Fault diagnosis method and system for nuclear power small reactor based on artificial intelligence

Abstract

The present application belongs to the technical field of nuclear power small reactor, and particularly relates to a fault diagnosis method and system for a nuclear power small reactor based on artificial intelligence. The method comprises the following steps: obtaining target operation data of a reactor cooling pump of a nuclear power small reactor collected by a production control area of the nuclear power small reactor based on a data diode in an information management area; processing the target operation data to obtain first characteristic information of the reactor cooling pump; inputting the first characteristic information into a preset fault diagnosis model to obtain first initial fault diagnosis information of the reactor cooling pump; and sending the first initial fault diagnosis information to the cloud to generate target fault diagnosis information of the reactor cooling pump in the production control area. The method has the advantages that the data diode is used to safely obtain data and perform diagnosis in the information management area, the network security problem of the nuclear power small reactor is solved, the data in the production control area can be safely obtained, the network security risk is avoided, and efficient fault diagnosis is realized.

Description

Technical Field

[0001] This invention belongs to the field of nuclear power small modular reactor (SMR) technology, specifically relating to a fault diagnosis method and system for nuclear power SMR based on artificial intelligence. Background Technology

[0002] With the increasing application of intelligent technologies in the construction and operation of Small Modular Reactors (SMRs), while these technologies have significantly improved operational safety and efficiency, they have also brought serious cybersecurity challenges. Particularly in the area of ​​fault diagnosis involving critical equipment such as reactor cooling pumps, there is a risk of malicious attackers exploiting these technologies, potentially leading to the alteration of diagnostic results or even reverse penetration that could affect the reactor's operational status.

[0003] Existing intelligent diagnostic systems in nuclear power plants often face a dilemma: on the one hand, they need to acquire high-precision real-time data from the production control area to ensure diagnostic accuracy, and on the other hand, they need to ensure that the production control area is not subject to any reverse interference.

[0004] While the introduction of data diode technology can achieve unidirectional isolation at the physical level, significant shortcomings remain in areas such as data integrity verification, encryption, and multi-region collaborative diagnostics during data transmission. Particularly in the monitoring of critical equipment like reactor cooling pumps, there is a lack of effective safety mechanisms for real-time processing and feature extraction of multi-dimensional parameters such as vibration, temperature, and pressure.

[0005] More significantly, existing systems struggle to achieve collaborative verification of diagnostic results between the information management area and the production control area. When the two areas conduct independent diagnoses based on different data sources, the lack of effective discrepancy analysis and result fusion mechanisms can lead to inconsistent or delayed diagnostic conclusions, affecting the timeliness and accuracy of operational decisions. Summary of the Invention

[0006] The purpose of this invention is to provide a fault diagnosis method and system for small modular reactors based on artificial intelligence, which can securely acquire data from the production control area, avoid network security risks, and achieve efficient fault diagnosis.

[0007] The technical solution of the present invention is as follows: a fault diagnosis method for small nuclear power reactors based on artificial intelligence, applied to the information management area of ​​the small nuclear power reactor, the method comprising:

[0008] Based on the data diodes in the information management area, the target operating data of the reactor cooling pump of the nuclear power small modular reactor (SMR) is acquired from the production control area of ​​the SMR.

[0009] The target operating data is processed to obtain the first characteristic information of the reactor cooling pump;

[0010] The first feature information is input into a preset fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump.

[0011] The first initial fault diagnosis information is sent to the cloud to generate target fault diagnosis information for the reactor cooling pump in the production control area.

[0012] The data diode based on the information management area acquires the target operating data of the reactor cooling pump of the nuclear power small modular reactor collected from the production control area of ​​the nuclear power small modular reactor, including:

[0013] Receive the first initial data collected from the reactor cooling pump in the production control area;

[0014] The first initial acquisition data is processed and the data diode is used for data transmission to obtain the target operating data of the reactor cooling pump from the first initial acquisition data.

[0015] The process of processing the first initial data acquisition and using the data diode for data transmission to obtain the target operating data of the reactor cooling pump includes:

[0016] The acquired data is processed using a preset signal conditioning circuit to obtain the first acquired data of the reaction cooling pump;

[0017] The first collected data is filtered to obtain the second collected data of the reaction cooling pump.

[0018] The data diode is used for data transmission to obtain the target operating data of the reactor cooling pump from the second acquired data.

[0019] The step of using the data diode for data transmission to obtain the target operating data of the reactor cooling pump from the first initial acquisition data includes:

[0020] The data diode is used to verify the second collected data to obtain the verification result of the second collected data;

[0021] Based on the verification result, the second collected data is encrypted to obtain the third collected data of the reactor cooling pump;

[0022] The data diode is used for data transmission to obtain the target operating data of the reactor cooling pump from the third acquired data.

[0023] The first feature information includes the first vibration information, the first temperature information, and the first pressure information of the reactor cooling pump;

[0024] The process of processing the target operating data to obtain the first characteristic information of the reactor cooling pump includes:

[0025] The target data is decrypted to obtain the decrypted target running data;

[0026] The decrypted target running data is filtered to obtain filtered target running data.

[0027] Feature extraction is performed on the filtered target operating data to obtain the first feature information, the first temperature information, and the first pressure information.

[0028] An artificial intelligence-based fault diagnosis method for small modular reactors (SMRs) is applied to the production control area of ​​the SMR, the method comprising:

[0029] Based on the cloud, the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor is obtained;

[0030] Based on the first initial fault diagnosis information, the second initial acquisition data of the reactor cooling pump of the nuclear power small reactor collected in the production control area is obtained;

[0031] The second initial data acquisition is processed to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information, and the second pressure information of the reactor cooling pump.

[0032] The second vibration information, the second temperature information, and the second pressure information are input into a preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump.

[0033] Based on the first initial fault diagnosis information and the second initial fault diagnosis information, target fault diagnosis information for the reactor cooling pump is generated.

[0034] The step of generating target fault diagnosis information for the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information includes:

[0035] Generate a first difference information between the first initial fault diagnosis information and the second initial fault diagnosis information;

[0036] If the first difference information is less than or equal to a preset first threshold, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information;

[0037] If the first difference information is greater than or equal to the first threshold, the second initial fault diagnosis information is used as the target fault diagnosis information.

