A full-digital networked control system fault detection method and application
By combining standardized processing with the Transformer-LSTM model to generate dynamic fault thresholds, the problems of low fault diagnosis coverage and long-term dependence in traditional methods are solved, achieving high reliability and real-time response fault detection, and adapting to the multi-level safety requirements of nuclear safety-grade tasks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA STATE SHIPBUILDING CORP LTD RESEARCH INSTITUTE 719
- Filing Date
- 2025-06-30
- Publication Date
- 2026-07-31
AI Technical Summary
Traditional fully digital networked control systems rely on threshold rules for fault diagnosis, which cannot cope with complex multimodal sensor data and has low fault detection coverage. Furthermore, existing AI algorithms have long-term dependency issues in time-series data processing and lack the ability to dynamically adapt to multi-level safety requirements, making it difficult to meet the real-time and reliability requirements of nuclear safety-grade tasks.
A data feature fusion architecture is constructed by combining standardized processing with Transformer model and LSTM network to generate dynamic fault thresholds. By mixing exponential moving average algorithm with learnable parameters, the system state can be adjusted in real time and the fault risk level can be determined.
It improves fault diagnosis coverage, reduces false alarm rate, achieves millisecond-level real-time response and high reliability, and adapts to multi-level safety requirements.
Smart Images
Figure CN120871807B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of control system fault detection, and in particular to a method for detecting faults in a fully digital networked control system, a device for detecting faults in a fully digital networked control system, an electronic device, and a computer-readable storage medium. Background Technology
[0002] Traditional fully digital networked control systems, such as Digital Instrument Control Systems (DCS), rely on threshold rules for fault diagnosis, which cannot handle complex multimodal sensor data and results in low fault detection coverage. Furthermore, existing AI algorithms (such as LSTM and CNN) suffer from long-term dependency issues in time-series data processing and lack the ability to dynamically adapt to multi-level safety requirements. In particular, nuclear safety-grade tasks have extremely high requirements for real-time performance and reliability, which current technologies struggle to meet. Summary of the Invention
[0003] To overcome the shortcomings of the prior art, this application provides a fault detection method and application for a fully digital networked control system, which can solve the problems of low fault detection coverage, long-term algorithm dependence, and lack of dynamic adaptability to safety requirements in existing DCS systems.
[0004] On the one hand, this application proposes a fault detection method for a fully digital networked control system, including: collecting raw sensor data of the control system and performing standardized processing to output standardized data; processing the standardized data through a transformer self-attention mechanism to generate global features, using a neural network to enhance the global features, and outputting fused features; generating a dynamic fault threshold based on the fused features using learnable parameters and an exponential moving average algorithm; and calculating the fault probability based on the dynamic fault threshold and the fused features.
[0005] In one embodiment of this application, the calculation formula for standardizing the raw sensor data is as follows: ;in, This represents the raw sensor data, where N is the number of sensors and T is the time step. To standardize data, For the first i The average value of each sensor, The standard deviation is denoted as .
[0006] In one embodiment of this application, the transformer self-attention mechanism satisfies: Where Q, K, and V are input matrices, representing the query, key, and value, respectively. For activation function, This is the dimension scaling factor.
[0007] In one embodiment of this application, the neural network uses LSTM to enhance the global features, and the gating mechanism of the LSTM is expressed as follows: ;in, For the Gate of Oblivion This is the weight matrix. For bias terms, This is a global feature.
[0008] In one embodiment of this application, the formula for generating the dynamic fault threshold is: ;in, As a learnable parameter, EMA stands for Exponential Moving Average. As a feature of fusion, This is the dynamic fault threshold.
[0009] In one embodiment of this application, the formula for calculating the failure probability is as follows: ;in, This represents the probability of failure. This is the activation function.
[0010] In one embodiment of this application, after calculating the fault probability, the method further includes: determining the relationship between the fault probability and the dynamic fault threshold, determining the fault risk level based on the determination result, and matching the corresponding fault handling strategy.
