Main-standby difference comparison method and device for industrial system, terminal equipment and storage medium

By constructing a weight distribution matrix and a neural network model, the problem of control disturbances after the switching of primary and backup equipment in the existing technology is solved, and the automatic comparison and advance perception of differences between primary and backup equipment are realized.

CN121479326APending Publication Date: 2026-02-06POWER DISPATCHING CONTROL CENT OF GUANGDONG POWER GRID CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511614044.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-11-06
Publication Date
2026-02-06

AI Technical Summary

Technical Problem

Existing technologies cannot detect control disturbances after the switching of primary and backup equipment in advance, mainly because the hidden hidden dangers inside the equipment have nonlinear characteristics, and existing methods cannot effectively handle huge amounts of data.

Method used

By acquiring the operating data of the primary and backup devices, constructing the first and second weight distribution matrices, performing weighted deviation calculations, and training the primary-backup difference comparison model using a single hidden layer neural network, the automated difference comparison of the primary and backup devices is achieved.

Benefits of technology

It enables automated comparison of differences between primary and backup devices, can quickly process massive amounts of data, and can detect control disturbances after the switchover between primary and backup devices in advance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121479326A_ABST
    Figure CN121479326A_ABST
Patent Text Reader

Abstract

The invention discloses a main-standby difference comparison method and device for an industrial system, terminal equipment and a storage medium, and belongs to the field of industrial data processing, and the method comprises the steps: obtaining the operation data of main equipment and standby equipment of a to-be-tested industrial system; determining a first weight distribution matrix according to the operation data; based on the first weight distribution matrix, performing weighted deviation calculation on the operation data to obtain an element service duration deviation matrix; determining a second weight distribution matrix according to the operation data; weighting the operation data based on the second weight distribution matrix to obtain an element maintenance fault deviation matrix; and comparing the element service time length deviation matrix with the element maintenance fault deviation matrix to obtain a difference comparison result between the master device and the slave device, so that the problem that control disturbance after switching of the master device and the slave device cannot be sensed in advance in the prior art can be solved.
Need to check novelty before this filing date? Find Prior Art

Description

TECHNICAL FIELD

[0001] The present application relates to the field of industrial data processing, and in particular to a master-slave difference comparison method and device for an industrial system, a terminal device and a storage medium. BACKGROUND

[0002] In an industrial system, in order to ensure the continuous and stable operation of key equipment, a redundant design of master-slave equipment is often used, that is, when the master equipment is abnormal, the standby equipment is switched to maintain the system operation. However, after the master-slave switching of the equipment, once control disturbances such as power fluctuations and unstable system operation occur, a warning and response can only be made after the problem occurs. This is because the hidden problems in the equipment have nonlinear characteristics, such as the slow degradation of electronic components and the microscopic fatigue damage of mechanical parts, and their development and evolution process is complex, and there is not a simple linear relationship between the equipment operating state. The prior art usually compares the differences between the master and slave equipment by manual means, but since there is a large amount of data processing in the equipment comparison, the amount of data that can be processed by manual comparison is much smaller than the amount of data that needs to be processed in the equipment comparison. Therefore, the prior art cannot pre-knowledge the control disturbance after the master-slave switching of the equipment. SUMMARY

[0003] The present application provides a master-slave difference comparison method and device for an industrial system, a terminal device and a storage medium, which can solve the problem that the prior art cannot pre-knowledge the control disturbance after the master-slave switching of the equipment.

[0004] The master-slave difference comparison method for an industrial system provided by the present application comprises:

[0005] Obtaining the running data of the master equipment and the standby equipment of the industrial system to be tested; wherein the running data comprises: the master equipment model, the master equipment element service scene, the master equipment element service time information, the master equipment element model, the master equipment element maintenance fault information, the standby equipment model, the standby equipment element service scene, the standby equipment element service time information, the standby equipment element model and the standby equipment element maintenance fault information;

[0006] According to the master equipment model, the master equipment element service scene, the standby equipment model and the standby equipment element service scene, a first weight distribution matrix is determined;

[0007] Based on the first weight distribution matrix, the master equipment element service time information and the standby equipment element service time information are weighted and deviated to obtain an element service time deviation matrix;

[0008] According to the master equipment element model and the standby equipment element model, a second weight distribution matrix is determined;

[0009] Based on the second weight distribution matrix, the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component are weighted to obtain the component maintenance fault deviation matrix.

[0010] The service life deviation matrix of the component and the maintenance failure deviation matrix of the component are compared to obtain the difference comparison results between the master equipment and the slave equipment.

[0011] Further, determining the first weight distribution matrix based on the main equipment model, the service scenario of the main equipment component, the backup equipment model, and the service scenario of the backup equipment component includes:

[0012] Get the list of equipment components corresponding to the main device and the backup device;

[0013] Extract the first device component tag number from the device component list, and collect device operating status record data when the first device component tag number is in a fault state, constrained by the main device model, backup device model, main device component service scenario, and backup device component service scenario; wherein, the device operating status record data includes: health status and fault status;

[0014] The percentage of records in the device's operating status that are in the fault state is statistically analyzed and set as the first initial weight for inducing a fault. If the first initial weight for inducing a fault is greater than or equal to the inducing weight threshold, it is added to the set of initial weights for inducing a fault; otherwise, the weight for inducing a fault by the first device component number is set to 0 and added to the first weight distribution matrix.

[0015] Further, the step of calculating the weighted deviation of the service duration information of the main equipment component and the service duration information of the backup equipment component based on the first weight distribution matrix to obtain the component service duration deviation matrix includes:

[0016] Extract the service duration list of the main equipment component and the service duration list of the backup equipment component from the service duration information of the main equipment component and the service duration information of the backup equipment component;

[0017] Calculate the service duration deviation vector matrix of components with the same position number in the main equipment component service duration list and the backup equipment component service duration list;

[0018] Based on the first weight distribution matrix, the service life deviation vector matrix of the corresponding numbered components is weighted to obtain the component service life deviation matrix.

[0019] Further, determining the second weight distribution matrix based on the main equipment component model and the backup equipment component model includes:

[0020] Obtain the maintenance fault list corresponding to the main equipment and the backup equipment;

[0021] Based on the main equipment component model and the backup equipment component model, extract the first component model;

[0022] In the list of maintenance faults, the first maintenance fault type is extracted, and the damage rating table bound to the first component model is input to obtain the damage rating of the first maintenance fault type. The damage rating table is a user-predefined data table that stores fault types and damage levels. The higher the damage level, the higher the degree of damage.

[0023] Calculate the ratio of the damage rating of the first maintenance fault type to the sum of the damage ratings, set it as the damage weight of the first maintenance fault type of the first component model, and add it to the second weight distribution matrix.

[0024] Further, the step of weighting the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component based on the second weight distribution matrix to obtain the component maintenance fault deviation matrix includes:

[0025] Based on the main equipment component maintenance fault information, extract the main equipment component model, main equipment component maintenance fault type, and main equipment component maintenance fault frequency.

[0026] Based on the fault information of the backup equipment components, extract the model of the faulty backup equipment component, the fault type of the backup equipment component, and the fault frequency of the backup equipment component.