[0038] The step of generating target fault diagnosis information for the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information includes:

[0039] Obtain the first feature information uploaded from the information management area to the cloud;

[0040] Determine the second difference information between the first feature information and the second feature information;

[0041] Based on the second difference information, the target fault diagnosis information of the reactor cooling pump is generated by processing the first fault diagnosis information and the second fault diagnosis information.

[0042] The step of generating target fault diagnosis information for the reactor cooling pump based on the second difference information, according to the first fault diagnosis information and the second fault diagnosis information, includes:

[0043] If the second difference information is less than or equal to a preset second threshold, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information;

[0044] If the second difference information is greater than or equal to the second threshold, the second initial fault diagnosis information is used as the target fault diagnosis information.

[0045] The AI-based fault diagnosis system for small modular reactors (SMRs) executes the AI-based fault diagnosis method for SMRs, comprising: a second acquisition unit, a third acquisition unit, a second processing unit, a second input unit, and a generation unit.

[0046] The second acquisition unit is used to acquire the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; the third acquisition unit is used to acquire the second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; the second processing unit is used to process the second initial acquisition data to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information and the second pressure information of the reactor cooling pump; the second input unit is used to input the second vibration information, the second temperature information and the second pressure information into a preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump; the generation unit is used to generate the target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0047] The beneficial effects of this invention are as follows: Based on the data diode in the information management area, target operating data of the reactor cooling pump of the nuclear power small modular reactor (SMR) is acquired from the production control area of ​​the SMR; the target operating data is processed to obtain the first characteristic information of the reactor cooling pump; the first characteristic information is input into a preset fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump; the first initial fault diagnosis information is sent to the cloud to generate the target fault diagnosis information of the reactor cooling pump in the production control area. This realizes the secure acquisition and diagnosis of data using the data diode in the information management area, solves the network security problem of the nuclear power small modular reactor, and enables secure acquisition of data in the production control area, avoiding network security risks, while achieving efficient fault diagnosis. Attached Figure Description

[0048] Figure 1 An architecture diagram of an artificial intelligence-based fault diagnosis method for small modular reactors provided in an embodiment of the present invention;

[0049] Figure 2 A schematic diagram of the first process of the fault diagnosis method for small nuclear power plants based on artificial intelligence provided in an embodiment of the present invention;

[0050] Figure 3 A schematic diagram of the second process of the fault diagnosis method for small nuclear power reactors based on artificial intelligence provided in an embodiment of the present invention;

[0051] Figure 4 A schematic block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] This invention provides a fault diagnosis method and system for small modular reactors (SMRs) based on artificial intelligence. The method includes: acquiring target operating data of the reactor cooling pumps of the SMR collected from the production control area of ​​the SMR using a data diode in the information management area; processing the target operating data to obtain first characteristic information of the reactor cooling pumps; inputting the first characteristic information into a preset fault diagnosis model to obtain first initial fault diagnosis information of the reactor cooling pumps; and sending the first initial fault diagnosis information to the cloud to generate target fault diagnosis information of the reactor cooling pumps in the production control area. This method enables secure data acquisition and diagnosis using data diodes in the information management area, solves the network security problem of SMRs, and thus enables secure acquisition of data from the production control area, avoiding network security risks while achieving efficient fault diagnosis.

[0054] like Figure 1 As shown, the fault diagnosis method for nuclear power small modular reactor 100 based on artificial intelligence provided in this embodiment of the invention can be applied to the information management area 110 or the production control area 120 of the nuclear power small modular reactor 100.

[0055] The nuclear power small modular reactor 100 is equipped with an information management area 110 and a production control area 120. Data transmission between the information management area 110 and the production control area is achieved using data diodes. Both the information management area 110 and the production control area 120 communicate with the cloud 200. Meanwhile, the fault diagnosis method for the nuclear power small modular reactor 100 based on artificial intelligence provided by this invention is mainly used to diagnose faults in the reactor cooling pump 130 of the nuclear power small modular reactor 100.

[0056] The following is a detailed description of the fault diagnosis method for small nuclear power reactors based on artificial intelligence provided by this invention.

[0057] like Figure 2 As shown, the fault diagnosis method for small nuclear power reactors based on artificial intelligence is applied to the information management area. The method includes the following steps S210~S240.

[0058] S210. Based on the data diodes in the information management area, acquire the target operating data of the reactor cooling pumps of the nuclear power small modular reactor collected in the production control area of ​​the nuclear power small modular reactor.

[0059] S220. Process the target operating data to obtain the first characteristic information of the reactor cooling pump;

[0060] S230. Input the first feature information into the preset fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump.

[0061] S240. Send the first initial fault diagnosis information to the cloud to generate target fault diagnosis information for the reactor cooling pump in the production control area.

[0062] In this embodiment, a small nuclear power reactor can be understood as a small modular reactor, characterized by its smaller power output, modular construction, and inherent safety.

[0063] The information management area can be understood as the area within a nuclear power plant used for non-production control information processing such as data management, analysis, and office work. It is usually physically or logically isolated from the production control area.

[0064] A data diode can be understood as a unidirectional data transmission device. Its physical or logical design ensures that data can only be transmitted in one direction, preventing any reverse data flow, thereby providing a high level of network isolation.

[0065] The production control area can be understood as the area within a nuclear power plant that is directly responsible for core production operations such as reactor operation, equipment control, and safety protection, and has extremely high requirements for network security.

[0066] Reactor cooling pumps can be understood as key equipment in small nuclear reactors used for circulating coolant, and their operating status directly affects reactor safety.

[0067] The target operating data can be understood as the raw or pre-processed operating parameters collected from the reactor cooling pump, such as vibration, temperature, and pressure.

[0068] The first feature information can be understood as structured data extracted from the target operating data to characterize the operating status or failure mode of the reactor cooling pump.

[0069] A fault diagnosis model can be understood as a pre-trained artificial intelligence model, such as a machine learning model or a deep learning model, used to determine whether a device has a fault and its type based on input feature information.

[0070] The initial fault diagnosis information can be understood as the preliminary diagnosis results output by the fault diagnosis model in the information management area based on the first feature information.

[0071] The cloud can be understood as a distributed computing environment that provides computing resources, storage services, and network services. In this embodiment, it serves as an intermediate platform for the secure transmission of diagnostic information between the information management area and the production control area.

[0072] The target fault diagnosis information can be understood as the fault diagnosis results of the reactor cooling pumps that are ultimately generated in the production control area and used to guide operational decisions.