[0011] On the other hand, this application also proposes a fault detection device for a fully digital networked control system, comprising: a data standardization processing module for collecting raw sensor data of the control system and performing standardization processing to output standardized data; a feature fusion module for processing the standardized data through a transformer self-attention mechanism to generate global features, using a neural network to enhance the global features, and outputting fused features; a dynamic fault threshold generation module for generating a dynamic fault threshold based on the fused features using a learnable parameter and an exponential moving average algorithm; and a fault probability calculation module for calculating the fault probability based on the dynamic fault threshold and the fused features.
[0012] In another aspect, embodiments of this application also propose an electronic device, including: a memory and one or more processors connected to the memory, the memory storing a computer program, and the processors executing the computer program to implement the fault detection method for a fully digital networked control system as described in any of the above embodiments.
[0013] In another aspect, embodiments of this application also propose a computer-readable storage medium storing computer-executable instructions for performing the fault detection method for a fully digital networked control system as described in any of the above embodiments.
[0014] As can be seen from the above, the embodiments of this application, compared with the prior art, can have at least one or more of the following beneficial effects: The fault detection method for a fully digital networked control system proposed in this application addresses the problems of long-term dependence and difficulty in local feature extraction in existing data processing algorithms by standardizing sensor data and constructing a data feature fusion architecture using the Transformer model and Long Short-Term Memory (LSTM) network, thereby improving fault diagnosis coverage. A method based on the exponential moving average algorithm and a hybrid learnable parameter-based dynamic fault threshold generation enables real-time adjustment of the system state, adapting to multi-level safety requirements and effectively reducing false alarm rates. Finally, the method determines the fault risk level based on the relationship between the calculated fault probability and the dynamic fault threshold, and matches corresponding fault handling strategies, thus achieving millisecond-level real-time response and high reliability. Attached Figure Description
[0015] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a fault detection method for a fully digital networked control system provided in this application embodiment; Figure 2 This application provides a schematic diagram illustrating the specific execution logic of a fault detection method for a fully digital networked control system. Figure 3 This is a schematic diagram of the Transformer-LSTM fusion model structure provided in an embodiment of this application; Figure 4 This is a schematic diagram of the dynamic threshold generation process provided in an embodiment of this application; Figure 5 This is a schematic diagram of a multi-level diagnostic decision-making process provided in an embodiment of this application; Figure 6 A schematic diagram of the structure of a fault detection device for a fully digital networked control system provided in this application embodiment; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application; Figure 8 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation
[0016] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described with reference to the accompanying drawings and embodiments.
[0017] To enable those skilled in the art to better understand the technical solutions of this application, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments, and all should fall within the protection scope of this application.
[0018] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are applicable in distinguishing similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or applicable to such processes, methods, products, or apparatus.
[0019] It should also be noted that the division of multiple embodiments in this application is only for the convenience of description and should not constitute a special limitation. Features in various embodiments can be combined and referenced in each other without contradiction.
[0020] like Figure 1 As shown, the first embodiment of this application proposes a fault detection method for a fully digital networked control system, which includes, for example: step S1, collecting the raw sensor data of the control system and performing standardization processing to output standardized data; step S2, processing the standardized data through a transformer self-attention mechanism to generate global features, using a neural network to enhance the global features, and outputting fused features; step S3, generating a dynamic fault threshold based on the fused features using a learnable parameter and an exponential moving average algorithm; and step S4, calculating the fault probability based on the dynamic fault threshold and the fused features.
[0021] Specifically, the fault detection method for the fully digital networked control system is applied, for example, to high-safety-requirement control tasks in DIMA (Distributed Avionics Architecture) systems of nuclear power plants, such as fault diagnosis of the main coolant pump of the nuclear reactor and fault diagnosis of the nuclear heating station. Combined with... Figure 2As shown, in step S1, for example, multi-source detection data is acquired through sensors and data fusion and standardization are performed. The acquired raw sensor data is represented as follows: Where N is the number of sensors and T is the time step. In one implementation, the formula is used... Data standardization processing is performed, among which To standardize data, For the first i The average value of each sensor, The standard deviation is denoted as .
