Methods, devices, and equipment for optimizing fault diagnosis sequence based on historical repair records

Abstract

This application belongs to the field of data processing technology, specifically disclosing a method, apparatus, and equipment for optimizing fault diagnosis order based on historical repair records. Through this application, after constructing a current fault state array, a historical fault state matrix, and a historical fault unit array, the fault diagnosis weight coefficient of each unit in the equipment is calculated. The target fault diagnosis order is determined based on the identification information of each fault diagnosis weight coefficient, and the fault diagnosis order in the maintenance manual is optimized based on the target fault diagnosis order. By fully utilizing historical repair records and combining them with current fault phenomena to construct the current fault state array, the historical fault state matrix, and the historical fault unit array, and after determining the target fault diagnosis order based on the identification information of the fault diagnosis weight coefficients, the fault diagnosis order is optimized, thereby effectively improving the accuracy and efficiency of optimizing the fault diagnosis order, and thus improving the efficiency of fault diagnosis.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and more specifically, relates to a method, apparatus and equipment for optimizing the fault diagnosis sequence based on historical repair records. Background Technology

[0002] "By observing several phenomena at the time of the fault, and combining knowledge of the equipment's working principles, we determine the possible range of faulty units, and then check them one by one in a certain order to finally find the faulty unit" is the general fault location process. Among these, the key to efficiently completing fault location is to reasonably select the order of fault investigation.

[0003] The commonly used troubleshooting sequence is derived from consulting the equipment's repair manual. However, these manuals are typically issued when the equipment is put into use, and their completion date is usually earlier than the equipment's commissioning time. This often results in a static troubleshooting sequence. As equipment becomes increasingly complex, the one-to-one mapping between fault symptoms and faulty units is no longer dominant. Due to the complex and many-to-many mapping between multiple states and multiple units of the equipment, the issued repair manuals do not record a significant number of fault symptoms. Furthermore, the repair manuals record typical fault symptoms, not a comprehensive list of all fault symptoms, rendering the troubleshooting sequence based on "equipment principle + typical fault symptoms" inapplicable. Currently, the common method for optimizing the troubleshooting sequence still relies on professional technicians. However, the increasing complexity of equipment and the complex and many-to-many mapping between multiple states and multiple units of the equipment demand a higher level of expertise from technicians, leading to significant time wastage. Moreover, even professional technicians are prone to errors during the optimization process. Therefore, the accuracy of optimizing the troubleshooting sequence using the above methods is relatively low. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of this application is to provide a method, apparatus and equipment for optimizing the fault diagnosis sequence based on historical repair records, which aims to solve the problem that the prior art has low accuracy in optimizing the fault diagnosis sequence because it requires a higher level of professionalism from the technicians and the technicians may make more or less mistakes during the optimization process.

[0005] To achieve the above objectives, in a first aspect, this application provides a method for optimizing the fault diagnosis sequence based on historical repair records, comprising: When a device malfunction is detected, the current malfunction phenomenon and historical repair records of the device are obtained; Construct a current fault status array, a historical fault status matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively. The fault investigation weight coefficient of each unit in the device is calculated based on the current fault status array, the historical fault status matrix, and the historical fault unit array. The target fault diagnosis order is determined based on the identification information of each fault diagnosis weight coefficient, and the fault diagnosis order in the maintenance manual is optimized based on the target fault diagnosis order. The step of calculating the fault investigation weight coefficient of each unit in the device based on the current fault state array, the historical fault state matrix, and the historical fault unit array includes: Calculate the modulus of the current fault state array, and iterate through and update the elements in the current fault state array according to the modulus of the current fault state array; Calculate the modulus of each column vector of the historical fault state matrix, and iterate through and update the elements in the historical fault state matrix according to the modulus of each column vector of the historical fault state matrix. Decompose the updated historical fault state matrix; The fault investigation weight coefficient of each unit in the device is calculated based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, the traversed and updated current fault state array, and the historical fault unit array.

[0006] In one embodiment, the step of constructing a current fault state array, a historical fault state matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records respectively includes: Iterate through the various working states involved in the current fault phenomenon and divide the various working states into working states supported by sensor data and working states without sensor data support. The state ratio value is calculated based on the current sensor value of the working state supported by sensor data and the normal value of the working state supported by sensor data when the device is working normally. A hierarchical quantization strategy is used to quantify the working state without sensor data support, and a current fault state array is constructed based on the quantization results and the state ratio values. Based on the current fault phenomenon and the historical repair records, construct a historical fault status matrix and a historical fault cell array, respectively.