[0073] Specifically, in the process of acquiring target operating data of the reactor cooling pump of a small modular reactor (SMR) from the production control area, this invention can monitor the operating status of the reactor cooling pump in real time using sensors within the production control area, such as collecting parameters like vibration, temperature, and pressure. The target operating data is transmitted to the input of a data diode through physical or logical isolation, for example, via a data link configured for unidirectional transmission. The data diode ensures that the target operating data can only flow from the production control area to the information management area, thereby preventing any potentially malicious instructions or data from the information management area back into the production control area. The output of the data diode transmits the received target operating data to the information management area. For example, a fiber optic-connected data diode can be used to transmit data from the production control area to the information management area in the form of optical signals. The information management area receives the optical signals and converts them into electrical signals for processing.

[0074] Meanwhile, during the processing of target operational data, the target operational data received by the information management area may contain noise or redundant information. Therefore, preprocessing of the target operational data is necessary. For example, a simple digital filter can be used to denoise the target operational data, or the data can be normalized to eliminate differences in the dimensions and ranges of different sensors. The processed data is then used to extract key features reflecting the operating status of the equipment. For example, statistical features of the data, such as mean, variance, peak value, kurtosis, etc., can be calculated, or time-domain and frequency-domain analysis can be performed to extract spectral features. These extracted feature information are combined to form the first feature information of the reactor cooling pump.

[0075] Furthermore, in the process of inputting the first feature information into a pre-set fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump, a fault diagnosis model can be pre-deployed and trained in the information management area. The fault diagnosis model can be a rule-based expert system or a simple decision tree model. Upon receiving the first feature information, the fault diagnosis model analyzes and reasons about the input data, and outputs a preliminary judgment about the current operating status of the reactor cooling pump based on its internal logic or learned patterns, such as "normal," "minor abnormality," or "potential fault." This preliminary judgment is the first initial fault diagnosis information.

[0076] Finally, the initial fault diagnosis information is sent to the cloud to generate target fault diagnosis information for the reactor cooling pump in the production control area. Specifically, the information management area uploads the generated initial fault diagnosis information to the cloud via a secure network channel. The cloud, acting as a relay platform, receives and stores the initial fault diagnosis information. The production control area retrieves this initial fault diagnosis information from the cloud through its own network interface. In the production control area, this information can be combined with other real-time operational data or operator experience within the production control area. After further evaluation and confirmation, the target fault diagnosis information for the reactor cooling pump is finally generated to guide operator decision-making. For example, the production control area can simply use the initial fault diagnosis information obtained from the cloud directly as the target fault diagnosis information, or use it after local format conversion.

[0077] In this invention, by deploying the fault diagnosis system in the information management area and utilizing data diodes to achieve secure one-way transmission of data from the production control area, the network security risks introduced by intelligent technologies are effectively isolated. Simultaneously, this invention processes operational data and performs model diagnosis to generate preliminary diagnostic information, which is then securely transmitted to the production control area via the cloud. Ultimately, reliable fault diagnosis information is generated, ensuring the secure acquisition of operational data from key equipment in small modular reactors (SMRs) and the reliable generation of diagnostic results, thus preventing potential network attacks from impacting nuclear safety.

[0078] In some embodiments, based on the data diodes of the information management area, the target operating data of the reactor cooling pump of the nuclear power small modular reactor (SMR) collected in the production control area includes: receiving first initial acquisition data of the reactor cooling pump collected in the production control area; processing the first initial acquisition data; and transmitting the data using the data diodes to obtain the target operating data of the reactor cooling pump from the first initial acquisition data.

[0079] In this embodiment, by receiving the initial data collected from the reactor cooling pump in the production control area, raw, unprocessed operational data can be directly obtained from the production control area of ​​the small modular reactor (SMR). The production control area is the core area for the operation of the SMR, and the data collected therein directly reflects the real-time operating status of the reactor cooling pump.

[0080] For example, a data acquisition unit can be deployed in the production control area. This unit is directly connected to the sensors of the reactor cooling pump (such as vibration sensors, temperature sensors, pressure sensors, etc.) to collect the analog or digital signals output by the sensors in real time and convert them into a unified data format.

[0081] Alternatively, the present invention can also obtain these raw data streams from existing monitoring systems (such as DCS systems) through the data bus or industrial Ethernet within the production control area, ensuring the real-time nature and integrity of the data.

[0082] Specifically, in processing the first initial acquired data, this invention can preprocess the received raw data to improve data quality, remove noise, ensure data format uniformity, and prepare for subsequent data transmission and fault diagnosis. The first initial acquired data includes vibration data collected by a vibration sensor, temperature data collected by a temperature sensor, and pressure data collected by a pressure sensor.

[0083] For example, it may include signal conditioning, such as amplifying, filtering, or performing analog-to-digital conversion on analog signals; or format conversion and data cleaning on digital signals to remove outliers or redundant data. Furthermore, the invention may also include data verification, such as checking the integrity and consistency of data using cyclic redundancy check (CRC) or hash algorithms to prevent errors or tampering during data acquisition or initial transmission.

[0084] Furthermore, this invention employs a data diode for data transmission. A data diode is a physical or logical unidirectional data transmission mechanism. Its core function is to ensure that data can only flow in one direction (from the production control area to the information management area) and cannot flow in the opposite direction. This is crucial for scenarios like small modular reactors in nuclear power plants, which have extremely high network security requirements, and can effectively prevent potential network attacks from the information management area from penetrating into the production control area.

[0085] For example, a physical data diode typically consists of two independent network interfaces: one interface connects to the production control area and the other interface connects to the information management area. The two interfaces transmit data unidirectionally via fiber optic cables or dedicated hardware, and the irreversibility of the data flow is ensured at the hardware level.

[0086] Alternatively, logic data diodes can be implemented through a combination of software and hardware. For example, a strict one-way access control policy can be configured on the gateway device, combined with technologies such as packet filtering and protocol conversion, to ensure that only specific types of data can pass through unidirectionally and that reverse connections cannot be established.

[0087] Simultaneously, this invention obtains target operating data of the reactor cooling pump from the initial acquired data, thereby extracting key operating parameters truly useful for fault diagnosis. For example, after data transmission is complete, the data receiving module in the information management area parses predefined data fields from the data stream, such as vibration amplitude, frequency, temperature, and pressure values, and stores them in a local database or memory for use by the fault diagnosis model. Alternatively, during data transmission, the receiving end of the data diode can further parse and verify the data to ensure its validity and integrity, and then output the compliant data as the target operating data.