[0022] In step S2, combined Figure 3 As shown, for example, time-series feature extraction based on the Transformer model, with standardized input data... Output global features .
[0023] In one implementation, the self-attention mechanism of the Transformer model is represented as: Where Q, K, and V are input matrices, representing the query, key, and value, respectively. For activation function, This is the dimension scaling factor.
[0024] In one implementation, the neural network employs a Long Short-Term Memory (LSTM) network for global features. Perform feature enhancement and output fused features. The gating mechanism of the Long Short-Term Memory (LSTM) network is expressed as follows: ; in, For the Gate of Oblivion This is the weight matrix. This is a bias term.
[0025] Next, in step S3, combined with Figure 4 As shown, the input fusion features A dynamic fault threshold is generated using learnable parameters and an exponential moving average algorithm. The calculation formula is as follows: ; in, As a learnable parameter, EMA stands for Exponential Moving Average. As a feature of fusion, This is the dynamic fault threshold.
[0026] In step S4, the multi-level diagnostic decision input fusion features and dynamic fault threshold The formula for calculating the failure probability is expressed as follows: ; in, This represents the probability of failure. This is the activation function.
[0027] In one implementation, combined Figure 5 As shown, after calculating the fault probability, the relationship between the fault probability and the dynamic fault threshold is determined, the fault risk level is determined based on the determination result, and the corresponding fault handling strategy is matched.
[0028] Specifically, if we are judging the probability of failure... >Dynamic fault threshold If the fault risk level is determined to be high, an emergency repair command (such as reactor shutdown protection) is triggered; if the fault probability is less than 0.5, the fault risk level is determined to be high. Dynamic fault threshold If the fault risk level is determined to be medium risk, the redundant link is switched; otherwise, the fault risk level is determined to be low risk, the log is recorded, and continuous monitoring is performed.
[0029] In one implementation, for example, the above-mentioned fault detection model is further validated and updated online, with real-time feedback fault probability data as input and updated model parameters as output. The calculation formula for online learning is expressed as follows: ; in, For learning rate, This is the loss function.
[0030] In summary, the fault detection method for a fully digital networked control system proposed in the first embodiment of this application solves the problems of long-term dependence and difficulty in local feature extraction in existing data processing algorithms by standardizing sensor data and constructing a data feature fusion architecture by combining the Transformer model and the Long Short-Term Memory (LSTM) network, thereby improving the fault diagnosis coverage. The method of generating dynamic fault thresholds based on the exponential moving average algorithm and learnable parameters enables real-time adjustment of the system state, which can adapt to multi-level safety requirements and effectively reduce the false alarm rate. The fault risk level is determined according to the relationship between the calculated fault probability and the dynamic fault threshold, and the corresponding fault handling strategy is matched, thereby achieving millisecond-level real-time response and high reliability.
[0031] In addition, such as Figure 6 As shown, the second embodiment of this application also proposes a fault detection device 20 for a fully digital networked control system, which includes, for example, a data standardization processing module 201, a feature fusion module 202, a dynamic fault threshold generation module 203, and a fault probability calculation module 204.
[0032] The data standardization processing module 201 is used to collect the raw sensor data of the control system and perform standardization processing to output standardized data; the feature fusion module 202 is used to process the standardized data through the transformer self-attention mechanism to generate global features, and to perform feature enhancement on the global features using a neural network to output fused features; the dynamic fault threshold generation module 203 is used to generate a dynamic fault threshold based on the fused features using a learnable parameter and an exponential moving average algorithm; and the fault probability calculation module 204 is used to calculate the fault probability based on the dynamic fault threshold and the fused features.
[0033] The fault detection method for a fully digital networked control system implemented by the fault detection device for a fully digital networked control system disclosed in the second embodiment of this application is as described in the first embodiment above, and therefore will not be described in detail here. Optionally, each module and the other operations or functions described above are for implementing the method described in the first embodiment, and the beneficial effects of the fault detection device for a fully digital networked control system provided in this embodiment are the same as the beneficial effects of the fault detection method for a fully digital networked control system provided in the first embodiment above. For the sake of brevity, they will not be repeated here.
