantitative measure for the possibility of the occurrence of the fault cause.

[0186] In order to determine the membership function corresponding to each fuzzy set, in an optional implementation manner, a membership function is constructed for each of the fuzzy sets, and the sixth determination module includes:

[0187] The first determination submodule constructs a target membership function, which is: , , , Satisfy 0≤ < < ≤1, represents the probability that the above historical fault causes lead to the above fault events, where a, b and c are all constants;

[0188] Specifically, we define a target membership function, , , are the parameters of the target membership function, which determine the shape of the membership function and ensure that the output of the function is within the range of valid probabilities. Through this function, we can quantify the fuzzy set. By adjusting the parameters of the membership function, customized analysis can be performed for different fault environments or conditions.

[0189] The second determination submodule constructs the membership function corresponding to the above “very small” ;

[0190] Specifically, the fuzzy set is constructed as a "very small" membership function , membership function The parameters are =0, =0.25.

[0191] The third determination submodule constructs the membership function corresponding to the above “small” ;

[0192] Specifically, construct the membership function of the fuzzy set as "small" , membership function The parameters are =0, =0.25, =0.5.

[0193] The fourth determination submodule constructs the membership function corresponding to the above “medium” ;

[0194] Specifically, the membership function of the fuzzy set is constructed as "medium" , membership function The parameters are =0.25, =0.5, =0.75.

[0195] The fifth determination submodule constructs the membership function corresponding to the above “large” ;

[0196] Specifically, construct the membership function of the fuzzy set as "large" , membership function The parameters are =0.5, =0.75, =1.

[0197] The sixth determination submodule constructs the membership function corresponding to the above “very large” .

[0198] Specifically, the fuzzy set is constructed as a membership function of "very large" , membership function The parameters are =0.75, =1.

[0199] In one embodiment, if Figure 5 As shown in the figure, taking the triangle membership function image as an example, the membership function image corresponding to each fuzzy set is constructed. Figure 5 The membership function images of the above fuzzy sets of "very small", "small", "medium", "large" and "very large" are respectively shown.

[0200] In order to obtain the average membership function, in an optional implementation, the average fuzzy number of each of the above-mentioned cut sets is calculated according to the above-mentioned number corresponding to each of the above-mentioned fuzzy sets, and the membership function corresponding to the above-mentioned average fuzzy number is generated to obtain the average membership function. The above-mentioned eighth determination module includes:

[0201] A seventh determination submodule normalizes the numbers corresponding to the fuzzy sets to obtain weights of the fuzzy sets;

[0202] Specifically, the statistical data of the number corresponding to each fuzzy set is standardized so that these numbers are converted into corresponding weight ratios. These weights represent the proportion of each fuzzy set in the total, reflecting the importance of each fuzzy set relative to the total data.

[0203] The eighth determination submodule performs weighted averaging of the lower limits of the cut sets of the fuzzy sets according to the weights of the fuzzy sets to obtain an average lower limit, performs weighted averaging of the upper limits of the cut sets of the fuzzy sets according to the weights of the fuzzy sets to obtain an average upper limit, uses the average lower limit as the lower limit of the average fuzzy number, uses the average upper limit as the upper limit of the average fuzzy number, and obtains the average fuzzy number. ;

[0204] Specifically, according to the weight of each fuzzy set, the weighted average calculation is performed on the lower limit and upper limit of the cut set of each fuzzy set to obtain an average lower limit value and an average upper limit value respectively. Then, the calculated average lower limit value and average upper limit value are defined as the corresponding limits of the average fuzzy number, and the average fuzzy number is obtained. , thus forming a specific numerical range of the average fuzzy number. This method ensures that the impact of different fuzzy sets on the final average fuzzy number matches its importance in the overall data through weight adjustment, providing a more balanced and representative estimate.

[0205] The ninth determination submodule constructs an average membership function corresponding to the average fuzzy number according to the average fuzzy number and the target membership function. , where d, e, f and g are all constants.

[0206] Specifically, according to the calculated average fuzzy number, an inverse truncation process is performed on it. By analogy with the target membership function, the average membership function corresponding to the average fuzzy number can be obtained. In this way, we can map the numerical range of the average fuzzy number to the membership function, thereby obtaining a function that describes its degree of membership.