[0088] In this invention, by receiving the initial data collected from the production control area, the originality and authority of the data required for fault diagnosis are ensured, as it directly originates from the core area of ​​the small modular reactor (SMR) operation. Processing the initial data effectively removes noise and redundant information, improving data quality and laying the foundation for subsequent accurate diagnosis. Simultaneously, data diodes are used for data transmission, achieving unidirectional isolation between the production control area and the information management area at the physical or logical level. This fundamentally eliminates the possibility of network security risks in the information management area spreading to the production control area, greatly enhancing the security of SMR operation data transmission. Finally, this invention obtains the target operation data from the processed and securely transmitted initial data, ensuring that the data input to the fault diagnosis model is of high quality and high security. This significantly improves the accuracy, reliability, and overall network security of the AI-based SMR fault diagnosis method, effectively avoiding the risk of misdiagnosis or safety accidents due to data quality issues or network attacks.

[0089] In some embodiments, processing the first initial acquired data and using a data diode for data transmission to obtain target operating data of the reactor cooling pump includes: processing the acquired data using a preset signal conditioning circuit to obtain first acquired data of the reactor cooling pump; filtering the first acquired data to obtain second acquired data of the reactor cooling pump; and using a data diode for data transmission to obtain the target operating data of the reactor cooling pump from the second acquired data.

[0090] In this embodiment, the signal conditioning circuit aims to optimize the raw sensor data. The signal conditioning circuit can be configured as an analog signal conditioning module, which may include amplifiers, attenuators, impedance matching circuits, and level shifting circuits, etc., to adjust the amplitude, frequency response, and impedance characteristics of the signal to ensure it meets the input requirements of subsequent processing units. For example, weak sensor signals can be amplified by a high-precision operational amplifier to effectively improve the signal-to-noise ratio.

[0091] Meanwhile, after analog signals are digitized, signal conditioning circuits can also be configured as digital signal conditioning modules, which can be processed by digital filters, sampling rate converters or data format conversion modules to correct sampling errors, eliminate quantization noise or convert data into standard formats.

[0092] Specifically, after obtaining the first acquired data, the present invention performs filtering processing on the first acquired data, thereby further refining the data. This allows for the selective removal of unwanted noise components from the first acquired data while retaining signal information crucial for fault diagnosis. The filtering process can be implemented using various digital filter algorithms, such as Butterworth filters, Chebyshev filters, elliptic filters, or FIR / IIR filters.

[0093] In this invention, by employing a pre-set signal conditioning circuit to perform preliminary processing on the initial acquired data, the signal morphology can be effectively adjusted, and some noise and interference in the original data can be removed, thereby obtaining more stable and standardized first acquired data. Simultaneously, further filtering of the first acquired data can more thoroughly eliminate residual noise components, ensuring the purity and accuracy of the second acquired data, providing a high-quality data foundation for subsequent fault diagnosis. Finally, this invention uses data diodes for unidirectional data transmission, which not only ensures the safe and reliable acquisition of target operating data from the second acquired data, but also utilizes the physical isolation characteristics of the data diodes to fundamentally eliminate the risk of reverse penetration attacks on the production control area, greatly enhancing the network security of the nuclear power small modular reactor fault diagnosis system. This significantly improves the accuracy and reliability of the target operating data, providing high-quality and high-security input for artificial intelligence-based fault diagnosis models, thereby improving the accuracy of fault diagnosis and the overall operational safety of the nuclear power small modular reactor.

[0094] In some embodiments, data diodes are used for data transmission to obtain target operating data of the reactor cooling pump from first initial acquisition data, including: verifying second acquisition data using data diodes to obtain a verification result of the second acquisition data; encrypting the second acquisition data based on the verification result to obtain third acquisition data of the reactor cooling pump; and using data diodes for data transmission to obtain target operating data of the reactor cooling pump from the third acquisition data.

[0095] In this embodiment, the present invention uses a data diode to verify the second acquired data, thereby ensuring the integrity and authenticity of the data. For example, the present invention can use a cyclic redundancy check (CRC) algorithm, which calculates the checksum of the second acquired data and appends it to the data. The receiving end then performs the same calculation and compares it with the received checksum. If the two match, it indicates that the data has not been tampered with.

[0096] Alternatively, the present invention may employ a hash verification algorithm, such as MD5 or SHA-256, to generate a unique hash value for the second collected data. This hash value is transmitted along with the data, and the receiving end verifies the integrity of the data by recalculating the hash value and comparing it with the received hash value.

[0097] Specifically, this invention encrypts the second collected data based on the verification result, which can protect the confidentiality of the data and prevent unauthorized access or eavesdropping. Furthermore, the encryption process can be implemented by combining asymmetric and symmetric encryption algorithms. For example, the RSA algorithm can be used to encrypt the session key used for symmetric encryption, and then this session key can be used to encrypt the second collected data, thereby ensuring secure key transmission while efficiently encrypting large amounts of data.

[0098] In this invention, by verifying the second acquired data, it is possible to effectively detect whether the data has been tampered with or erroneous before or during transmission, ensuring the integrity and authenticity of the data. Based on the verification results, the data is encrypted, further enhancing the confidentiality of the data and preventing sensitive operational data from being illegally stolen or eavesdropped on. Finally, this invention utilizes the unidirectional transmission characteristics of the data diode to securely transmit the encrypted third acquired data. It not only leverages the physical isolation advantage of the data diode but also combines data verification and encryption software protection to form a multi-layered, highly secure data transmission mechanism. This significantly reduces network security risks during data transmission and effectively prevents attackers from threatening the safe operation of the nuclear power small modular reactor by tampering with or stealing data, thereby ensuring the reliability of the reactor cooling pump target operation data and the overall safety of the nuclear power system.

[0099] In some embodiments, the first feature information includes first vibration information, first temperature information, and first pressure information of the reactor cooling pump; processing the target operating data to obtain the first feature information of the reactor cooling pump includes: decrypting the target data to obtain decrypted target operating data; filtering the decrypted target operating data to obtain filtered target operating data; and extracting features from the filtered target operating data to obtain the first feature information, first temperature information, and first pressure information.