[0027] Based on the main equipment faulty component model, main equipment fault type, main equipment fault frequency, standby equipment faulty component model, standby equipment fault type, and standby equipment fault frequency, a comparison of the same model and fault type is performed, and the deviation between the standby equipment fault frequency and the main equipment fault frequency is calculated to obtain an initial component fault deviation matrix; wherein, when the main equipment data or standby equipment data for the compared fault frequency is empty, zero padding is performed;

[0028] Based on the second weight distribution matrix, the initial component maintenance fault deviation matrix is ​​weighted by the same model and fault type to obtain the component maintenance fault deviation matrix.

[0029] Further, the comparison of the component service life deviation matrix and the component maintenance failure deviation matrix to obtain the difference comparison result between the master equipment and the slave equipment includes:

[0030] The component service life deviation matrix and the component maintenance fault deviation matrix are input into a preset master-slave difference comparison model for comparison, and the difference comparison results between the master equipment and the slave equipment output by the master-slave difference comparison model are obtained.

[0031] Furthermore, the training of the primary / backup difference comparison model includes:

[0032] Constrained by the main equipment model, backup equipment model, main equipment component service scenario, and backup equipment component service scenario, a component service duration deviation record matrix and a component maintenance fault deviation record matrix with switching result identifiers are collected; wherein, the switching result identifier represents the historical switching control state. When the control result does not meet the working requirements, the switching result identifier is uncontrollable, and when the control result meets the working requirements, the switching result identifier is controllable.

[0033] Using the switching result identifier as supervision, and the component service duration deviation record matrix and the component maintenance fault deviation record matrix as input, a single-hidden-layer neural network is integrated. The neural network with the most votes among the single-hidden-layer neural networks is used as output. Training is performed on the single-hidden-layer neural networks to obtain the master-slave difference comparison model.

[0034] Another embodiment of the present invention provides a primary / backup difference comparison device for an industrial system, comprising: a data acquisition module, a first weight module, a first matrix calculation module, a second weight module, a second matrix calculation module, and a result generation module;

[0035] The data acquisition module is used to acquire the operating data of the main equipment and backup equipment of the industrial system under test; wherein, the operating data includes: main equipment model, main equipment component service scenario, main equipment component service duration information, main equipment component model, main equipment component maintenance and fault information, backup equipment model, backup equipment component service scenario, backup equipment component service duration information, backup equipment component model, and backup equipment component maintenance and fault information.

[0036] The first weighting module is used to determine a first weighting distribution matrix based on the main equipment model, the service scenario of the main equipment component, the backup equipment model, and the service scenario of the backup equipment component;

[0037] The first matrix calculation module is used to perform weighted deviation calculation on the service duration information of the main equipment component and the service duration information of the backup equipment component based on the first weight distribution matrix, so as to obtain the component service duration deviation matrix;

[0038] The second weighting module is used to determine a second weighting distribution matrix based on the main equipment component model and the backup equipment component model;

[0039] The second matrix calculation module is used to weight the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component based on the second weight distribution matrix to obtain the component maintenance fault deviation matrix.

[0040] The result generation module is used to compare the component service life deviation matrix and the component maintenance fault deviation matrix to obtain the difference comparison results between the master equipment and the slave equipment.

[0041] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the steps of the master-slave difference comparison method for industrial systems provided by the present invention.

[0042] Another embodiment of the present invention also provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the steps of the master-slave difference comparison method for industrial systems provided by the present invention.

[0043] The following benefits can be obtained by implementing the present invention:

[0044] This invention discloses a method for comparing the differences between primary and backup equipment in an industrial system. The method involves acquiring operational data of the primary and backup equipment of the industrial system under test. This operational data includes: primary equipment model, primary equipment component service scenario, primary equipment component service duration information, primary equipment component model, primary equipment component maintenance / fault information, backup equipment model, backup equipment component service scenario, backup equipment component service duration information, backup equipment component model, and backup equipment component maintenance / fault information. A first weight distribution is determined based on the primary equipment model, the primary equipment component service scenario, the backup equipment model, and the backup equipment component service scenario. The system employs a matrix-based approach. Based on the first weighted distribution matrix, a weighted deviation calculation is performed on the service duration information of the primary equipment component and the backup equipment component to obtain a component service duration deviation matrix. A second weighted distribution matrix is ​​determined according to the model numbers of the primary and backup equipment components. Based on the second weighted distribution matrix, the maintenance fault information of the primary equipment component and the backup equipment component is weighted to obtain a component maintenance fault deviation matrix. The component service duration deviation matrix and the component maintenance fault deviation matrix are compared to obtain a difference comparison result between the primary and backup equipment. This application achieves automated comparison of differences between primary and backup equipment, enabling rapid processing of the massive data volume involved in primary and backup equipment comparison, thereby obtaining the differences between the primary and backup equipment. Based on the obtained primary and backup equipment differences, it enables advance detection of control disturbances that may occur after primary and backup equipment switching. Attached Figure Description

[0045] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0046] Figure 1 This is a flowchart illustrating a primary / backup difference comparison method for an industrial system provided by an embodiment of the present invention.

[0047] Figure 2 This is a schematic diagram of the structure of a master-slave difference comparison device for an industrial system provided in an embodiment of the present invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0050] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0051] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0052] In the description of the embodiments in this application, the term "and / or" is merely a description of 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, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0053] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0054] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0055] See Figure 1 To address the problem that existing technologies cannot detect control disturbances after a switchover between primary and backup equipment in advance, an embodiment of the present invention provides a method for comparing primary and backup differences in industrial systems, comprising:

[0056] 101. Obtain the operating data of the main equipment and backup equipment of the industrial system under test; wherein, the operating data includes: main equipment model, main equipment component service scenario, main equipment component service duration information, main equipment component model, main equipment component maintenance fault information, backup equipment model, backup equipment component service scenario, backup equipment component service duration information, backup equipment component model, and backup equipment component maintenance fault information.

[0057] In a specific embodiment, service information of the main and backup equipment components is first obtained. This service information includes service duration information and maintenance / fault information of the main and backup equipment components. For example, the service information of the main and backup equipment components is fundamental data for assessing equipment reliability. It includes service duration information and maintenance / fault information of the main and backup equipment components. This information can be extracted from embedded sensors, the Device Management System (DMS), and the maintenance database through standardized interfaces. The service duration information includes the cumulative operating time of components in the main and backup equipment, used to quantify the progressive wear and tear of the components. The maintenance / fault information includes the historical fault types, frequencies, and last maintenance times of the main and backup equipment components. For example, the main equipment relay experiences an average of two contact oxidation events per year, and the same model component in the backup equipment experiences one coil open circuit due to long-term idleness, reflecting the historical fault information of the components.

[0058] Understandably, the service life information of the main and backup equipment components reflects the natural aging process of the equipment, while the maintenance and failure information of the main and backup equipment components reflects the hidden defects that the equipment may expose. The combination of the two constructs a health profile of the main and backup equipment components throughout their entire life cycle, which can provide early warning of the failure risk after the backup component is switched over, and provide a quantitative basis for subsequent comparison of differences.

[0059] 102. Determine the first weight distribution matrix based on the main equipment model, the service scenario of the main equipment component, the backup equipment model, and the service scenario of the backup equipment component.