[0207] In order to obtain the occurrence probability of the cut set corresponding to the historical fault cause, in an optional implementation manner, the occurrence probability of the cut set corresponding to the historical fault cause is determined according to the average membership function, and the ninth determination module includes:

[0208] The tenth determination submodule constructs the maximum fuzzy set membership function and the minimum fuzzy set membership function. The maximum fuzzy set membership function is , the above minimum fuzzy set membership function is ;

[0209] Specifically, we define two membership functions, namely the maximum fuzzy set membership function and the minimum fuzzy set membership function. The maximum fuzzy set membership function represents the maximum possible degree to which an element belongs to the fuzzy set, while the minimum fuzzy set membership function represents the minimum possible degree to which an element belongs to the fuzzy set. These two membership functions reflect the performance of the relationship between the element and the fuzzy set under the most optimistic and pessimistic circumstances, respectively.

[0210] The eleventh submodule is determined by The right fuzzy probability of the above average fuzzy number is calculated, where represents the supremum, represents the above average membership function, is the membership function of the above maximum fuzzy set;

[0211] Specifically, in order to determine the right fuzzy probability of the average fuzzy number, we need to calculate the intersection of the average membership function and the maximum fuzzy set membership function, and find the supremum value from them. The supremum value of this intersection is defined as the right fuzzy probability.

[0212] The twelfth determination submodule is passed The left fuzzy probability of the above average fuzzy number is calculated, where represents the supremum, represents the above average membership function, is the minimum fuzzy set membership function mentioned above;

[0213] Specifically, in order to determine the left fuzzy probability of the average fuzzy number, we need to calculate the intersection of the average membership function and the minimum fuzzy set membership function, and find the supremum value from them. The supremum value of this intersection is defined as the left fuzzy probability.

[0214] The thirteenth submodule is determined by The occurrence probability of the above cut set corresponding to the above historical failure causes is calculated.

[0215] Specifically, using the preset formula , we can calculate the probability of occurrence of the cut sets corresponding to the historical fault causes. By combining the membership function and fuzzy set theory, we can convert the membership of the fuzzy set into a specific probability value. Then we can determine which fault causes have the highest probability of occurrence based on the probability of each cut set, so as to conduct troubleshooting.

[0216] The above-mentioned fault diagnosis device based on the combination of diagnostic logic and fault tree includes a processor and a memory, and the above-mentioned first acquisition unit, first determination unit, second determination unit, first control unit and third determination unit are all stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in the form of any combination.

[0217] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the efficiency of substation secondary system fault troubleshooting can be improved by adjusting kernel parameters. The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one storage chip.

[0218] The embodiment of the present invention provides a processor, which is used to run a program, wherein the program executes the above-mentioned fault diagnosis method based on the combination of diagnostic logic and fault tree when it is running. Specifically, a fault diagnosis method based on the combination of diagnostic logic and fault tree includes: step S201, obtaining the fault event of the substation, and constructing a fault tree with the above-mentioned fault event as the top event, wherein the above-mentioned fault event is the event that triggers the fault alarm of the above-mentioned substation; step S202, determining multiple cut sets based on the above-mentioned fault tree, wherein the above-mentioned cut set is a set of bottom events of the above-mentioned fault tree, and the occurrence of all the above-mentioned bottom events in the above-mentioned cut set leads to the occurrence of the above-mentioned top event; step S203, determining the occurrence probability of each of the above-mentioned cut sets according to the historical fault causes of the above-mentioned fault events, wherein the above-mentioned historical fault causes correspond to the above-mentioned cut sets one by one; step S204, checking whether the above-mentioned bottom events of the above-mentioned cut sets occur in order from large to small according to the occurrence probability of each of the above-mentioned cut sets, until the above-mentioned cut set in which all the above-mentioned bottom events occur is obtained; step S205, determining the above-mentioned cut set in which all the above-mentioned bottom events occur as the fault cause of the above-mentioned fault event.