[0100] In this embodiment, the first feature information includes the first vibration information, the first temperature information, and the first pressure information of the reactor cooling pump. The first vibration information can be obtained by an acceleration sensor or vibration sensor installed on the reactor cooling pump housing or bearing housing. It can be the instantaneous value, root mean square value, or peak value of vibration acceleration, vibration velocity, or vibration displacement. Alternatively, the first vibration information can also be a specific frequency component or harmonic amplitude obtained by performing spectral analysis on the original vibration signal.

[0101] The first temperature information can be obtained through thermocouples, resistance temperature detectors (RTDs), or infrared thermometers to monitor the temperature of the pump body, bearings, or coolant; alternatively, the first temperature information can be the local temperature of critical components inside the pump (such as bearings or seals).

[0102] The first pressure information can be obtained through a pressure sensor to monitor the pump's inlet and outlet pressures; alternatively, the first pressure information can be the pump's head or differential pressure.

[0103] Furthermore, the present invention decrypts the target data to obtain decrypted target operational data, thereby ensuring that the received target operational data has not been illegally tampered with or stolen during transmission, restoring its original readable form, and guaranteeing the integrity and confidentiality of the data. The decryption process can employ an asymmetric encryption algorithm (such as RSA), where the production control area uses a public key to encrypt the data, and the information management area uses a private key to decrypt the data.

[0104] Simultaneously, this invention filters the decrypted target running data to obtain filtered target running data, thereby eliminating noise and interference in the data, improving data quality, and providing a cleaner and more reliable input for subsequent feature extraction, avoiding noise from misleading diagnostic results. For example, algorithms such as moving average filtering, median filtering, or Kalman filtering can be used to effectively suppress random noise and sudden outliers by smoothing the data sequence or estimating its state.

[0105] In this invention, target data is decrypted to obtain decrypted target operating data; the decrypted target operating data is then filtered to obtain filtered target operating data; and feature extraction is performed on the filtered target operating data to obtain first feature information, first temperature information, and first pressure information. This provides accurate and reliable fault diagnosis information, significantly improving the accuracy and reliability of fault diagnosis for small modular reactors (SMRs), effectively avoiding misdiagnosis or missed diagnosis due to data issues, and thus ensuring the safe and stable operation of SMRs.

[0106] In the artificial intelligence-based fault diagnosis method for small modular reactors (SMRs) provided in this invention embodiment, target operating data of the reactor cooling pump of the SMR is acquired from the production control area of ​​the SMR based on a data diode in the information management area. The target operating data is processed to obtain first characteristic information of the reactor cooling pump. The first characteristic information is input into a preset fault diagnosis model to obtain first initial fault diagnosis information of the reactor cooling pump. The first initial fault diagnosis information is sent to the cloud to generate target fault diagnosis information of the reactor cooling pump in the production control area. This method enables secure data acquisition and diagnosis using a data diode in the information management area, solves the network security problem of SMRs, and enables secure acquisition of data from the production control area, avoiding network security risks and achieving efficient fault diagnosis.

[0107] In some embodiments, such as Figure 3 As shown, the present invention also provides a fault diagnosis method for small nuclear power reactors based on artificial intelligence, which is applied to the production control area of ​​small nuclear power reactors. The method includes steps S310, S320, S330, S340 and S350.

[0108] S310. Based on the cloud, obtain the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor;

[0109] S320. Based on the first initial fault diagnosis information, obtain the second initial acquisition data of the reactor cooling pump of the nuclear power small reactor collected in the production control area.

[0110] S330. The second initial data acquisition is processed to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information and the second pressure information of the reactor cooling pump.

[0111] S340. Input the second vibration information, the second temperature information and the second pressure information into the preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump.

[0112] S350. Based on the first initial fault diagnosis information and the second initial fault diagnosis information, generate the target fault diagnosis information for the reactor cooling pump.

[0113] In this embodiment, the initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor is obtained from the cloud. The cloud is used as a security intermediary to achieve physical isolation between the information management area and the production control area, thereby blocking potential attack paths. Specifically, the cloud receives the diagnostic information transmitted unidirectionally from the information management area via a data diode, ensuring that the production control area can obtain external diagnostic results without directly connecting to the information management area, effectively avoiding the exposure of the network attack surface.

[0114] Furthermore, in the process of acquiring the second initial acquisition data of the reactor cooling pump of the nuclear power small modular reactor collected in the production control area based on the first initial fault diagnosis information, the cloud can send the first initial fault diagnosis information to the production control area. The production control area can parse the first initial fault diagnosis information to determine the first initial acquisition data corresponding to the first initial fault diagnosis information generated by the information management area, i.e., the second initial acquisition data. The production control area then generates the second initial fault diagnosis information based on the first initial acquisition data to determine whether the first initial fault diagnosis information generated by the information management area is accurate, and thus whether the information management area has been attacked. This not only avoids the resource waste caused by the production control area performing full data acquisition for fault diagnosis, but also ensures the timeliness and relevance of the data. The production control area has local storage, and the second initial acquisition data can be obtained from the local storage. The second initial acquisition data can be determined based on the timestamp or tag information of the first initial fault diagnosis information. The timestamp of the first initial acquisition data can be determined through the timestamp or tag information of the first initial fault diagnosis information, thus allowing the second initial acquisition data to be obtained from the local storage.

[0115] Meanwhile, the present invention processes the second initial acquisition data to obtain the second characteristic information of the reactor cooling pump. Specifically, it can denoise, normalize and extract features from the original parameters such as vibration, temperature and pressure, thereby forming structured data characterizing the operating status of the equipment.

[0116] Based on this, the present invention inputs the second feature information into a pre-set fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump. The fault diagnosis model is simultaneously deployed within the production control area, trained based on local historical data, and capable of independently performing fault mode recognition. It outputs a local diagnosis result that is mutually verified with the first initial fault diagnosis information. Finally, the present invention generates the target fault diagnosis information of the reactor cooling pump based on the first and second initial fault diagnosis information, thereby effectively resisting the risk of diagnostic information tampering caused by network attacks that may occur in the information management area.

