Method for detecting signal anomaly of neutron detector of out-of-core nuclear measurement system

Abstract

The invention particularly relates to a signal anomaly detection method for a neutron detector of an ex-core nuclear measurement system. The signal anomaly detection method comprises the following steps: step 101, acquiring a current signal of the neutron detector in real time; 102, performing feature engineering on the current signal of the neutron detector to obtain feature data of the neutron detector; step 103, dividing the characteristic data of the neutron detector into data modes, and inputting the data modes into the trained twin model of the neutron detector; and step 104, the trained twin model of the neutron detector outputs feature data reconstructed by the neutron detector and performs signal anomaly detection of the neutron detector. The invention further relates to a signal anomaly detection device for the neutron detector of the out-of-core nuclear measurement system, computer equipment and a storage medium. According to the invention, the signal abnormity of the neutron detector of the out-of-core nuclear measurement system is efficiently detected, and an auxiliary judgment basis is provided for state monitoring of related reactor core equipment or parts in a nuclear power plant.

Description

Technical Field

[0001] The present invention relates to the technical field of neutron detector signal anomaly detection in an ex-core nuclear measurement system, and in particular to a method, device, computer equipment and storage medium for neutron detector signal anomaly detection in an ex-core nuclear measurement system. Background Art

[0002] The normal operation and power regulation of nuclear power units are based on accurate measurement of nuclear reactor power. When a neutron detector measurement failure occurs, it may cause unplanned shutdown of the nuclear power unit, resulting in greater economic losses and reduced system reliability. Therefore, in order to ensure the normal operation of the nuclear power unit and reduce the economic losses caused by unplanned shutdowns, it is necessary to perform fault detection on the neutron detectors of the off-core nuclear measurement system and replace the failed neutron detectors in a timely manner to ensure the stable operation of the nuclear power unit and avoid greater economic losses.

[0003] In order to ensure the normal output of neutron detectors, it is necessary to detect the status of neutron detectors. At present, there are studies on the detection of neutron detector failures of various structures in nuclear power plants. For example, some scholars at home and abroad have realized the diagnosis of neutron detector failures through various methods such as generalized likelihood ratio, principal component analysis, and kmeans algorithm. At present, there are some methods for detecting neutron detector failures in ex-core nuclear measurement systems, but they cannot detect neutron detector signal anomalies in ex-core nuclear measurement systems very well. Signal anomaly detection is a common maintenance measure. When a signal anomaly occurs, it can detect the device with signal anomalies in time and issue an early warning to reduce the damage and risks that may be caused by signal anomalies. Summary of the invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment and storage medium for detecting abnormality of neutron detector signals in an extra-core nuclear measurement system of a nuclear power plant, in order to address the problems of difficulty in judging abnormality of neutron detector signals and complexity in evaluating health status in an extra-core nuclear measurement system of a nuclear power plant, so as to realize efficient detection of abnormality of neutron detector signals in an extra-core nuclear measurement system and provide auxiliary judgment basis for status monitoring of relevant core equipment or components in a nuclear power plant.

[0005] In order to solve the above problems, the present invention provides a method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system, comprising the following steps:

[0006] Step 101, obtaining a current signal of a neutron detector in real time;

[0007] Step 102: Perform feature engineering on the current signal of the neutron detector to obtain feature data of the neutron detector;

[0008] Step 103, dividing the characteristic data of the neutron detector into data patterns and inputting the data patterns into the trained neutron detector twin model;

[0009] Step 104: The trained neutron detector twin model outputs the reconstructed feature data of the neutron detector and performs neutron detector signal anomaly detection.