[0060] In this embodiment, determining the first weight distribution matrix based on the main equipment model, the service scenario of the main equipment component, the backup equipment model, and the service scenario of the backup equipment component includes:

[0061] Get the list of equipment components corresponding to the main device and the backup device;

[0062] Extract the first device component tag number from the device component list, and collect device operating status record data when the first device component tag number is in a fault state, constrained by the main device model, backup device model, main device component service scenario, and backup device component service scenario; wherein, the device operating status record data includes: health status and fault status;

[0063] The percentage of records in the device's operating status that are in the fault state is statistically analyzed and set as the first initial weight for inducing a fault. If the first initial weight for inducing a fault is greater than or equal to the inducing weight threshold, it is added to the set of initial weights for inducing a fault; otherwise, the weight for inducing a fault by the first device component number is set to 0 and added to the first weight distribution matrix.

[0064] In a specific embodiment, the sources of equipment operating status recording data are as follows: First, real-time monitoring and control systems, such as Supervisory Control and Data Acquisition (SCADA) and Distributed Control Systems (DCS), which capture the operating status of component tags in real time according to timestamps; second, equipment operation logs, such as operation logs generated by temperature, vibration, and other sensors on the components and PLCs, which can supplement the details of parameter changes and state switching before and after the fault occurs; third, equipment maintenance management systems, such as Computerized Maintenance Management Systems (CMMS) and Enterprise Asset Management Systems (EAM), where manually or automatically entered maintenance records clearly indicate the fault occurrence time, type, and handling result of the component tag, used to trace historical faults; and fourth, historical maintenance archives, such as paper or electronically archived maintenance reports, fault analysis documents, and maintenance handover records.

[0065] In a specific embodiment, firstly, a first equipment component tag number is extracted from the equipment component list. For example, the first equipment component tag number (e.g., P-001, M-102) is extracted from the equipment component list. This first equipment component tag number ensures that components of different equipment and different models can be accurately located; for example, it distinguishes between the main equipment compressor A (e.g., tag number C-001) and the backup equipment compressor B (e.g., tag number C-002). Further, the equipment component list is a list document formed by systematically sorting, identifying, and recording various components in the equipment. It contains key attribute information of the equipment components, such as the component tag number (a unique identifier for each component, used for accurate location and differentiation of different components, such as C-001), component type, and model (e.g., mechanical components, electronic components, etc.). Secondly, constrained by the equipment model and the service scenario number, equipment operating status record data for the first equipment component tag number in a fault state is collected. Each piece of equipment operating status record data includes both a healthy state and a fault state. Specifically, constrained by the equipment model (such as PLC model PLC-500) and service scenario number (such as high temperature environment S-003), the working status record data of this type of component in a specific scenario is collected. Each working status record data includes a health status and a fault status (such as 1 indicating normal and 0 indicating fault).

[0066] For example, constrained by the equipment model (e.g., PLC-500) and service scenario number (e.g., high-temperature environment S-003) of the first equipment component tag number, at least the following four types of data carriers can be used to collect equipment operation status record data where the first equipment component tag number is in a fault state: First, real-time monitoring and control systems, such as Supervisory Control and Data Acquisition (SCADA) and Distributed Control Systems (DCS); second, equipment operation logs, such as operation logs generated by temperature, vibration, and other sensors mounted on the component and by the PLC; third, equipment maintenance management systems, such as Computerized Maintenance Management Systems (CMMS) and Enterprise Asset Management Systems (EAM); and fourth, historical maintenance archives, such as paper or electronically archived maintenance reports, fault analysis documents, and maintenance handover records. For instance, 100 equipment operation status record data entries where the first equipment component tag number is in a fault state can be collected from historical maintenance archives, of which 20 entries show the fault state.

[0067] Next, the percentage of records in the equipment operating status data that are in the fault state is calculated and set as the first initial weight for inducing a fault. If the first initial weight for inducing a fault is greater than or equal to the inducing weight threshold, it is added to the initial weight set for inducing a fault; otherwise, the weight for inducing a fault by the first equipment component reference number is set to 0 and added to the first weight distribution matrix. Specifically, the first initial weight for inducing a fault = number of fault states / total number of records in the equipment operating status data. For example, if 100 records of equipment operating status data are collected, and 20 of them show a fault state, then the first initial weight for inducing a fault = 20 / 100 = 0.2. Further, the initial weight of the first induced fault is compared with the preset induced fault weight threshold. If the initial weight of the first induced fault is greater than or equal to the induced fault weight threshold, it is added to the set of initial induced fault weights. Otherwise, the induced fault weight of the first device component reference number is set to 0 and added to the first weight distribution matrix. The preset induced fault weight threshold is used to filter low-frequency fault components and focus on high-risk components. Those skilled in the art can dynamically set it according to the actual equipment type, service scenario, etc. For example, the induced fault weight threshold can be set to 20%. For example, the initial weights of the first induced faults of reference number P-001 and reference number P-002 are calculated to be 0.2 and 0.3, respectively, and the induced fault weight threshold is 20%. Since the initial weights of the first induced faults are both greater than or equal to the induced fault weight threshold, the initial weights of the first induced faults of 0.2 and 0.3 are added to the set of initial induced fault weights.

[0068] Finally, the initial set of induced fault weights is normalized and added to the first weight distribution matrix. After normalization, the sum of the weights in the first weight distribution matrix is ​​1. For example, the initial set of induced fault weights includes first induced fault weights of 0.2 and 0.3, which are normalized to: 0.2 / (0.2+0.3) = 0.4 and 0.3 / (0.2+0.3) = 0.6. Then, 0.4 and 0.6 are added to the first weight distribution matrix [0.4, 0.6]. The first weight reflects the probability of different components failing; the larger the first weight, the greater the probability of the component failing.

[0069] 103. Based on the first weight distribution matrix, perform weighted deviation calculation on the service duration information of the main equipment component and the service duration information of the backup equipment component to obtain the component service duration deviation matrix.

[0070] In this embodiment, the step of calculating the weighted deviation of the service duration information of the main equipment component and the service duration information of the backup equipment component based on the first weight distribution matrix to obtain the component service duration deviation matrix includes:

[0071] Extract the service duration list of the main equipment component and the service duration list of the backup equipment component from the service duration information of the main equipment component and the service duration information of the backup equipment component;

[0072] Calculate the service duration deviation vector matrix of components with the same position number in the main equipment component service duration list and the backup equipment component service duration list;

[0073] Based on the first weight distribution matrix, the service life deviation vector matrix of the corresponding numbered components is weighted to obtain the component service life deviation matrix.

[0074] It should be noted that there is a complex nonlinear relationship between equipment service life and failure rate. For example, in high-temperature and high-humidity chemical industry scenarios, electronic components experience accelerated aging, and the failure rate after 1000 hours of service can be three times that under normal temperature conditions. In wind power generation scenarios, mechanical components such as gearboxes, under high load fluctuations, show an exponential increase in the probability of failure due to fatigue wear for every 500 hours of service life. Therefore, when comparing the service life information of primary and backup equipment components, one cannot rely solely on a simple numerical comparison of service time. However, traditional methods often use service life as the sole criterion, directly comparing the cumulative operating hours of primary and backup equipment components, while ignoring the impact of different equipment models, service scenarios, and individual component characteristics on the aging rate. This approach lacks dynamic adaptability and cannot accurately capture the nonlinear degradation patterns of components under complex operating conditions.