[0007] In one embodiment, the step of constructing a historical fault state matrix and a historical fault cell array based on the current fault phenomenon and the historical repair records respectively includes: Search the historical repair records for valid cases involving various working conditions related to the current fault phenomenon; The various working states of the fault phenomena in each of the effective cases are proportionally quantified, and the proportional quantization results are saved in the initial fault state matrix in the form of column vectors to construct the historical fault state matrix. Obtain the fault unit number of the valid case and save the number to the initial fault unit array to construct the historical fault unit array.

[0008] In one embodiment, the step of calculating the fault investigation weight coefficient of each unit in the device based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, the traversed and updated current fault state array, and the historical fault unit array includes: The feature matrix of the historical fault state is calculated based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, and the traversed and updated historical fault state matrix. The feature vector of the current fault is calculated based on the matrix composed of the left singular vectors of the decomposition, the transpose of the matrix composed of the left singular vectors of the decomposition, and the updated current fault state array. The historical fault cell array is traversed element by element, and the current identification information of each traversed element in the historical fault cell array is obtained. Based on the current identification information, determine the column vectors corresponding to the feature matrix of the historical fault states, and calculate the correlation degree between each column vector and the feature vector of the current fault. The maximum correlation degree is selected from all the correlation degrees, and the maximum correlation degree is used as the fault diagnosis weight coefficient of each unit in the device.

[0009] In one embodiment, the step of determining the target fault diagnosis order based on the identification information of each of the fault diagnosis weight coefficients, and optimizing the fault diagnosis order in the maintenance manual based on the target fault diagnosis order, includes: The identification information of each of the aforementioned fault investigation weight coefficients is sorted according to a preset order; The order of troubleshooting targets is determined based on the sorted identification information. If there is no fault troubleshooting order in the maintenance manual that is strongly associated with the target fault troubleshooting order, then the fault troubleshooting order in the maintenance manual will be supplemented and optimized. If the maintenance manual contains a troubleshooting order that is strongly correlated with the target troubleshooting order, then the troubleshooting order in the maintenance manual shall be revised and optimized.

[0010] Secondly, this application provides a fault diagnosis sequence optimization device based on historical repair records, comprising: The acquisition module is used to acquire the current fault phenomenon and historical repair records of the device when a fault is detected. The construction module is used to construct a current fault state array, a historical fault state matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively. The calculation module is used to calculate the fault investigation weight coefficient of each unit in the device based on the current fault status array, the historical fault status matrix and the historical fault unit array; The optimization module is used to determine the target fault diagnosis order based on the identification information of each fault diagnosis weight coefficient, and to optimize the fault diagnosis order in the maintenance manual based on the target fault diagnosis order. The calculation module is further configured to calculate the modulus of the current fault state array, and traverse and update the elements in the current fault state array according to the modulus of the current fault state array; calculate the modulus of each column vector of the historical fault state matrix, and traverse and update the elements in the historical fault state matrix according to the modulus of each column vector of the historical fault state matrix; decompose the traversed and updated historical fault state matrix; and calculate the fault investigation weight coefficient of each unit in the device according to the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, and the traversed and updated current fault state array.

[0011] Thirdly, this application provides an electronic device, comprising: at least one memory for storing a program; and at least one processor for executing the program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to execute the method described in the first aspect or any possible implementation thereof.

[0012] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0013] Fifthly, this application provides a computer program product that, when run on a processor, causes the processor to perform the method described in the first aspect or any possible implementation thereof.

[0014] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0015] Overall, the technical solutions conceived in this application have the following beneficial effects compared with the prior art: (1) When a fault is detected in the equipment, this application introduces historical repair records. Since the repair cases record more and more fault phenomena and fault causes, the fault phenomena imply the relationship between the working principle and unit structure composition of the equipment. Therefore, this application makes full use of historical repair records to construct the current fault status array, the historical fault status matrix and the historical fault unit array. The target fault troubleshooting order is determined according to the identification information of each fault troubleshooting weight coefficient. The fault troubleshooting order in the maintenance manual is optimized according to the target fault troubleshooting order, thereby effectively improving the accuracy of optimizing the fault troubleshooting order, thereby improving the efficiency of fault troubleshooting and completing equipment repair more quickly.

[0016] (2) Since the sensor records a certain attribute value of the device, such as the speed of the car engine, the temperature of the coolant, the speed of the car, etc., the physical dimensions of these states are generally different. For working states supported by sensor data, in order to eliminate the influence of physical dimensions, this application calculates the state ratio value based on the current sensor value of the working state supported by sensor data and the normal value of the working state supported by sensor data when the device is working normally. Specifically, the current sensor value is divided by the normal value. For working states without sensor data support, a hierarchical quantization strategy can be used to quantify the working states without sensor data support.