[0117] In this invention, a local fault diagnosis closed loop is constructed in the production control area and dynamically integrated with diagnostic information from the information management area, thereby achieving cross-validation and enhanced reliability of diagnostic results under strict network security isolation conditions. Simultaneously, the cloud-based security mediation mechanism ensures unidirectional data flow into the information management area, while the complete process of local data acquisition, feature extraction, and model diagnosis establishes a secure diagnostic capability independent of external systems. Therefore, even if the intelligent system in the information management area suffers a network attack, causing distortion of the initial fault diagnosis information, the production control area can still generate reliable second initial fault diagnosis information based on locally acquired data, ultimately generating target fault diagnosis information unaffected by external interference. This fundamentally guarantees the accuracy of fault diagnosis for key equipment in small modular reactors and nuclear power plants, as well as nuclear safety.

[0118] In some embodiments, generating target fault diagnosis information for a reactor cooling pump based on first initial fault diagnosis information and second initial fault diagnosis information includes: generating first difference information between the first initial fault diagnosis information and the second initial fault diagnosis information; if the first difference information is less than or equal to a preset first threshold, fusing the first initial fault diagnosis information and the second initial fault diagnosis information to obtain target fault diagnosis information; if the first difference information is greater than or equal to the first threshold, using the second initial fault diagnosis information as target fault diagnosis information.

[0119] In this embodiment, the first difference information can quantify the degree of inconsistency between the first initial fault diagnosis information from the information management area and the second initial fault diagnosis information from the production control area.

[0120] The first threshold can be understood as a reference value used to determine whether the difference between two initial fault diagnosis information is significant. The first threshold can be set based on the operating experience of nuclear power small modular reactors, expert knowledge, or historical fault data.

[0121] Specifically, this invention employs different processing strategies based on the comparison result between the first difference information and a preset first threshold. Specifically, if the first difference information is less than or equal to the preset first threshold, it indicates that the two diagnostic information sets have a high degree of consistency. In this case, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information. The fusion operation aims to comprehensively utilize diagnostic information from different regions to obtain a more comprehensive and reliable diagnostic result. For example, different weights can be assigned based on the confidence level or source reliability of the first and second initial fault diagnosis information, and then a weighted average can be performed.

[0122] If the first difference information is greater than or equal to the first threshold, it indicates that there is a significant difference between the two diagnostic information. In this case, the second initial fault diagnosis information is used as the target fault diagnosis information. This operation means that when there is a significant difference between the two initial fault diagnosis information, the second initial fault diagnosis information from the production control area is preferentially adopted as the final target fault diagnosis information. This is usually based on the real-time nature and directness of the data source in the production control area, as well as the higher trust in local diagnostic results in nuclear safety scenarios.

[0123] In this invention, by generating first difference information and comparing it with a preset first threshold, the system can intelligently determine the degree of consistency between two initial fault diagnosis information, providing an objective basis for subsequent decisions, avoiding subjective misjudgments, and thus effectively addressing network security risks and preventing information tampering from leading to incorrect decisions. Especially in scenarios with extremely high safety requirements, such as small modular reactors (SMRs), prioritizing the use of local diagnostic results that are closer to the physical equipment and less susceptible to external interference can effectively avoid the impact of potential network security risks or data transmission errors, ensuring the robustness of fault diagnosis results and the operational safety of SMRs.

[0124] In some embodiments, generating target fault diagnosis information for the reactor cooling pump based on first initial fault diagnosis information and second initial fault diagnosis information includes: acquiring first feature information uploaded from the information management area to the cloud; determining second difference information between the first feature information and the second feature information; and generating target fault diagnosis information for the reactor cooling pump based on the second difference information and processing according to the first fault diagnosis information and the second fault diagnosis information.

[0125] In this embodiment, the first feature information can be understood as an abstract representation of the key operating parameters of the reactor cooling pump obtained from the information management area and uploaded to the cloud. The first feature information may include the reactor cooling pump's first vibration information, first temperature information, and first pressure information.

[0126] Specifically, by acquiring first feature information, the present invention can provide a securely isolated data source, independent of the production control area, for subsequent comparison with feature information generated in the production control area, thereby enhancing the reliability and security of fault diagnosis.

[0127] The second difference information is an indicator used to quantify the degree of inconsistency between the first and second feature information. The first feature information originates from the information management area and represents the data characteristics after secure transmission and processing via data diodes; the second feature information originates from the data characteristics collected and processed locally in the production control area. By comparing these two sets of feature information, this invention can detect anomalies, tampering, or deviations that may occur during data transmission, processing, or collection.

[0128] The first fault diagnosis information is the diagnosis result generated in the cloud model based on data from the information management area; the second fault diagnosis information is the diagnosis result generated in the local model based on local data from the production control area.

[0129] In this invention, by determining the second difference information between the first and second feature information, the system can quantify this inconsistency, providing a reliable basis for subsequent diagnostic decisions. Simultaneously, based on this second difference information, the system intelligently processes the first and second fault diagnosis information, dynamically adjusting the diagnostic strategy according to the degree of data consistency, thereby improving the accuracy and robustness of the diagnosis. Furthermore, when there are significant differences in feature information, prioritizing the adoption of local diagnostic results or triggering alarms effectively avoids the risk of misdiagnosis caused by external attacks or data transmission errors, thus significantly enhancing the safety and reliability of the nuclear power small modular reactor (SMR) fault diagnosis system and ensuring the safe and stable operation of the SMR.

[0130] In some embodiments, based on the second difference information, the target fault diagnosis information of the reactor cooling pump is generated by processing the first fault diagnosis information and the second fault diagnosis information, including: if the second difference information is less than or equal to a preset second threshold, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information; if the second difference information is greater than or equal to the second threshold, the second initial fault diagnosis information is used as the target fault diagnosis information.

[0131] In this embodiment, the second difference information can be understood as an indicator used in the fault diagnosis process of a small modular reactor (SMR) to quantify the degree of difference between the first characteristic information from the information management area and the second characteristic information from the production control area. The second difference information can be used to assess the consistency or deviation between data acquired or processed in different safety areas.

[0132] The second threshold can be understood as a predetermined value. The second threshold can serve as a benchmark for judging the magnitude of the second difference information, thereby guiding the processing strategy of subsequent fault diagnosis information.

[0133] Specifically, when the second difference information is less than or equal to the preset second threshold, it indicates that there is a high degree of consistency between the first feature information from the information management area and the second feature information from the production control area. At this time, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to generate more comprehensive, accurate and more confident target fault diagnosis information.