[0075] In this embodiment, the service duration lists of the main equipment components and the backup equipment components are first extracted from the service duration information of the main and backup equipment components. For example, the service durations of the main equipment and the backup equipment are extracted from the service duration information and aligned according to the bit order in the first weight distribution matrix. For instance, the main equipment bit P-001 corresponds to a duration of 5000h, and bit P-002 corresponds to 3000h; the backup equipment bit P-001 corresponds to a duration of 1000h, and bit P-002 corresponds to 2000h. Thus, the aligned service duration lists of the main equipment (e.g., 5000h, 3000h) and the backup equipment (e.g., 1000h, 2000h) are obtained.

[0076] Next, the service duration deviation vector matrix of corresponding element numbers in the main equipment component service duration list and the backup equipment component service duration list is calculated. Specifically, the service duration of the first equipment component number in the main and backup equipment is processed by the first-number service attenuation parameter evaluator to obtain the first-number service attenuation parameter of the main and backup equipment. The ratio of the first-number service attenuation parameter of the main and backup equipment to the sum of the service attenuation parameters of the main and backup equipment is calculated and added to the service duration attenuation weight matrix of the main and backup equipment. Then, based on the service duration attenuation weight matrix of the main and backup equipment, the service duration list of the main and backup equipment components is preprocessed and weighted to obtain the corrected service duration list of the main and backup equipment components. Finally, the service duration deviation vector matrix of corresponding element numbers in the corrected service duration list of the main equipment components and the corrected service duration list of the backup equipment components is calculated.

[0077] Finally, based on the first weight distribution matrix, the service life deviation vector matrix of the corresponding component is weighted to obtain the component service life deviation matrix. Specifically, each element in the first weight distribution matrix is ​​multiplied by the corresponding element in the service life deviation vector matrix of the corresponding component to obtain the component service life deviation matrix, thus highlighting the deviation impact of high-risk components. For example, if the first weight distribution matrix is ​​[0.4, 0.6] and the service life deviation vector matrix of the corresponding component is [1400, 700], a weighted calculation is performed: [0.4*1400, 0.6*700] = [560, 420], obtaining the component service life deviation matrix. In this way, through the weighted calculation of the first weight distribution matrix and the service life deviation vector matrix of the corresponding component, the weight ratio of different components in the service life deviation comparison can be dynamically adjusted, so that the service life deviation is tilted towards components with high failure probability.

[0078] In summary, this application dynamically adjusts the weight ratio of different components in the service duration deviation comparison by weighting the first weight distribution matrix and the service duration deviation vector matrix of components with the same position number. Specifically, based on the historical failure probability of the components (i.e., the first weight distribution matrix), components with high failure probability are given higher weight coefficients, which significantly amplifies the deviation impact of these components in the comparison of primary and backup equipment. For example, a certain type of capacitor has a high failure probability in a specific service scenario, and its corresponding weight coefficient is 0.8, while the weight of low-risk components is only 0.2. When calculating the service duration deviation of components in primary and backup equipment, the actual deviation value is multiplied by different weights. This weighted calculation method breaks the limitation of the traditional method that only performs linear comparison based on time, and instead fully considers the nonlinear aging law and failure characteristics of equipment components in actual operation. By quantifying the difference in failure risk of different components, the comparison of service duration information of primary and backup equipment can more accurately reflect the difference in the true health status of components, effectively avoiding the evaluation deviation caused by simple linear comparison, and providing more reliable data support for primary and backup switching decisions.

[0079] In a specific embodiment, calculating the service duration deviation vector matrix of co-position element service durations in the main equipment element service duration list and the backup equipment element service duration list includes:

[0080] The service duration of the first equipment component number of the main and backup equipment is processed by the first equipment number service attenuation parameter evaluator to obtain the first equipment number service attenuation parameter of the main and backup equipment. The first equipment number service attenuation parameter of the main and backup equipment includes the first equipment number service attenuation parameter of the main equipment and the first equipment number service attenuation parameter of the backup equipment.

[0081] Calculate the ratio of the first service attenuation parameter of the main equipment to the sum of the service attenuation parameters of the main equipment, and add it to the service duration attenuation weight matrix of the main equipment;

[0082] Calculate the ratio of the first service attenuation parameter of the backup equipment to the sum of the service attenuation parameters of the backup equipment, and add it to the service time attenuation weight matrix of the backup equipment;

[0083] Based on the main equipment service life attenuation weight matrix, the main equipment component service life list is preprocessed and weighted to obtain the main equipment component corrected service life list.

[0084] Based on the backup equipment service life attenuation weight matrix, the backup equipment component service life list is preprocessed and weighted to obtain the backup equipment component corrected service life list.

[0085] Calculate the service duration deviation vector matrix of the corresponding element in the main equipment element modified service duration list and the backup equipment element modified service duration list.

[0086] Specifically, the elements in the main equipment service time attenuation weight matrix are weighted and calculated together with the corresponding elements in the main equipment component service time list to obtain the corrected service time of the main equipment component. The main equipment service time attenuation weight matrix is ​​then traversed, and the corrected service time list of main equipment components is obtained through weighted calculation. For example, if the main equipment component service time is 5000 hours and the main equipment service time attenuation weight is 0.3, after preprocessing and weighting, the corrected service time of the main equipment component is obtained as 5000 * 0.3 = 1500 hours.

[0087] Specifically, the elements in the backup equipment service duration attenuation weight matrix are weighted and calculated together with the corresponding elements in the backup equipment component service duration list to obtain the corrected service duration of the backup equipment component. The backup equipment service duration attenuation weight matrix is ​​then traversed, and the corrected service duration list of backup equipment components is obtained through weighted calculation. For example, if the backup equipment component service duration is 1000 hours and the backup equipment service duration attenuation weight is 0.1, after preprocessing and weighting, the corrected service duration of the backup equipment component is obtained as 1000 * 0.1 = 100 hours.

[0088] Finally, the service duration deviation vector matrix of the corresponding components in the corrected service duration list of the main equipment components and the corrected service duration list of the backup equipment components is calculated. Specifically, the service duration deviation vector of the corresponding components is generated by subtracting the corrected service duration lists of the main and backup components. For example, the corrected service duration list of the main equipment components is [1500h, 900h], and the corrected service duration list of the backup equipment components is [100h, 200h]. The difference is used to obtain the service duration deviation vector matrix of the corresponding components [1400h, 700h]. If no correction is made and the deviation is directly calculated based on the original duration data, it will be difficult to truly reflect the actual aging difference between the main and backup equipment components. The corrected service duration deviation vector matrix of the corresponding components better reflects the actual deviation degree of the main and backup equipment components.

[0089] In one specific embodiment, the construction steps of the first bit-number service attenuation parameter evaluator include:

[0090] Extract the first device component tag number from the device component list, and extract the service duration of the first device component tag number of the main and backup devices from the service duration information of the main and backup devices.

[0091] Constrained by the equipment model and the service scenario number, a data set of component service duration records and a data set of component failure rate records of the first equipment component tag number are collected. The component failure rate record data is characterized by the proportion of component failure states in several component status record data of the preset component service duration record data.

[0092] Using the component service duration record dataset and the component failure rate record dataset, a first-bit service attenuation parameter estimator belonging to the equipment model and the service scenario number is trained through a neural network.