[0017] In summary, when a device malfunction is detected, this application acquires the current malfunction phenomenon and historical repair records of the device; constructs a current malfunction state array, a historical malfunction state matrix, and a historical malfunction unit array based on the current malfunction phenomenon and the historical repair records; calculates the fault investigation weight coefficient for each unit in the device based on the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array; determines the target fault investigation order based on the identification information of each fault investigation weight coefficient, and optimizes the fault investigation order in the maintenance manual based on the target fault investigation order. Through this method, historical repair records are fully utilized, and the current malfunction phenomenon is combined to construct the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array. After determining the target fault investigation order based on the identification information of the fault investigation weight coefficients, the fault investigation order is optimized, thereby effectively improving the accuracy of the optimized fault investigation order and thus improving the efficiency of fault investigation. Attached Figure Description

[0018] Figure 1 This is one of the flowcharts illustrating the fault diagnosis sequence optimization method based on historical repair records provided in this application embodiment; Figure 2This is a schematic diagram of the probability distribution simulation results of the number of investigations provided in the embodiments of this application; Figure 3 This is the second flowchart of the fault diagnosis order optimization method based on historical repair records provided in the embodiments of this application; Figure 4 This is a schematic diagram of the module structure of the fault diagnosis sequence optimization device based on historical repair records provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] In this article, the term "and / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The symbol " / " in this article indicates that the related objects are in an "or" relationship; for example, A / B means A or B.

[0021] The terms "first" and "second," etc., used in the specification and claims herein are used to distinguish different objects, not to describe a specific order of objects. For example, "first response message" and "second response message," etc., are used to distinguish different response messages, not to describe a specific order of response messages.

[0022] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0023] Based on this, embodiments of this application provide a method for optimizing the fault diagnosis sequence based on historical repair records, referring to... Figure 1 , Figure 1 This is one of the flowcharts illustrating the fault diagnosis order optimization method based on historical repair records provided in this application. In this embodiment, the fault diagnosis order optimization method based on historical repair records includes steps S10 to S40: Step S10: When a device malfunction is detected, the current malfunction phenomenon and historical repair records of the device are obtained.

[0024] It should be noted that this embodiment stipulates that there is only one cause for equipment failure, meaning that multiple units will not fail simultaneously. Furthermore, complex equipment is often equipped with multiple sensors to reflect various operating states in a timely manner. This embodiment refers to the set of operating state values ​​when equipment failure occurs as the "fault phenomenon." Therefore, when a equipment failure is detected, the current fault phenomenon and historical repair records can be obtained.

[0025] It should be understood that after the equipment is put into use, the number of actual repair cases increases continuously, and these repair cases record more and more fault phenomena and causes. These fault phenomena imply the relationship between the working principle and unit structure of the equipment. Therefore, this embodiment makes full use of historical repair records to make up for the deficiencies of the maintenance manual, further improve the efficiency of fault diagnosis, and complete equipment repair more quickly.

[0026] Step S20: Construct a current fault status array, a historical fault status matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively.

[0027] Understandably, the current fault state array refers to the array transformed from various working states involved in the current fault phenomenon. The historical fault state matrix stores the state ratio values ​​of various working states related to the current fault phenomenon. Each column corresponds to a valid case. The historical fault unit array stores the fault unit number of the valid case.

[0028] Further, step S20 includes: traversing various working states involved in the current fault phenomenon, and dividing the various working states into working states supported by sensor data and working states without sensor data support; calculating a state ratio value based on the current sensor values ​​of the working states supported by sensor data and the normal values ​​of the working states supported by sensor data when the equipment is working normally; quantizing the working states without sensor data support using a hierarchical quantization strategy, and constructing a current fault state array based on the quantization results and the state ratio value; and constructing a historical fault state matrix and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively.

[0029] It should be noted that, generally speaking, sensors record a certain attribute value of a device, such as the engine speed of a car, the temperature of the coolant, and the speed of the car. The physical dimensions of these states are generally different. For working states supported by sensor data, in order to eliminate the influence of physical dimensions, a state ratio value is calculated based on the current sensor value of the working state supported by sensor data and the normal value of the working state supported by sensor data when the device is working normally. Specifically, the current sensor value is divided by the normal value, where the normal value can be the average value of the device over a period of normal operation. In short, the state values ​​of each fault phenomenon can be state ratio values, reflecting the degree of change of each state of the device during a fault relative to its normal operation.