[0219] The present application provides a computer program product, which, when executed on a data processing device, is suitable for executing a program that is initialized with at least the following method steps, including: step S201, obtaining a fault event of a substation, and constructing a fault tree with the above fault event as the top event, wherein the above fault event is an event that triggers a fault alarm of the above substation; step S202, determining multiple cut sets based on the above fault tree, wherein the above cut set is a collection of bottom events of the above fault tree, and the occurrence of all the above bottom events of the above cut set leads to the occurrence of the above top event; step S203, determining the occurrence probability of each of the above cut sets according to the historical fault causes of the above fault events, wherein the above historical fault causes correspond to the above cut sets one by one; step S204, checking whether the above bottom events of the above cut sets occur in order of the occurrence probability of each of the above cut sets from large to small, until the above cut sets in which all the above bottom events occur are obtained; step S205, determining the above cut sets in which all the above bottom events occur as the fault cause of the above fault event.

[0220] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0221] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0222] 1) In a fault diagnosis method based on the combination of diagnostic logic and fault tree of the present application, a fault tree model of the secondary system of the intelligent substation is constructed through fault events, a cut set is determined by using the fault tree, the historical fault causes of the fault events are used as a reference, and the probability of occurrence of the cut set is calculated in combination with the membership function. In the process of troubleshooting, maintenance personnel only need to check whether the equipment related to the bottom event of the fault tree has a fault according to the probability of occurrence of the cut set from large to small, without checking the faulty equipment one by one, thereby improving the work efficiency of maintenance personnel and solving the problem that when a fault alarm occurs in the secondary system of the intelligent substation in the prior art, it is necessary to check the equipment that may have a fault one by one, resulting in low work efficiency of the staff.

[0223] 2) In a fault diagnosis device based on the combination of diagnostic logic and fault tree of the present application, a fault tree model of the secondary system of the intelligent substation is constructed through fault events, and the cut set is determined by using the fault tree. The historical fault causes of the fault events are used as a reference, and the probability of occurrence of the cut set is calculated in combination with the membership function. In the fault troubleshooting process, the maintenance personnel only need to check whether the equipment related to the bottom event of the fault tree has a fault according to the probability of occurrence of the cut set from large to small, without having to check the faulty equipment one by one, thereby improving the work efficiency of the maintenance personnel and solving the problem in the prior art that when a fault alarm occurs in the secondary system of the intelligent substation, it is necessary to check the equipment that may have a fault one by one, resulting in low work efficiency of the staff.

[0224] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A fault diagnosis method based on the combination of diagnostic logic and fault tree, characterized in that: include: Acquire a fault event of the substation, and use the fault event as a top event to construct a fault tree, wherein the fault event is an event that triggers a fault alarm of the substation; Determine a plurality of cut sets based on the fault tree, wherein the cut set is a set of bottom events of the fault tree, and occurrence of all the bottom events of the cut set leads to occurrence of the top event; Determine the occurrence probability of each of the cut sets according to the historical fault causes of the fault event, wherein the historical fault causes correspond to the cut sets one by one; Checking whether the bottom events of the cut sets occur in order of the occurrence probabilities of the cut sets from large to small, until the cut sets in which all the bottom events occur are obtained; The cut set in which all the bottom events occur is determined as the fault cause of the fault event.

2. The method according to claim 1, characterized in that The fault event is taken as the top event to construct a fault tree, including: Determine the cause that may cause the top event to occur, and obtain at least one intermediate event; Determine the steps, determine the causes that may cause the intermediate event to occur, and obtain the next level event; The next level event is taken as an intermediate event, and the determination step is repeated at least once until the intermediate event has no next level event, and the intermediate event without the next level event is determined as the bottom event of the fault tree.

3. The method according to claim 1, characterized in that Determining a plurality of cut sets based on the fault tree includes: When a target bottom event occurs and causes the top event to occur, the target bottom event is used as a cut set, and the target bottom event is any one of the bottom events; In the case that the occurrence of multiple target bottom events leads to the occurrence of the top event, the multiple target bottom events are taken as one of the cut sets.

4. The method according to claim 1, characterized in that: After determining a plurality of cut sets based on the fault tree, the method further comprises: In the case that any one of the cut sets is a subset of another cut set, the other cut set is deleted so that any one of the cut sets is not a subset of the other cut sets.