[0093] In this embodiment, by comparing the attenuation and aging levels of the main device and the backup device, it is possible to assess in advance whether the device can operate normally after switching to the backup device. However, the actual aging rate of device components is not solely determined by usage time, but is also affected by the service environment (such as ambient temperature, humidity, load, etc.) and the component's own characteristics (such as material corrosion resistance, design life). For example, the attenuation rate of the same type of resistor component in a high-temperature environment may be twice that in a normal-temperature environment. If the aging state is directly measured based on the original duration, it will lead to assessment bias. Therefore, it is necessary to correct the original duration based on the device model and service environment using an attenuation parameter evaluator to eliminate the interference of differences in aging rates between different components. Specifically:

[0094] First, the service duration of the first equipment component reference number in both the primary and backup equipment is processed by the first-number service attenuation parameter evaluator to obtain the first-number service attenuation parameter for both the primary and backup equipment. This first-number service attenuation parameter includes the first-number service attenuation parameter for the primary equipment and the first-number service attenuation parameter for the backup equipment. For example, the service duration of the first equipment component reference number in both the primary and backup equipment (e.g., 5000h, 1000h) is input into the first-number service attenuation parameter evaluator, which outputs the first-number service attenuation parameter for the primary equipment and the first-number service attenuation parameter for the backup equipment (e.g., 0.3, 0.1).

[0095] Next, the ratio of the first service attenuation parameter of the main equipment to the sum of the service attenuation parameters of the main equipment is calculated and added to the service duration attenuation weight matrix of the main equipment. The first service attenuation weight of the main equipment is calculated as: first attenuation parameter / sum of attenuation parameters of all components of the main equipment. For example, if the first service attenuation parameter of the main equipment is 0.3 and the sum of the service attenuation parameters of the main equipment is 1.5, then the first service attenuation weight of the main equipment is 0.3 / 1.5 = 0.2. This first service attenuation weight is then added to the service duration attenuation weight matrix of the main equipment.

[0096] Next, calculate the ratio of the first service attenuation parameter of the backup device to the sum of the service attenuation parameters of the backup device, and add it to the service duration attenuation weight matrix of the backup device. The first service attenuation weight of the backup device is calculated as: First service attenuation parameter / Sum of attenuation parameters of all components of the backup device. For example, if the first service attenuation parameter of the backup device is 0.1 and the sum of the service attenuation parameters of the backup device is 1, then the first service attenuation weight of the backup device is 0.1 / 1 = 0.1. This first service attenuation weight of the backup device is then added to the service duration attenuation weight matrix of the backup device.

[0097] In this embodiment, the first device component tag number is extracted from the device component list, and the service duration of the first device component tag number of the primary and backup devices is extracted from the service duration information of the primary and backup devices. For example, the first device component tag number (e.g., T-021) is extracted from the device component list, and the service duration of the first device component tag number of the primary and backup devices (e.g., 3000h for the primary device and 500h for the backup device) is extracted from the service duration information of the primary and backup devices.

[0098] Secondly, constrained by the equipment model and the service scenario number, a dataset of component service duration records and a dataset of component failure rate records for the first equipment component tag number are collected. The component failure rate record data is characterized by the proportion of component failure states in several component status records within the preset component service duration record data. For example, constrained by the equipment model (e.g., PLC-500) and the service scenario number (e.g., high-temperature environment S-003), a dataset of component service duration records (e.g., 100h, 500h, 1000h, ...) and a dataset of component failure rate records (e.g., 5%, 7%, 10%, ...) for the first equipment component tag number (e.g., T-021) are collected. The component failure rate record data is calculated by the proportion of component failure states in several component status records. For example, if the service duration is 1000h and 10 out of 100 status records are faulty, the failure rate is 10%, and this is used as the component failure rate record data for the 1000h service duration.

[0099] Finally, using the component service duration record dataset and the component failure rate record dataset, a first-bit service attenuation parameter estimator belonging to the equipment model and the service scenario number is trained through a neural network. For example, the component service duration record dataset and the component failure rate record dataset are used as training data and divided into training, validation, and test sets in a 7:1.5:1.5 ratio. The neural network structure can be a single-hidden-layer neural network, mainly composed of an input layer, hidden layers, and an output layer. The input layer receives service duration data, the hidden layer composed of ReLU neurons learns the mapping relationship between service duration and failure rate, and the output layer predicts the output failure rate through linear activation. Furthermore, the training process can use a mean squared error loss function and an Adam optimizer. After 100 iterations, the mean absolute error (MAE) of the validation set reaches 0.032, and the coefficient of determination (R²) is... 2 The result is over 0.91, proving that the model successfully captured the nonlinear decay law.

[0100] It is clear that, compared to existing technologies, this application, based on equipment model and service scenario number, performs induced failure weight distribution on the equipment component list to obtain a first weight distribution matrix. Then, it calculates the weighted deviation of the service duration information of the primary and backup equipment components to generate a component service duration deviation matrix. Thus, by using the first weight distribution matrix for weighted calculation, it breaks through the limitations of traditional methods that rely solely on linear comparisons based on time. Instead, it fully considers the nonlinear aging patterns and failure characteristics of equipment components in actual operation. By quantifying the differences in failure risk among different components, the comparison of primary and backup equipment service duration information can more accurately reflect the differences in the true health status of the components, effectively avoiding comparison bias caused by simple linear comparisons.

[0101] 104. Determine the second weight distribution matrix based on the main equipment component model and the backup equipment component model. In this embodiment, a first component model is first extracted from the component model. For example, the first component model is extracted from the component model, such as capacitor C101, resistor R-200, etc.

[0102] Secondly, from the list of maintenance faults, the first maintenance fault type is extracted and input into a damage rating table bound to the first component model to obtain the damage rating of the first maintenance fault type. The damage rating table is a user-predefined data table storing fault types and damage levels; the higher the damage level, the greater the damage. For example, the first maintenance fault type (e.g., breakdown) corresponding to the first component model (e.g., capacitor C101) is extracted from the maintenance fault list, establishing a unique correspondence between component models and fault types. The first component model (e.g., capacitor C101) and the first maintenance fault type (e.g., breakdown) are input into the predefined damage rating table, and the damage rating of the first maintenance fault type (e.g., level 8) is output. Furthermore, the damage rating table is a user-predefined data table of stored fault types and damage levels. It can be a mapping table of fault types and damage levels that can be pre-set and stored by those skilled in the art based on industry standards or historical experience. The levels are usually 1-10 (the higher the damage level, the higher the degree of damage, such as level 10 indicating fatal damage). For example, the damage rating table of capacitor C101 pre-set based on historical experience is as follows: leakage current - damage level 5, breakdown - damage level 8.

[0103] Finally, the ratio of the damage rating of the first maintenance fault type to the sum of the damage ratings is calculated and set as the damage weight of the first maintenance fault type for the first component model. This ratio is then added to the second weight distribution matrix, where the damage weight of the first maintenance fault type = damage rating of the first maintenance fault type / sum of damage ratings. For example, if capacitor C101 has a leakage damage rating of 5, a breakdown damage rating of 8, and a sum of damage ratings of 13, then the damage weight of leakage = 5 / 13 = 0.38, and the damage weight of breakdown = 8 / 13 = 0.62. This ratio is set as the damage weight of the first maintenance fault type for the first component model and added to the second weight distribution matrix.