[0030] It should be understood that for working states without sensor data support, a hierarchical quantification strategy can be used to quantify the working states without sensor data support. For example, 1 indicates that the state has changed slightly, 2 indicates that the state has changed to a moderate degree, and 3 indicates that the state has changed significantly. In this embodiment, the hierarchical description is set to no more than 4 levels.

[0031] Furthermore, the step of constructing a historical fault state matrix and a historical fault unit array based on the current fault phenomenon and the historical repair records respectively includes: searching for valid cases of various working states involved in the current fault phenomenon from the historical repair records; quantizing the various working states of the fault phenomenon of each valid case proportionally, and saving the proportional quantization results in the initial fault state matrix in the form of a column vector to construct the historical fault state matrix; obtaining the fault unit number of the valid case, and saving the number in the initial fault unit array to construct the historical fault unit array.

[0032] Understandably, in practical applications, the number of sensors is likely greater than the number of states involved in the current fault phenomenon. That is, the set of states involved in the fault phenomenon is mostly a subset of the total set of device states. Therefore, using the m states involved in the current fault phenomenon as a standard, cases containing m state information at the time of the fault are identified from historical repair records as valid cases. The numerical values ​​of these m state types are extracted and converted into the ratio of fault-time to normal-time values. After quantifying the various states of the fault phenomenon in each valid case, these values ​​are saved as column vectors in the initial fault state matrix to construct the historical fault state matrix M. The number of valid cases is denoted as n. The fault unit numbers of the valid cases are saved to the initial fault unit array to construct the historical fault unit array N. It is agreed that the historical fault unit array N involves k types of units. The column vector of fault phenomena of the same fault unit is placed in the adjacent columns of the historical fault state matrix M. For example, if unit 1 has 1 valid case and unit 2 has 2 valid cases, then the first column of the historical fault state matrix M is the fault phenomenon of fault unit 1, and the second and third columns of the historical fault state matrix M are the fault phenomena of fault unit 2. The corresponding elements of the historical fault unit array N are N1=1, N2=2, and N3=2.

[0033] It should be noted that, taking a fault phenomenon involving 10 states as an example, i.e., the number of fault phenomenon states m=10, for ease of discussion, the working states in the current fault phenomenon are directly converted into the ratios to normal states. In this case, the historical fault cell array N=[1,1,2,2,3,4,4,4,5,5,5,6] corresponding to the 12 valid cases, where the constructed current fault state array, historical fault state matrix, and historical fault cell array can be referred to Table 1, specifically as follows: Table 1:

[0034] Step S30: Calculate the fault investigation weight coefficient of each unit in the device based on the current fault status array, the historical fault status matrix, and the historical fault unit array.

[0035] It should be understood that the fault diagnosis weighting coefficient refers to the weighting coefficient for fault diagnosis of each unit in the equipment. The larger the fault diagnosis weighting coefficient, the higher the priority of the corresponding unit for fault diagnosis. For example, if the fault diagnosis weighting coefficients of unit 1, unit 2, and unit 3 in the equipment are 0.4, 0.3, and 0.8 respectively, then unit 3 will be diagnosed first, followed by unit 1, and finally unit 2.

[0036] Step S40: Determine the target fault troubleshooting order based on the identification information of each of the fault troubleshooting weight coefficients, and optimize the fault troubleshooting order in the maintenance manual based on the target fault troubleshooting order.

[0037] It should be noted that the identification information refers to the information that can identify each unit in the equipment. This identification information can be the unit number of the subscript. For example, if the unit numbers of the subscript are 1, 2, 3 and 4, the fault diagnosis weight coefficients of each unit in the equipment are 0.4, 0.3, 0.8 and 0.9, respectively. The final determined target fault diagnosis order is 4, 3, 1 and 2. That is, the fault diagnosis order in the maintenance manual can be optimized according to the target fault diagnosis order, and the fault diagnosis can be further completed.

[0038] Further, step S40 includes: sorting the identification information of each of the fault diagnosis weight coefficients according to a preset order; determining the target fault diagnosis order based on the sorted identification information; if there is no fault diagnosis order in the maintenance manual that is strongly correlated with the target fault diagnosis order, then supplementing and optimizing the fault diagnosis order in the maintenance manual; if there is a fault diagnosis order in the maintenance manual that is strongly correlated with the target fault diagnosis order, then revising and optimizing the fault diagnosis order in the maintenance manual.