[0034] like Figure 7 As shown, the third embodiment of this application also proposes an electronic device 30, which includes, for example, at least one processing unit 31 and at least one storage unit 32, wherein the storage unit 32 stores a computer program, and when the computer program is executed by the processing unit 31, the processing unit 31 performs the method described in the first embodiment, and the beneficial effects of the electronic device 30 provided in this embodiment are the same as the beneficial effects of the fault detection method of the fully digital networked control system provided in the first embodiment.
[0035] like Figure 8 As shown, the fourth embodiment of this application also provides a computer-readable storage medium 40, on which a computer program is stored. When the program is executed by a processor, it implements the steps of the above-described method. The beneficial effects of the computer-readable storage medium 40 provided in this embodiment are the same as those of the fault detection method for the fully digital networked control system provided in the first embodiment.
[0036] The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs, CD-ROMs, microdrives, as well as magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.
[0037] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.
[0038] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.
[0039] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between devices or units may be electrical or other forms.
[0040] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0041] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.
[0042] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.
[0043] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.
[0044] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of embodiments of this disclosure upon considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.
[0045] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0046] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this application should be included within the scope of protection of this application.
Claims
1. A fault detection method for a fully digital networked control system, characterized in that, include: The raw sensor data of the control system is collected and standardized, and standardized data is output. The standardized data is processed by a transformer self-attention mechanism to generate global features, and the global features are enhanced by a neural network to output fused features. Based on the fusion features, a dynamic fault threshold is generated using a learnable parameter and an exponential moving average algorithm; the formula for generating the dynamic fault threshold is: ； in, EMA is a learnable parameter, representing the exponential moving average. As a feature of fusion, Dynamic fault threshold; The fault probability is calculated based on the dynamic fault threshold and the fused features; the formula for the fault probability is: ； in, This represents the probability of failure. This is the activation function.
2. The fault detection method for a fully digital networked control system according to claim 1, characterized in that, The calculation formula for standardizing the raw sensor data is as follows: ； in, This represents the raw sensor data, where N is the number of sensors and T is the time step. To standardize data, For the first i The average value of each sensor, The standard deviation is denoted as .
3. The fault detection method for a fully digital networked control system according to claim 1, characterized in that, The transformer self-attention mechanism satisfies: ； Where Q, K, and V are input matrices representing the query, key, and value, respectively. For activation function, This is the dimension scaling factor.
4. The fault detection method for a fully digital networked control system according to claim 1, characterized in that, The neural network uses LSTM to enhance the global features, and the gating mechanism of the LSTM is expressed as follows: ； in, For the Gate of Oblivion This is the weight matrix. For bias terms, This is a global feature.
5. The fault detection method for a fully digital networked control system according to claim 1, characterized in that, After calculating the failure probability, the method further includes: Determine the relationship between the fault probability and the dynamic fault threshold, determine the fault risk level based on the determination result, and match the corresponding fault handling strategy.
6. A fault detection device for a fully digital networked control system, characterized in that, include: The data standardization processing module is used to collect the raw sensor data of the control system, perform standardization processing, and output standardized data; The feature fusion module is used to process the standardized data through the transformer self-attention mechanism to generate global features, use a neural network to enhance the global features, and output fused features. A dynamic fault threshold generation module is used to generate a dynamic fault threshold based on the fused features using a learnable parameter and an exponential moving average algorithm; the formula for generating the dynamic fault threshold is: ； in, EMA is a learnable parameter, representing the exponential moving average. As a feature of fusion, Dynamic fault threshold; The fault probability calculation module is used to calculate the fault probability based on the dynamic fault threshold and the fused features; the formula for the fault probability is: ； in, This represents the probability of failure. This is the activation function.
7. An electronic device, characterized in that, include: A memory and one or more processors connected to the memory, the memory storing a computer program, the processors executing the computer program to implement the fault detection method for a fully digital networked control system as described in any one of claims 1-5.
8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable commands for performing the fault detection method for a fully digital networked control system as described in any one of claims 1-5.