[0104] In this embodiment, determining the second weight distribution matrix based on the main equipment component model and the backup equipment component model includes:

[0105] Obtain the maintenance fault list corresponding to the main equipment and the backup equipment;

[0106] Based on the main equipment component model and the backup equipment component model, extract the first component model;

[0107] In the list of maintenance faults, the first maintenance fault type is extracted, and the damage rating table bound to the first component model is input to obtain the damage rating of the first maintenance fault type. The damage rating table is a user-predefined data table that stores fault types and damage levels. The higher the damage level, the higher the degree of damage.

[0108] Calculate the ratio of the damage rating of the first maintenance fault type to the sum of the damage ratings, set it as the damage weight of the first maintenance fault type of the first component model, and add it to the second weight distribution matrix.

[0109] In one specific embodiment, different types of maintenance faults cause varying degrees of damage to components. For example, in electronic components, a capacitor leakage fault may only result in a slight performance degradation, allowing the equipment to maintain basic operation. However, a breakdown fault will cause the capacitor to fail completely, directly leading to equipment shutdown and even damage to other components. Therefore, when comparing component maintenance faults, it is necessary to dynamically calculate the damage weight distribution based on the degree of damage caused by different maintenance faults, and then perform a weighted calculation to generate a component maintenance fault deviation matrix.

[0110] 105. Based on the second weight distribution matrix, the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component are weighted to obtain the component maintenance fault deviation matrix.

[0111] In this embodiment, the step of weighting the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component based on the second weight distribution matrix to obtain the component maintenance fault deviation matrix includes:

[0112] Based on the main equipment component maintenance fault information, extract the main equipment component model, main equipment component maintenance fault type, and main equipment component maintenance fault frequency.

[0113] Based on the fault information of the backup equipment components, extract the model of the faulty backup equipment component, the fault type of the backup equipment component, and the fault frequency of the backup equipment component.

[0114] Based on the main equipment faulty component model, main equipment fault type, main equipment fault frequency, standby equipment faulty component model, standby equipment fault type, and standby equipment fault frequency, a comparison of the same model and fault type is performed, and the deviation between the standby equipment fault frequency and the main equipment fault frequency is calculated to obtain an initial component fault deviation matrix; wherein, when the main equipment data or standby equipment data for the compared fault frequency is empty, zero padding is performed;

[0115] Based on the second weight distribution matrix, the initial component maintenance fault deviation matrix is ​​weighted by the same model and fault type to obtain the component maintenance fault deviation matrix.

[0116] In this embodiment, the main equipment component model, fault type, and fault frequency are first extracted from the main and backup equipment component maintenance fault information. For example, the main equipment component model (e.g., capacitor C101), fault type (e.g., leakage current), and fault frequency (e.g., twice / month) are extracted from the main and backup equipment component maintenance fault information.

[0117] Secondly, from the main and backup equipment component maintenance and fault information, the model of the backup equipment component that is experiencing maintenance and fault, the type of maintenance and fault, and the frequency of maintenance and fault are extracted. For example, from the backup equipment component maintenance and fault information, the model of the backup equipment component that is experiencing maintenance and fault (e.g., capacitor C101), the type of maintenance and fault (e.g., leakage current), and the frequency of maintenance and fault (e.g., once per month) are extracted.

[0118] Next, based on the model and fault type of the main equipment component under maintenance, and the model and fault type of the backup equipment component under maintenance, a comparison of the same model and fault type is performed. The deviation between the fault frequency of the backup equipment component and the fault frequency of the main equipment component is calculated to obtain an initial component maintenance fault deviation matrix. When the main equipment data or backup equipment data for comparing fault frequencies is empty, zero-padding is performed. For example, if either the main or backup equipment data is empty (e.g., the backup equipment has no data for that fault type), zero-padding is performed before calculation (e.g., if the main equipment frequency is 2, the backup equipment frequency is padded with 0, resulting in a deviation of 2-0=2). For example, using the model and fault type of the component under maintenance (e.g., capacitor C101, leakage current) as constraints, a comparison of the same model and fault type is performed to calculate the deviation between the fault frequency of the backup equipment component and the fault frequency of the main equipment component (e.g., 2-1=1), obtaining an initial component maintenance fault deviation. This process is repeated to obtain the initial component maintenance fault deviation matrix.

[0119] Finally, based on the second weight distribution matrix, the initial component maintenance fault deviation matrix is ​​weighted according to the same model and fault type to obtain the component maintenance fault deviation matrix. Specifically, each element in the initial component maintenance fault deviation matrix is ​​multiplied by the corresponding weight in the second weight distribution matrix to obtain the component maintenance fault deviation, where component maintenance fault deviation = initial deviation * second weight. For example, for capacitor C101, the initial deviation for leakage is 2, and the damage weight for leakage is 0.38. The weighted deviation = 2 × 0.38 = 0.76. The weighted calculation makes the frequency deviation of high-damage faults have a greater impact on the final result. For example, the fault weight for breakdown is 0.62, and its frequency deviation of 2 will be amplified to 1.24, while the deviation impact of low-weight faults will be weakened. This step is repeated to obtain the component maintenance fault deviation matrix.

[0120] In summary, compared to existing technologies, this application distributes damage weights across a list of maintenance faults based on component model to obtain a second weight distribution matrix. This matrix is ​​then used to weight the maintenance fault information of the primary and backup equipment components, generating a component maintenance fault deviation matrix. Thus, by converting the severity of fault types into calculable weights through a damage rating table, and by weighting frequency deviations with damage weights, a comprehensive risk assessment of fault occurrence frequency and damage severity is achieved, accurately quantifying the differences in maintenance faults between primary and backup equipment components.

[0121] 106. Compare the component service life deviation matrix and the component maintenance fault deviation matrix to obtain the difference comparison results between the master equipment and the slave equipment.

[0122] In this embodiment, comparing the component service life deviation matrix and the component maintenance failure deviation matrix to obtain the difference comparison result between the master equipment and the slave equipment includes:

[0123] The component service life deviation matrix and the component maintenance fault deviation matrix are input into a preset master-slave difference comparison model for comparison, and the difference comparison results between the master equipment and the slave equipment output by the master-slave difference comparison model are obtained.

[0124] In this embodiment, the training of the primary / backup difference comparison model includes:

[0125] Constrained by the main equipment model, backup equipment model, main equipment component service scenario, and backup equipment component service scenario, a component service duration deviation record matrix and a component maintenance fault deviation record matrix with switching result identifiers are collected; wherein, the switching result identifier represents the historical switching control state. When the control result does not meet the working requirements, the switching result identifier is uncontrollable, and when the control result meets the working requirements, the switching result identifier is controllable.

[0126] Using the switching result identifier as supervision, and the component service duration deviation record matrix and the component maintenance fault deviation record matrix as input, a single-hidden-layer neural network is integrated. The neural network with the most votes among the single-hidden-layer neural networks is used as output. Training is performed on the single-hidden-layer neural networks to obtain the master-slave difference comparison model.