[0039] Understandably, after calculating the fault diagnosis weight coefficients of each unit in the equipment, the identification information of each fault diagnosis weight coefficient can be sorted according to a preset order, which can be from largest to smallest. For the maintenance manual, it is necessary to determine whether there is a fault diagnosis order that is strongly correlated with the target fault diagnosis order. If not, it means that the maintenance manual does not have a fault diagnosis order for the faulty unit at all. In this case, the target fault diagnosis order is added to the maintenance manual to complete the optimization of the maintenance manual. Conversely, it means that the maintenance manual has a fault diagnosis order for the faulty unit that is difficult to find quickly. In this case, the fault diagnosis order is revised to complete the optimization of the maintenance manual.

[0040] It is important to emphasize that this embodiment also includes a simulation verification mechanism. By simulating the fault phenomena of the equipment multiple times, the above-mentioned technical solution is used to determine the target fault troubleshooting order, and the number of times the faulty component is found is recorded. Finally, the probability distribution of the number of troubleshooting attempts is statistically analyzed. Taking the above example, the simulation results of the probability distribution of the number of troubleshooting attempts can be referenced. Figure 2 Specifically: in Figure 2The document also lists the probability distribution of the number of inspections using the existing random order. By comparison, it can be seen that the probability of finding the faulty unit in this embodiment with only one inspection is about 0.60, and the probability of finding the faulty unit with no more than two inspections is about 0.82. Compared with the random order, the effect of optimizing the fault inspection order using the technical solution of this embodiment is extremely significant.

[0041] It should be understood that when compiling maintenance manuals, the logic of "starting from the equipment principle and deducing the fault phenomenon" is used to establish the association between fault phenomena and faulty units, and then to provide a troubleshooting order for possible faulty units. However, it is very difficult to fully implement the above logic on complex equipment, and it is difficult to cover or determine all fault phenomena. With the rapid development of AI (Artificial Intelligence) large language models, quickly extracting the fault state data mentioned in this embodiment from a large number of unstructured actual repair cases is no longer a time-consuming and laborious operation. This makes the technical solution of this embodiment feasible in terms of actual repair data foundation. When using multiple states to describe fault phenomena in this embodiment, it is not necessary to know in advance information such as "Is a certain state related to the fault?" or "If so, to what extent is the correlation?". It is only necessary to know that some of the states are related to the fault, thereby effectively reducing the difficulty of obtaining information and thus effectively improving the efficiency of fault diagnosis.

[0042] This embodiment, when a device malfunction is detected, acquires the current malfunction phenomenon and historical repair records of the device; constructs a current malfunction state array, a historical malfunction state matrix, and a historical malfunction unit array based on the current malfunction phenomenon and the historical repair records; calculates the fault investigation weight coefficient for each unit in the device based on the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array; determines the target fault investigation order based on the identification information of each fault investigation weight coefficient, and optimizes the fault investigation order in the maintenance manual based on the target fault investigation order. Through the above method, by fully utilizing historical repair records, combining the current malfunction phenomenon to construct the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array, and optimizing the fault investigation order after determining the target fault investigation order based on the identification information of the fault investigation weight coefficients, the accuracy of the optimized fault investigation order can be effectively improved, thereby increasing the efficiency of fault investigation.

[0043] In one specific implementation, this application provides steps for determining the fault diagnosis weighting coefficients for each unit in a computing device. Please refer to... Figure 3 , Figure 3This is the second flowchart illustrating the fault diagnosis sequence optimization method based on historical repair records provided in this application embodiment. Step S30 includes steps S301 to S304: Step S301: Calculate the modulus of the current fault state array, and iterate through and update the elements in the current fault state array according to the modulus of the current fault state array.

[0044] It should be noted that after constructing the current fault state array, the historical fault state matrix, and the historical fault cell array, normalization is required for both the current fault state array and the historical fault state matrix to achieve element traversal and updates. For the current fault state array, its modulus can be calculated first, specifically as follows: .

[0045] in, This represents the modulus of the current fault state array. This represents each element in the current fault status array.

[0046] It should be noted that after calculating the modulus of the current fault state array, the elements in the current fault state array can be iterated and updated based on the modulus of the current fault state array, specifically as follows: .

[0047] in, This represents the modulus of the current fault state array. This represents each element in the current fault status array.

[0048] Step S302: Calculate the modulus of each column vector of the historical fault state matrix, and update the elements in the historical fault state matrix by traversing according to the modulus of each column vector of the historical fault state matrix.

[0049] Understandably, for the historical fault state matrix, the same normalization process needs to be performed on each column of data. This involves calculating the modulus of each column vector in the historical fault state matrix and updating the elements of the matrix. For example, if a column vector in the historical fault state matrix is ​​denoted as... Y Then, traversing each updated element can be represented as: .