5. The method according to claim 1, characterized in that Determining the occurrence probability of each of the cut sets according to the historical fault causes of the fault event includes: Obtain the number of times the fault event occurred due to the historical fault cause in multiple historical time periods, and obtain multiple numbers, where the duration of all the historical time periods is the same; Determine a range according to the maximum and minimum values ​​of the multiple times, and divide the range into a plurality of equally spaced intervals, wherein the intervals correspond to fuzzy sets one by one, and the fuzzy sets include "very small", "small", "medium", "large" and "very large"; Randomly selecting a predetermined number of the times to be classified into corresponding intervals, and counting the number of the times in each interval, wherein the fuzzy sets correspond to the numbers one by one; Constructing membership functions for each of the fuzzy sets respectively; Determine the cut-off set corresponding to each fuzzy set according to the membership function of each fuzzy set , the cut set represents all the members in the fuzzy set that are greater than or equal to The set of elements of is the lower limit of the cut set, is the upper limit of the cut set; Calculating the average fuzzy number of each of the cut sets according to the number corresponding to each of the fuzzy sets, and generating a membership function corresponding to the average fuzzy number to obtain an average membership function; The occurrence probability of the cut set corresponding to the historical fault cause is determined according to the average membership function.

6. The method according to claim 5, characterized in that A membership function is constructed for each of the fuzzy sets, including: Construct a target membership function, which is: , , , Satisfy 0≤ < < ≤1, represents the probability that the historical fault cause leads to the fault event, where a, b and c are all constants; Construct the membership function corresponding to the "very small" ; Construct the membership function corresponding to the "small" ; Construct the membership function corresponding to the "medium" ; Construct the membership function corresponding to the "big" ; Construct the membership function corresponding to the "very large" .

7. The method according to claim 6, characterized in that The average fuzzy number of each of the cut sets is calculated according to the number corresponding to each of the fuzzy sets, and a membership function corresponding to the average fuzzy number is generated to obtain an average membership function, including: Normalizing the numbers corresponding to each of the fuzzy sets to obtain the weight of each of the fuzzy sets; According to the weight of each fuzzy set, the lower limit of the cut set of each fuzzy set is weighted averaged to obtain an average lower limit; according to the weight of each fuzzy set, the upper limit of the cut set of each fuzzy set is weighted averaged to obtain an average upper limit; the average lower limit is used as the lower limit of the average fuzzy number, and the average upper limit is used as the upper limit of the average fuzzy number to obtain the average fuzzy number. ; Construct an average membership function corresponding to the average fuzzy number according to the average fuzzy number and the target membership function , where d, e, f and g are all constants.

8. The method according to claim 5, characterized in that Determining the occurrence probability of the cut set corresponding to the historical fault cause according to the average membership function includes: Construct a maximum fuzzy set membership function and a minimum fuzzy set membership function, the maximum fuzzy set membership function is , the minimum fuzzy set membership function is ; pass The right fuzzy probability of the average fuzzy number is calculated, where represents the supremum, represents the average membership function, is the maximum fuzzy set membership function; pass The left fuzzy probability of the average fuzzy number is calculated, where represents the supremum, represents the average membership function, is the minimum fuzzy set membership function; pass The occurrence probability of the cut set corresponding to the historical fault cause is calculated.

9. A fault diagnosis device based on the combination of diagnostic logic and fault tree, characterized in that: include: A first acquisition unit is used to acquire a fault event of the substation and use the fault event as a top event to construct a fault tree, wherein the fault event is an event that triggers a fault alarm of the substation; A first determining unit is configured to determine a plurality of cut sets based on the fault tree, wherein the cut set is a set of bottom events of the fault tree, and occurrence of all the bottom events of the cut set leads to occurrence of the top event; A second determining unit, configured to determine the occurrence probability of each of the cut sets according to the historical fault causes of the fault event, wherein the historical fault causes correspond to the cut sets in a one-to-one manner; A first control unit is used to check whether the bottom event of the cut set occurs in order of the occurrence probability of each cut set from large to small, until the cut set in which all the bottom events occur is obtained; The third determining unit is configured to determine the cut set in which all the bottom events occur as the fault cause of the fault event.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images