[0127] It should be noted that the component service life deviation matrix quantifies the differences in the aging degree of primary and backup equipment components due to different service life, while the component maintenance fault deviation matrix reflects the differences in the frequency and severity of faults of primary and backup equipment components. Based on this, the differences between primary and backup equipment can be compared and early warnings can be issued.

[0128] In this embodiment of the application, firstly, based on the equipment model and the service scenario number, a component service duration deviation record matrix and a component maintenance fault deviation record matrix with switching result identifiers are collected. The switching result identifier represents the historical switching control state. When the control result does not meet the working requirements, the switching result identifier is uncontrollable. When the control result meets the working requirements, the switching result identifier is controllable. For example, based on equipment operation and maintenance and asset management systems (such as CMMS / EAM), or manually recorded equipment operation ledgers and historical maintenance records, and constrained by equipment model and service scenario number, a component service duration deviation record matrix and a component maintenance failure deviation record matrix with switching result identifiers can be collected. These matrices record the deviations in service duration and maintenance failures of primary and backup equipment components under specific equipment models and service scenarios in the past. The switching result identifier has only two states: controllable and uncontrollable. It is the only identifier for whether the working requirements can be met after the primary and backup equipment is switched in history. When the working requirements cannot be met after the switch, for example, after the primary equipment fails and the backup equipment fails to operate normally due to insufficient performance, the switching result identifier is uncontrollable. Conversely, if the control result meets the working requirements after the switch, the switching result identifier is controllable. In this way, supervision signals are provided for subsequent model training.

[0129] Secondly, using the switching result identifier as supervision, and the component service duration deviation record matrix and the component maintenance fault deviation record matrix as input, an integrated single-hidden-layer neural network is used. The output is the majority vote of several single-hidden-layer neural networks, and training is performed to obtain the primary / backup difference comparison model. For example, an architecture integrating a single-hidden-layer neural network can be used for model construction. Single-hidden-layer neural networks have a relatively simple structure, high computational efficiency, and are suitable for processing this type of structured data. Multiple single-hidden-layer neural networks are trained in parallel, with each network independently learning the mapping relationship between the input data (component service duration deviation record matrix and component maintenance fault deviation record matrix) and the output result (switching result identifier). For example, during training, the component service duration deviation record matrix, the component maintenance fault deviation record matrix, and the switching result identifier are used as training data, and divided into a training set, a validation set, and a test set in a 7:1.5:1.5 ratio. Then, the collected component service duration deviation record matrix and component maintenance fault deviation record matrix are used as training data. Using the fault deviation record matrix as input and the switching result identifier as the supervision target, the parameters of each single hidden layer neural network are continuously adjusted through the backpropagation algorithm to make the network's prediction result as close as possible to the actual switching result identifier. The final output of the model can adopt the voting majority principle, that is, multiple single hidden layer neural networks predict the input data respectively, and the result predicted by the majority network (such as the mode) is used as the final output. This method can effectively reduce the risk of erroneous predictions caused by fluctuations in training data or model deviations of individual networks, and improve the overall stability and reliability of the model. The trained master-slave difference comparison model can make full use of historical data and the learning ability of neural networks to accurately evaluate and predict the state differences of master and standby devices.

[0130] Finally, the component service life deviation matrix and the component maintenance failure deviation matrix are processed using a pre-trained primary / backup difference comparison model to generate a primary / backup difference comparison result. When the primary / backup difference comparison result is uncontrollable, a switching warning indicator is issued for the backup equipment. For example, the component service life deviation matrix (e.g., [560, 420]) and the component maintenance failure deviation matrix (e.g., [0.76, 1.24]) are input into the pre-trained primary / backup difference comparison model to generate a primary / backup difference comparison result (e.g., uncontrollable), indicating that the control result still does not meet the working requirements after switching the backup equipment. At this time, a switching warning indicator is issued for the backup equipment.

[0131] In summary, compared to existing technologies, this application constructs a primary / backup difference comparison model to process the component service life deviation matrix and the component maintenance fault deviation matrix, generating a primary / backup difference comparison result. When the primary / backup difference comparison result is uncontrollable, a switching warning indicator is issued for the backup equipment. In this way, the primary / backup difference comparison model can be fully utilized to accurately compare the differences between primary and backup equipment and provide early warning of control disturbances after equipment switching.

[0132] In summary, through the above technical solutions, this application constructs a dual-weight distribution matrix based on multi-dimensional data such as equipment model, service scenario, component model, and maintenance fault type: The first weight distribution matrix focuses on the nonlinear relationship between service duration and fault risk. Based on historical data of equipment model and service scenario, components with significant nonlinear risk are given high weight. When there is a slight difference in service duration between the main and backup equipment and components with significant nonlinear risk, the weighting will amplify the impact of this deviation, thereby capturing hidden risks where the absolute duration difference is not large but the risk has been significantly differentiated. Components whose risk changes linearly with time are given low weight to avoid them masking key risks. The second weighted distribution matrix focuses on the nonlinear correlation between fault type and equipment damage. Based on component model and a predefined damage rating table, it assigns high weights to high-damage fault types. When the primary and backup equipment exhibit different frequencies of high-damage fault types, the weighted distribution amplifies the impact of this deviation, capturing hidden cumulative damage that occurs infrequently but has severe consequences. Conversely, it assigns low weights to low-damage faults to prevent them from masking core risks due to their high frequency. Then, the weighted deviation of the dual weighted distribution matrix is ​​fused with the component service life deviation matrix and the component maintenance fault deviation matrix through a primary / backup difference comparison model trained by a neural network, ultimately generating the primary / backup difference comparison result. Thus, when the primary / backup difference comparison result is uncontrollable, a switchover warning is issued for the backup equipment, achieving advance perception of control disturbances after equipment switchover.

[0133] like Figure 2 As shown, based on the above method embodiments, corresponding apparatus embodiments are provided;

[0134] An embodiment of the present invention provides a primary / backup difference comparison device for an industrial system, comprising: a data acquisition module 201, a first weight module 202, a first matrix calculation module 203, a second weight module 204, a second matrix calculation module 205, and a result generation module 206;

[0135] The data acquisition module is used to acquire the operating data of the main equipment and backup equipment of the industrial system under test; wherein, the operating data includes: main equipment model, main equipment component service scenario, main equipment component service duration information, main equipment component model, main equipment component maintenance and fault information, backup equipment model, backup equipment component service scenario, backup equipment component service duration information, backup equipment component model, and backup equipment component maintenance and fault information.

[0136] The first weighting module is used to determine a first weighting distribution matrix based on the main equipment model, the service scenario of the main equipment component, the backup equipment model, and the service scenario of the backup equipment component;

[0137] The first matrix calculation module is used to perform weighted deviation calculation on the service duration information of the main equipment component and the service duration information of the backup equipment component based on the first weight distribution matrix, so as to obtain the component service duration deviation matrix;

[0138] The second weighting module is used to determine a second weighting distribution matrix based on the main equipment component model and the backup equipment component model;

[0139] The second matrix calculation module is used to weight the maintenance fault information of the main equipment component and the maintenance fault information of the backup equipment component based on the second weight distribution matrix to obtain the component maintenance fault deviation matrix.

[0140] The result generation module is used to compare the component service life deviation matrix and the component maintenance fault deviation matrix to obtain the difference comparison results between the master equipment and the slave equipment.