[0050] in, This represents the magnitude of each column vector in the historical fault state matrix. This represents each element in the historical fault state matrix.

[0051] It should be noted that the data for each element in the normalized current fault state array and historical fault state matrix can be found in Table 2. Table 2:

[0052] Step S303: Decompose the updated historical fault state matrix.

[0053] It should be understood that after traversing and updating the elements in the historical fault state matrix, the traversed and updated historical fault state matrix also needs to be decomposed, specifically as follows: .

[0054] Where M represents the historical fault state matrix, Let A denote the matrix composed of the left singular vectors, and let A denote the singular value matrix. This represents the matrix composed of right singular value vectors. This represents the identifier for the matrix transpose operation.

[0055] It should be noted that for matrices composed of left singular vectors, the data for each element can be found in Table 3; for singular value matrices, the data for each element can be found in Table 4; and for matrices composed of right singular value vectors, the data for each element can be found in Table 5. Specifically: Table 3:

[0056] Table 4:

[0057] Table 5:

[0058] Step S304: Calculate the fault investigation weight coefficient of each unit in the device based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, and the traversed and updated current fault state array.

[0059] Further, the step of calculating the fault investigation weight coefficient of each unit in the device based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, and the traversed and updated current fault state array includes: calculating the feature matrix of the historical fault state based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, and the traversed and updated historical fault state matrix; calculating the feature vector of the current fault based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, and the traversed and updated current fault state array; traversing the historical fault unit array element by element and obtaining the current identification information of each traversed element in the historical fault unit array; determining the column vector corresponding to the feature matrix of the historical fault state based on the current identification information, and calculating the correlation degree between each column vector and the feature vector of the current fault; selecting the maximum correlation degree from each correlation degree, and using the maximum correlation degree as the fault investigation weight coefficient of each unit in the device.

[0060] It should be understood that after decomposing the matrix composed of left singular vectors and the transpose of the matrix composed of the decomposed left singular vectors, the feature matrix of the historical fault state can be calculated by combining it with the updated historical fault state matrix. Specifically: .

[0061] in, The feature matrix representing historical fault states, This represents the matrix composed of left singular vectors. This represents the identifier for the matrix transpose operation, and M represents the historical fault state matrix.

[0062] It is understandable that after decomposing the matrix composed of the left singular vectors and the transpose of the matrix composed of the decomposed left singular vectors, the feature vector of the current fault can also be calculated by combining it with the updated current fault state array, specifically: .

[0063] in, The feature vector representing the current fault, This represents the matrix composed of left singular vectors. This represents the identifier for the matrix transpose operation. This represents the current fault status array.

[0064] It should be noted that, for the historical fault cell array, taking cell j as an example, all elements of cell j can be found, and based on the current identifier information of these elements in the historical fault cell array, all column vectors corresponding to the feature matrix of the historical fault state are determined. This current identifier information can be an index number. Then, the correlation degree between each column vector and the feature vector of the current fault is calculated sequentially, specifically: .

[0065] in, Indicates the degree of relevance. Represents a column vector. The feature vector representing the current fault.

[0066] It should be understood that after calculating multiple correlation degrees, the maximum correlation degree can be selected from among them, and the maximum correlation degree can be used as the fault investigation weight coefficient of unit j. For details, please refer to Table 6: Table 6:

[0067] This embodiment calculates the modulus of the current fault state array and updates the elements in the current fault state array by traversing based on the modulus of the current fault state array; calculates the modulus of each column vector of the historical fault state matrix and updates the elements in the historical fault state matrix by traversing based on the modulus of each column vector of the historical fault state matrix; decomposes the traversed and updated historical fault state matrix; calculates the fault investigation weight coefficient of each unit in the device based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, the traversed and updated current fault state array, and the historical fault unit array. Through the above method, after traversing and updating the elements in the current fault state array and the historical fault state matrix according to their respective moduli, the feature matrix of the historical fault state and the feature vector of the current fault are calculated by combining the matrix composed of the decomposed left singular vectors and the transpose matrix, and the maximum correlation between the column vector and the feature vector of the current fault is used as the fault investigation weight coefficient of the unit, thereby effectively improving the accuracy of calculating the fault investigation weight coefficient.