[0010] As one of the achievable methods, in step 103, the neutron detector twin model is formed by fusing the twin sub-models of each group of neutron detectors in the ex-core nuclear measurement system;

[0011] The twin model of each group of neutron detectors is built based on the twin relationship between the characteristic data of this group of neutron detectors and the characteristic data of other groups of neutron detectors, combined with the random forest regression model;

[0012] The twin model of each group of neutron detectors takes the characteristic data of other groups of neutron detectors as input, takes the reconstructed characteristic data of the neutron detectors of this group as output and compares the reconstructed characteristic data of the neutron detectors of this group with the actual characteristic data;

[0013] If the difference between the reconstructed characteristic data and the actual characteristic data of this group of neutron detectors exceeds the signal anomaly detection threshold of this group of neutron detectors, the twin model of this group of neutron detectors determines that the signal of this group of neutron detectors is abnormal and issues an early warning.

[0014] As one of the achievable methods, in step 103, the trained neutron detector twin model is obtained by the following steps:

[0015] Step 201, obtaining historical current signals of each group of neutron detectors of the out-of-core nuclear measurement system through monitoring equipment;

[0016] Step 202: Perform feature engineering on the historical current signals of each group of neutron detectors to obtain feature data of each group of neutron detectors; divide the feature data of each group of neutron detectors into training data and test data, the training data of each group of neutron detectors constitutes a training data set, and the test data of each group of neutron detectors constitutes a test data set;

[0017] Step 203: preprocess the data in the training data set and the test data set respectively;

[0018] Step 204: training the neutron detector twin model using the preprocessed data in the training data set to obtain a trained neutron detector twin model; the trained neutron detector twin model implements neutron detector signal anomaly detection;

[0019] Step 205: Test the trained neutron detector twin model using the preprocessed data in the test data set. If the test passes, the training of the neutron detector twin model is terminated; if the test fails, return to step 204.

[0020] As one possible implementation method, in step 203, preprocessing the data in the training data set and the test data set respectively includes the following steps:

[0021] The characteristic data of each group of neutron detectors in the training data set and the test data set are divided into data patterns of each group of neutron detectors; the data pattern of each group of neutron detectors includes the characteristic data of other groups of neutron detectors at the same time and the characteristic data of the neutron detectors in this group.

[0022] As one of the achievable methods, in step 204, the neutron detector twin model is trained by using the preprocessed data in the training data set to obtain the trained neutron detector twin model, including the following steps:

[0023] Step 301: input the characteristic data of each group of neutron detectors in the training data set at the same time as other groups of neutron detectors into the twin model of the current group of neutron detectors, and the twin model of the current group of neutron detectors outputs the reconstructed characteristic data of the current group of neutron detectors;

[0024] Step 302: The twin model of the neutron detectors in the present group determines the signal anomaly detection threshold of the neutron detectors in the present group according to the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in the present group, and obtains the trained twin model of the neutron detectors in the present group;

[0025] Step 303: The trained twin models of each group of neutron detectors are fused to form a trained neutron detector twin model;

[0026] The characteristic data of each group of neutron detectors in the training data set are the characteristic data of each group of neutron detectors in a normal state.

[0027] As one of the achievable methods, in step 205, the trained neutron detector twin model is tested by using the preprocessed data in the test data set, including the following steps:

[0028] Step 401, input the characteristic data of each group of neutron detectors in the test data set at the same time as other groups of neutron detectors into the trained twin model of the group of neutron detectors;

[0029] The trained twin model of the group of neutron detectors outputs the reconstructed feature data of the group of neutron detectors and compares the actual feature data of the group of neutron detectors with the reconstructed feature data;

[0030] If the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in this group exceeds the signal anomaly detection threshold of the neutron detectors in this group, the trained twin model of the neutron detectors in this group determines that the signal of the neutron detectors in this group is abnormal;

[0031] Step 402: Calculate the mean square error of the signal anomaly detection result of the trained twin model of the neutron detectors in the group and the actual state of the neutron detectors in the group to obtain the signal anomaly detection error of the trained twin model of the neutron detectors in the group;

[0032] If the error of the signal anomaly detection of the neutron detectors in this group by the trained twin model exceeds the set threshold, it is determined that the trained twin model of the neutron detectors in this group has failed the test;

[0033] If the error of the trained twin model for detecting abnormal signals of the neutron detectors in this group is less than the set threshold, the trained twin model of the neutron detectors in this group is judged to have passed the test;

[0034] Step 403: If all the trained twin models of each group of neutron detectors pass the test, the trained neutron detector twin model passes the test, and the training of the neutron detector twin model ends;

[0035] If the trained twin model of any group of neutron detectors fails the test, the process returns to step 204 to train the twin model of the current group of neutron detectors.