[0141] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the master-slave difference comparison method for industrial systems provided by any of the above-described method embodiments of the present invention.

[0142] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0143] Based on the above embodiments of the primary / backup difference comparison method for industrial systems, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the primary / backup difference comparison method for industrial systems according to any embodiment of the present invention.

[0144] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0145] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0146] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0147] Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the master-slave difference comparison method for industrial systems described in any of the above-described method embodiments of the present invention.

[0148] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0149] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A master-backup difference comparison method for an industrial system, characterized by, The method comprises: obtaining running data of a main device and a backup device of an industrial system to be tested, wherein the running data comprises a main device model, a main device element service scene, main device element service time length information, a main device element model, main device element maintenance fault information, a backup device model, a backup device element service scene, backup device element service time length information, a backup device element model, and backup device element maintenance fault information; determining a first weight distribution matrix according to the main device model, the main device element service scene, the backup device model, and the backup device element service scene; performing weighted deviation calculation on the main device element service time length information and the backup device element service time length information based on the first weight distribution matrix to obtain an element service time length deviation matrix; determining a second weight distribution matrix according to the main device element model and the backup device element model; performing weighting on the main device element maintenance fault information and the backup device element maintenance fault information based on the second weight distribution matrix to obtain an element maintenance fault deviation matrix; comparing the element service time length deviation matrix and the element maintenance fault deviation matrix to obtain a difference comparison result between the main device and the backup device.

2. The master-backup discrepancy comparison method for an industrial system according to claim 1, wherein, The method comprises: obtaining a device element list corresponding to the main device and the backup device; extracting a first device element bit number in the device element list, and collecting device working state record data of the first device element bit number in a fault state with the main device model, the backup device model, the main device element service scene, and the backup device element service scene as constraints; wherein the device working state record data comprises a healthy state and a fault state; statistically determining a proportion of the number of the device working state record data in the fault state, and setting the proportion as a first induced fault initial weight; when the first induced fault initial weight is greater than or equal to an induced weight threshold, adding the first induced fault initial weight to an induced fault initial weight set; otherwise, setting a first device element bit number induced fault weight to 0 and adding the first device element bit number induced fault weight to the first weight distribution matrix.

3. The master-backup discrepancy comparison method for an industrial system according to claim 2, wherein, The method comprises: extracting a main device element service time length list and a backup device element service time length list from the main device element service time length information and the backup device element service time length information; calculating a same-bit element service time length deviation vector matrix of the main device element service time length list and the backup device element service time length list; performing weighting on the same-bit element service time length deviation vector matrix based on the first weight distribution matrix to obtain the element service time length deviation matrix.

4. The master-backup discrepancy comparison method for an industrial system according to claim 3, wherein, The method comprises: obtaining a maintenance fault list corresponding to the main device and the backup device; According to the main device element model and the standby device element model, a first element model is extracted; In the maintenance fault list, a first maintenance fault type is extracted, an injury rating table bound with the first element model is input, and a first maintenance fault type injury rating is obtained, wherein the injury rating table is a user-predefined data table storing fault types and injury levels, and the greater the injury level, the higher the injury degree; The ratio of the sum of the first maintenance fault type injury rating and the injury rating is calculated, which is set as the first maintenance fault type injury weight of the first element model, and is added to the second weight distribution matrix.

5. The master-backup discrepancy comparison method for an industrial system according to claim 4, wherein, Based on the second weight distribution matrix, the main device element maintenance fault information and the standby device element maintenance fault information are weighted to obtain an element maintenance fault deviation matrix, including: According to the main device element maintenance fault information, the main device maintenance fault element model, the main device element maintenance fault type and the main device element maintenance fault frequency are extracted; According to the standby device element maintenance fault information, the standby device maintenance fault element model, the standby device element maintenance fault type and the standby device element maintenance fault frequency are extracted; Based on the main device maintenance fault element model, the main device element maintenance fault type, the main device element maintenance fault frequency, the standby device maintenance fault element model, the standby device element maintenance fault type and the standby device element maintenance fault frequency, the same type and same fault type comparison is performed, the deviation of the standby device element maintenance fault frequency and the main device element maintenance fault frequency is calculated, and an initial element maintenance fault deviation matrix is obtained; wherein when the main device data or the standby device data of the compared maintenance fault frequency is empty, zero filling is performed; According to the second weight distribution matrix, the same type and same fault type of the initial element maintenance fault deviation matrix is weighted to obtain the element maintenance fault deviation matrix.

6. The master-backup discrepancy comparison method for an industrial system according to claim 5, wherein, The element service time length deviation matrix and the element maintenance fault deviation matrix are compared to obtain a difference comparison result between the main device and the standby device, including: The element service time length deviation matrix and the element maintenance fault deviation matrix are input into a preset main-standby difference comparison model for comparison to obtain a difference comparison result between the main device and the standby device output by the main-standby difference comparison model.

7. The master-backup discrepancy comparison method for an industrial system according to Claim 6, wherein, The training of the main-standby difference comparison model includes: With the main device model, the standby device model, the main device element service scene and the standby device element service scene as constraints, the element service time length deviation record matrix and the element maintenance fault deviation record matrix with a switching result identifier are collected; wherein the switching result identifier represents the historical switching control state, when the control result does not meet the work demand, the switching result identifier is uncontrollable, and when the control result meets the work demand, the switching result identifier is controllable; With the switching result identifier as supervision, the element service time length deviation record matrix and the element maintenance fault deviation record matrix as input, a single hidden layer neural network is integrated, the neural network corresponding to the most votes in a plurality of single hidden layer neural networks is output, a plurality of single hidden layer neural networks are trained, and the main-standby difference comparison model is obtained.

8. A primary / backup difference comparison device for an industrial system, characterized in that, Comprising: a data acquisition module, a first weight module, a first matrix calculation module, a second weight module, a second matrix calculation module, and a result generation module; the data acquisition module is configured to acquire operation data of a main device and a backup device of an industrial system to be tested, wherein the operation data comprises a main device model, a main device element service scene, main device element service time length information, a main device element model, main device element maintenance fault information, a backup device model, a backup device element service scene, backup device element service time length information, a backup device element model, and backup device element maintenance fault information; the first weight module is configured to determine a first weight distribution matrix according to the main device model, the main device element service scene, the backup device model, and the backup device element service scene; the first matrix calculation module is configured to perform weighted deviation calculation on the main device element service time length information and the backup device element service time length information based on the first weight distribution matrix to obtain an element service time length deviation matrix; the second weight module is configured to determine a second weight distribution matrix according to the main device element model and the backup device element model; the second matrix calculation module is configured to perform weighting on the main device element maintenance fault information and the backup device element maintenance fault information based on the second weight distribution matrix to obtain an element maintenance fault deviation matrix; the result generation module is configured to compare the element service time length deviation matrix and the element maintenance fault deviation matrix to obtain a difference comparison result between the main device and the backup device.

9. A terminal device, comprising: Comprising:

10. A computer-readable storage medium, characterized in that, a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, when the computer program is executed by the processor, the processor implements the main-backup difference comparison method for an industrial system according to any one of claims 1-7. Comprising: a stored computer program, wherein when the computer program is running, the device where the computer readable storage medium is located is controlled to perform the main-backup difference comparison method for an industrial system according to any one of claims 1-7.