[0068] The following describes the fault diagnosis order optimization device based on historical repair records provided in this application. The fault diagnosis order optimization device described below corresponds to the fault diagnosis order optimization method described above. Please refer to... Figure 4 , Figure 4This is a schematic diagram of the module structure of the fault diagnosis sequence optimization device based on historical repair records provided in this application embodiment, including: The acquisition module T10 is used to acquire the current fault phenomenon and historical repair records of the device when a fault is detected.

[0069] The construction module T20 is used to construct a current fault state array, a historical fault state matrix, and a historical fault unit array based on the current fault phenomenon and the historical repair records, respectively.

[0070] The calculation module T30 is used to calculate the fault investigation weight coefficient of each unit in the device based on the current fault status array, the historical fault status matrix and the historical fault unit array.

[0071] The optimization module T40 is used to determine the target fault diagnosis order based on the identification information of each fault diagnosis weight coefficient, and to optimize the fault diagnosis order in the maintenance manual based on the target fault diagnosis order.

[0072] This embodiment, when a device malfunction is detected, acquires the current malfunction phenomenon and historical repair records of the device; constructs a current malfunction state array, a historical malfunction state matrix, and a historical malfunction unit array based on the current malfunction phenomenon and the historical repair records; calculates the fault investigation weight coefficient for each unit in the device based on the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array; determines the target fault investigation order based on the identification information of each fault investigation weight coefficient, and optimizes the fault investigation order in the maintenance manual based on the target fault investigation order. Through the above method, by fully utilizing historical repair records, combining the current malfunction phenomenon to construct the current malfunction state array, the historical malfunction state matrix, and the historical malfunction unit array, and optimizing the fault investigation order after determining the target fault investigation order based on the identification information of the fault investigation weight coefficients, the accuracy of the optimized fault investigation order can be effectively improved, thereby increasing the efficiency of fault investigation.

[0073] It is understood that the detailed functional implementation of each of the above modules can be found in the description of the aforementioned method embodiments, and will not be repeated here.

[0074] It should be understood that the above-described device is used to execute the methods in the above embodiments. The implementation principle and technical effect of the corresponding program modules in the device are similar to those described in the above methods. The working process of the device can be referred to the corresponding process in the above methods, and will not be repeated here.

[0075] Based on the methods in the above embodiments, this application provides an electronic device, please refer to... Figure 5 , Figure 5This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application.

[0076] It should be noted that the system may include: a processor 10, a communications interface 20, a memory 30, and a communication bus 40. The processor 10, communications interface 20, and memory 30 communicate with each other via the communication bus 40. The processor 10 can invoke logical instructions stored in the memory 30 to execute the methods described in the above embodiments.

[0077] Furthermore, the logical instructions in the aforementioned memory 30 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0078] Based on the methods in the above embodiments, this application provides a computer-readable storage medium storing a computer program that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0079] Based on the methods in the above embodiments, this application provides a computer program product that, when run on a processor, causes the processor to execute the methods in the above embodiments.

[0080] It is understood that the processor in the embodiments of this application can be a central processing unit, or other general-purpose processors, digital signal processors, application-specific integrated circuits, field-programmable gate arrays, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. A general-purpose processor can be a microprocessor or any conventional processor.

[0081] The method steps in this application embodiment can be implemented in hardware or by a processor executing software instructions. The software instructions can consist of corresponding software modules, which can be stored in random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, registers, hard disks, portable hard disks, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium can also be a component of the processor.

[0082] It is understood that the various numerical designations used in the embodiments of this application are merely for descriptive convenience and are not intended to limit the scope of the embodiments of this application. Those skilled in the art will readily understand that the above descriptions are merely preferred embodiments of this application and are not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for optimizing the fault diagnosis sequence based on historical repair records, characterized in that, include: When a device malfunction is detected, the current malfunction phenomenon and historical repair records of the device are obtained; Construct a current fault status array, a historical fault status matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively. The fault investigation weight coefficient of each unit in the device is calculated based on the current fault status array, the historical fault status matrix, and the historical fault unit array. The target fault diagnosis order is determined based on the identification information of each fault diagnosis weight coefficient, and the fault diagnosis order in the maintenance manual is optimized based on the target fault diagnosis order. The step of calculating the fault investigation weight coefficient of each unit in the device based on the current fault state array, the historical fault state matrix, and the historical fault unit array includes: Calculate the modulus of the current fault state array, and iterate through and update the elements in the current fault state array according to the modulus of the current fault state array; Calculate the modulus of each column vector of the historical fault state matrix, and iterate through and update the elements in the historical fault state matrix according to the modulus of each column vector of the historical fault state matrix. Decompose the updated historical fault state matrix; The fault investigation weight coefficient of each unit in the device is calculated based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, the traversed and updated current fault state array, and the historical fault unit array.