[0134] When the second difference information is greater than or equal to a preset second threshold, it indicates a significant difference between the first feature information from the information management area and the second feature information from the production control area. This may indicate abnormal data transmission, sensor malfunction, or even potential network attack risks. In this case, the present invention adopts a security-first strategy, using the second initial fault diagnosis information as the target fault diagnosis information, thereby effectively avoiding the risk of misjudgment caused by possible tampering or anomalies in the data of the information management area.

[0135] In this invention, by introducing a mechanism for processing fault diagnosis information based on second difference information, the problem of how to ensure the accuracy and safety of diagnosis results when there are large differences in data from different safety areas during the fault diagnosis process of nuclear power small modular reactors is effectively solved. Thus, while ensuring the accuracy of diagnosis, the network security resilience of the system can be significantly improved.

[0136] In the artificial intelligence-based fault diagnosis method for small modular reactors (SMRs) provided in this invention embodiment, first initial fault diagnosis information generated by the information management area of ​​the SMR is obtained based on the cloud; second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area is obtained based on the first initial fault diagnosis information; the second initial acquisition data is processed to obtain second characteristic information of the reactor cooling pump; the second characteristic information includes second vibration information, second temperature information, and second pressure information of the reactor cooling pump; the second vibration information, second temperature information, and second pressure information are input into a preset fault diagnosis model to obtain second initial fault diagnosis information of the reactor cooling pump; and target fault diagnosis information of the reactor cooling pump is generated based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0137] In some embodiments, the present invention also provides an artificial intelligence-based fault diagnosis system for small modular reactors, which is used to perform any of the aforementioned artificial intelligence-based fault diagnosis methods for small modular reactors.

[0138] The fault diagnosis system for small modular reactors based on artificial intelligence provided by the present invention includes: a first acquisition unit, a first processing unit, a first input unit, and a transmission unit.

[0139] The first acquisition unit is used to acquire target operating data of the reactor cooling pump of the nuclear power small modular reactor (SMR) collected in the production control area of ​​the nuclear power small modular reactor based on the data diodes in the information management area; the first processing unit is used to process the target operating data to obtain the first characteristic information of the reactor cooling pump; the first input unit is used to input the first characteristic information into a preset fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump; and the sending unit is used to send the first initial fault diagnosis information to the cloud to generate the target fault diagnosis information of the reactor cooling pump in the production control area.

[0140] The artificial intelligence-based fault diagnosis system for small modular reactors (SMRs) provided in this invention can acquire target operating data of the reactor cooling pumps collected in the production control area of ​​the SMR based on data diodes in the information management area; process the target operating data to obtain first characteristic information of the reactor cooling pumps; input the first characteristic information into a preset fault diagnosis model to obtain first initial fault diagnosis information of the reactor cooling pumps; and send the first initial fault diagnosis information to the cloud to generate target fault diagnosis information of the reactor cooling pumps in the production control area. This system enables secure data acquisition and diagnosis using data diodes in the information management area, solves the network security problem of SMRs, and thus enables secure acquisition of data from the production control area, avoiding network security risks while achieving efficient fault diagnosis.

[0141] In some embodiments, the present invention also provides an artificial intelligence-based fault diagnosis system for small modular reactors, which is used to perform any of the aforementioned artificial intelligence-based fault diagnosis methods for small modular reactors.

[0142] The fault diagnosis system for small modular reactors based on artificial intelligence provided by the present invention includes: a second acquisition unit, a third acquisition unit, a second processing unit, a second input unit, and a generation unit.

[0143] The second acquisition unit is used to acquire the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; the third acquisition unit is used to acquire the second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; the second processing unit is used to process the second initial acquisition data to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information and the second pressure information of the reactor cooling pump; the second input unit is used to input the second vibration information, the second temperature information and the second pressure information into a preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump; the generation unit is used to generate the target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0144] The artificial intelligence-based fault diagnosis system for small modular reactors (SMRs) provided in this invention can acquire first initial fault diagnosis information generated by the information management area of ​​the SMR in the cloud; based on the first initial fault diagnosis information, acquire second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area; process the second initial acquisition data to obtain second characteristic information of the reactor cooling pump; the second characteristic information includes second vibration information, second temperature information, and second pressure information of the reactor cooling pump; input the second vibration information, second temperature information, and second pressure information into a preset fault diagnosis model to obtain second initial fault diagnosis information of the reactor cooling pump; and generate target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0145] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above-mentioned artificial intelligence-based fault diagnosis system for small modular reactors and its various units can be found in the corresponding descriptions in the aforementioned method embodiments. For the sake of convenience and brevity, these details will not be repeated here.

[0146] The aforementioned AI-based fault diagnosis system for small modular reactors can be implemented as a computer program, which can perform tasks such as... Figure 4 It runs on the electronic device shown.

[0147] See Figure 4 The device 400 includes a processor 402, a memory, and a network interface 405 connected via a system bus 401, wherein the memory may include a storage medium 403 and internal memory 404.

[0148] The storage medium 403 may store an operating system 4031 and a computer program 4032. When the computer program 4032 is executed, it enables the processor 402 to execute an artificial intelligence-based fault diagnosis method for small nuclear power reactors.

[0149] The processor 402 provides computing and control capabilities to support the operation of the entire device 400.

[0150] The internal memory 404 provides an environment for the operation of the computer program 4032 in the non-volatile storage medium 403. When the computer program 4032 is executed by the processor 402, the processor 402 can execute a fault diagnosis method for small nuclear power reactors based on artificial intelligence.

[0151] This network interface 405 is used for network communication, such as providing data transmission. Those skilled in the art will understand that... Figure 4The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the device 400 to which the present invention is applied. The specific device 400 may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0152] The processor 402 is used to run the computer program 4032 stored in the memory to perform the following functions: based on the data diodes in the information management area, it acquires the target operating data of the reactor cooling pump of the nuclear power small modular reactor collected in the production control area; processes the target operating data to obtain the first characteristic information of the reactor cooling pump; inputs the first characteristic information into a preset fault diagnosis model to obtain the first initial fault diagnosis information of the reactor cooling pump; and sends the first initial fault diagnosis information to the cloud to generate the target fault diagnosis information of the reactor cooling pump in the production control area.