2. The method as described in claim 1, characterized in that, The steps of constructing a current fault state array, a historical fault state matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records respectively include: Iterate through the various working states involved in the current fault phenomenon and divide the various working states into working states supported by sensor data and working states without sensor data support. The state ratio value is calculated based on the current sensor value of the working state supported by sensor data and the normal value of the working state supported by sensor data when the device is working normally. A hierarchical quantization strategy is used to quantify the working state without sensor data support, and a current fault state array is constructed based on the quantization results and the state ratio values. Based on the current fault phenomenon and the historical repair records, construct a historical fault status matrix and a historical fault cell array, respectively.

3. The method as described in claim 2, characterized in that, The steps of constructing a historical fault state matrix and a historical fault cell array based on the current fault phenomenon and the historical repair records respectively include: Search the historical repair records for valid cases involving various working conditions related to the current fault phenomenon; The various working states of the fault phenomena in each of the effective cases are proportionally quantified, and the proportional quantization results are saved in the initial fault state matrix in the form of column vectors to construct the historical fault state matrix. Obtain the fault unit number of the valid case and save the number to the initial fault unit array to construct the historical fault unit array.

4. The method as described in claim 1, characterized in that, The step of calculating the fault investigation weight coefficient of each unit in the device based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the updated historical fault state matrix, the updated current fault state array, and the historical fault unit array includes: The feature matrix of the historical fault state is calculated based on the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, and the traversed and updated historical fault state matrix. The feature vector of the current fault is calculated based on the matrix composed of the left singular vectors of the decomposition, the transpose of the matrix composed of the left singular vectors of the decomposition, and the updated current fault state array. The historical fault cell array is traversed element by element, and the current identification information of each traversed element in the historical fault cell array is obtained. Based on the current identification information, determine the column vectors corresponding to the feature matrix of the historical fault states, and calculate the correlation degree between each column vector and the feature vector of the current fault. The maximum correlation degree is selected from all the correlation degrees, and the maximum correlation degree is used as the fault diagnosis weight coefficient of each unit in the device.

5. The method according to any one of claims 1 to 4, characterized in that, The step of determining the target fault diagnosis order based on the identification information of each fault diagnosis weight coefficient, and optimizing the fault diagnosis order in the maintenance manual based on the target fault diagnosis order, includes: The identification information of each of the aforementioned fault investigation weight coefficients is sorted according to a preset order; The order of troubleshooting targets is determined based on the sorted identification information. If there is no fault troubleshooting order in the maintenance manual that is strongly associated with the target fault troubleshooting order, then the fault troubleshooting order in the maintenance manual will be supplemented and optimized. If the maintenance manual contains a troubleshooting order that is strongly correlated with the target troubleshooting order, then the troubleshooting order in the maintenance manual shall be revised and optimized.

6. A fault diagnosis sequence optimization device based on historical repair records, characterized in that, include: The acquisition module is used to acquire the current fault phenomenon and historical repair records of the device when a fault is detected. The construction module is used to construct a current fault state array, a historical fault state matrix, and a historical fault cell array based on the current fault phenomenon and the historical repair records, respectively. The calculation module is used to calculate the fault investigation weight coefficient of each unit in the device based on the current fault status array, the historical fault status matrix and the historical fault unit array; The optimization module is used to determine the target fault diagnosis order based on the identification information of each fault diagnosis weight coefficient, and to optimize the fault diagnosis order in the maintenance manual based on the target fault diagnosis order. The calculation module is also used to calculate the modulus of the current fault state array, and to traverse and update the elements in the current fault state array according to the modulus of the current fault state array. Calculate the modulus of each column vector of the historical fault state matrix, and iterate through and update the elements in the historical fault state matrix according to the modulus of each column vector of the historical fault state matrix; decompose the traversed and updated historical fault state matrix; calculate the fault investigation weight coefficient of each unit in the device according to the matrix composed of the decomposed left singular vectors, the transpose of the matrix composed of the decomposed left singular vectors, the traversed and updated historical fault state matrix, the traversed and updated current fault state array, and the historical fault unit array.

7. An electronic device, characterized in that, include: At least one memory for storing computer programs; At least one processor is configured to execute a program stored in the memory, wherein when the program stored in the memory is executed, the processor is configured to perform the method as described in any one of claims 1-5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is run on the processor, it causes the processor to perform the method as described in any one of claims 1-5.

9. A computer program product, characterized in that, When the computer program product is run on a processor, the processor causes the processor to perform the method as described in any one of claims 1-5.

Images