[0153] The processor 402 is used to run a computer program 4032 stored in the memory to perform the following functions: acquiring first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; acquiring second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; processing the second initial acquisition data to obtain second characteristic information of the reactor cooling pump; the second characteristic information includes second vibration information, second temperature information, and second pressure information of the reactor cooling pump; inputting the second vibration information, second temperature information, and second pressure information into a preset fault diagnosis model to obtain second initial fault diagnosis information of the reactor cooling pump; and generating target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0154] Those skilled in the art will understand that Figure 4 The embodiments of device 400 shown do not constitute a limitation on the specific configuration of device 400. In other embodiments, device 400 may include more or fewer components than shown, or combine certain components, or have different component arrangements. For example, in some embodiments, device 400 may include only memory and processor 402. In such embodiments, the structure and function of memory and processor 402 are similar to those shown. Figure 4 The embodiments shown are consistent and will not be repeated here.

[0155] It should be understood that, in this embodiment of the invention, the processor 402 may be a Central Processing Unit (CPU), or it may be another general-purpose processor 402, a digital signal processor 402 (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor 402 may be a microprocessor 402, or it may be any conventional processor 402, etc.

[0156] According to one aspect of the present invention, a computer program product or computer program is also provided, comprising computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the following steps: acquiring target operating data of the reactor cooling pump of the nuclear power small modular reactor (SMR) collected in the production control area of ​​the SMR based on a data diode in the information management area; processing the target operating data to obtain first characteristic information of the reactor cooling pump; inputting the first characteristic information into a preset fault diagnosis model to obtain first initial fault diagnosis information of the reactor cooling pump; and sending the first initial fault diagnosis information to the cloud to generate target fault diagnosis information of the reactor cooling pump in the production control area.

[0157] The processor of the electronic device reads computer instructions from a computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the following steps: acquiring first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; acquiring second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; processing the second initial acquisition data to obtain second characteristic information of the reactor cooling pump; the second characteristic information includes second vibration information, second temperature information, and second pressure information of the reactor cooling pump; inputting the second vibration information, second temperature information, and second pressure information into a preset fault diagnosis model to obtain second initial fault diagnosis information of the reactor cooling pump; and generating target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0158] It will be understood by those skilled in the art that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program includes program instructions and can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0159] In another embodiment of the present invention, a computer storage medium is provided. This storage medium may be a non-volatile computer-readable storage medium or a volatile storage medium. The storage medium stores a computer program 4032, wherein when executed by a processor 402, the computer program 4032 performs the following steps: in response to a first operation request from a user to learn online, determining the course category for which the user is learning online; allocating course content corresponding to the course category to the user's user terminal for online learning; the course content includes at least one of information security course content, corporate culture course content, skills and techniques course content, and rules and regulations course content; in response to a second operation request from a user to take an online exam of the course content, allocating exam content corresponding to the course content to the user terminal for online exams; obtaining learning information from the user's online learning and exam information from the user's online exams; evaluating the user's online learning based on the learning information and exam information to obtain the user's learning effectiveness.

[0160] The storage medium stores a computer program 4032, which, when executed by the processor 402, performs the following steps: acquiring first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; acquiring second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; processing the second initial acquisition data to obtain second characteristic information of the reactor cooling pump; the second characteristic information includes second vibration information, second temperature information, and second pressure information of the reactor cooling pump; inputting the second vibration information, second temperature information, and second pressure information into a preset fault diagnosis model to obtain second initial fault diagnosis information of the reactor cooling pump; and generating target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

[0161] The storage medium can be any computer-readable storage medium that can store program code, such as a USB flash drive, external hard drive, read-only memory (ROM), magnetic disk, or optical disk.

[0162] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0163] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for diagnosing a failure of a nuclear small modular reactor based on artificial intelligence, characterized by, The method, applied to the production control area of ​​a small modular reactor (SMR), includes: Based on the cloud, the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor is obtained; Based on the first initial fault diagnosis information, the second initial acquisition data of the reactor cooling pump of the nuclear power small reactor collected in the production control area is obtained; The second initial data acquisition is processed to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information, and the second pressure information of the reactor cooling pump. The second vibration information, the second temperature information, and the second pressure information are input into a preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump. Based on the first initial fault diagnosis information and the second initial fault diagnosis information, target fault diagnosis information for the reactor cooling pump is generated. The step of generating target fault diagnosis information for the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information includes: Generate a first difference information between the first initial fault diagnosis information and the second initial fault diagnosis information; If the first difference information is less than or equal to a preset first threshold, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information; If the first difference information is greater than or equal to the first threshold, the second initial fault diagnosis information is used as the target fault diagnosis information.

2. The fault diagnosis method for small modular reactors based on artificial intelligence as described in claim 1, characterized in that, The step of generating target fault diagnosis information for the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information includes: Obtain the first feature information uploaded from the information management area to the cloud; Determine the second difference information between the first feature information and the second feature information; Based on the second difference information, the target fault diagnosis information of the reactor cooling pump is generated by processing the first initial fault diagnosis information and the second initial fault diagnosis information.

3. The fault diagnosis method for small modular reactors based on artificial intelligence as described in claim 2, characterized in that, The step of generating target fault diagnosis information for the reactor cooling pump based on the second difference information, according to the first initial fault diagnosis information and the second initial fault diagnosis information, includes: If the second difference information is less than or equal to a preset second threshold, the first initial fault diagnosis information and the second initial fault diagnosis information are fused to obtain the target fault diagnosis information; If the second difference information is greater than or equal to the second threshold, the second initial fault diagnosis information is used as the target fault diagnosis information.

4. A fault diagnosis system for small modular reactors based on artificial intelligence, characterized in that: The fault diagnosis method for small nuclear power reactors based on artificial intelligence according to any one of claims 1-3 includes: a second acquisition unit, a third acquisition unit, a second processing unit, a second input unit, and a generation unit; The second acquisition unit is used to acquire the first initial fault diagnosis information generated by the information management area of ​​the nuclear power small modular reactor (SMR) based on the cloud; the third acquisition unit is used to acquire the second initial acquisition data of the reactor cooling pump of the SMR collected in the production control area based on the first initial fault diagnosis information; the second processing unit is used to process the second initial acquisition data to obtain the second characteristic information of the reactor cooling pump; the second characteristic information includes the second vibration information, the second temperature information and the second pressure information of the reactor cooling pump; the second input unit is used to input the second vibration information, the second temperature information and the second pressure information into a preset fault diagnosis model to obtain the second initial fault diagnosis information of the reactor cooling pump; the generation unit is used to generate the target fault diagnosis information of the reactor cooling pump based on the first initial fault diagnosis information and the second initial fault diagnosis information.

Images