[0036] The characteristic data of each group of neutron detectors in the test data set are characteristic data of known states of each group of neutron detectors.

[0037] In order to solve the above problems, the present invention also provides a neutron detector signal abnormality detection device in an out-of-core nuclear testing system, comprising:

[0038] A data acquisition module, used to acquire the current signal of the neutron detector in real time and send it to a data processing module;

[0039] A data processing module is used to receive the current signal of the neutron detector sent by the data acquisition module and perform feature engineering to obtain the feature data of the neutron detector; divide the feature data of the neutron detector into data modes and send them to the judgment module;

[0040] The judgment module has an embedded trained neutron detector twin model, which is used to receive the data pattern of the neutron detector sent by the data processing module and input it into the trained neutron detector twin model; the trained neutron detector twin model outputs the characteristic data reconstructed by the neutron detector and performs neutron detector signal anomaly detection.

[0041] In order to solve the above technical problems, the present invention provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the above-mentioned method for detecting abnormality of neutron detector signals in an extra-core nuclear measurement system are implemented.

[0042] In order to solve the above technical problems, the present invention provides a computer-readable storage medium, on which computer-readable instructions are stored. When the computer-readable instructions are executed, the steps of the above-mentioned method for detecting abnormality of neutron detector signals in the extra-core nuclear measurement system are implemented.

[0043] Beneficial technical effects of the present invention:

[0044] The method, device, computer equipment and storage medium for detecting abnormal signals of neutron detectors in an ex-core nuclear measurement system of the present invention are based on the twin relationship between the characteristic data of neutron detectors in the ex-core nuclear measurement system, and are combined with a random forest regression model with high regression prediction accuracy to build a neutron detector twin model, so as to achieve 100% accuracy in detecting abnormal signals of neutron detectors in the ex-core nuclear measurement system, and provide an auxiliary judgment basis for status monitoring of relevant core equipment or components in nuclear power plants. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 It is a flow chart of an embodiment of a method for detecting abnormality of a neutron detector signal in an out-of-core nuclear testing system of the present invention;

[0046] Figure 2 It is a schematic structural view of an embodiment of a device for detecting abnormality of a sub-detector signal in an out-of-core nuclear testing system of the present invention.

[0047] Figure 3 The figure is a schematic structural view of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which this application belongs; the terms used in the specification of the application are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned figure descriptions are intended to cover non-exclusive inclusions.

[0049] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0050] The technical solution of the present invention is clearly and completely described below in conjunction with the accompanying drawings and specific embodiments.

[0051] refer to Figure 1 , a flow chart of an embodiment of a method for detecting abnormality of a neutron detector signal in an ex-core nuclear measurement system is shown, wherein the method for detecting abnormality of a neutron detector signal in an ex-core nuclear measurement system comprises the following steps:

[0052] Step 101, obtaining a current signal of a neutron detector in real time through a monitoring device;

[0053] Step 102: Perform feature engineering on the current signal of the neutron detector to obtain feature data of the neutron detector;

[0054] Step 103, dividing the characteristic data of the neutron detector into data patterns and inputting the data patterns into the trained neutron detector twin model;

[0055] Step 104: the trained neutron detector twin model outputs the reconstructed feature data of the neutron detector and compares the reconstructed feature data of the neutron detector with the actual feature data;

[0056] If the difference between the reconstructed feature data and the actual feature data of the neutron detector exceeds the signal anomaly detection threshold of the neutron detector, the trained neutron detector twin model determines that the neutron detector signal is abnormal and issues an early warning.

[0057] In this embodiment, in step 103, the neutron detector twin model is formed by fusing the twin sub-models of each group of neutron detectors in the ex-core nuclear measurement system;

[0058] The twin model of each group of neutron detectors is built based on the twin relationship between the characteristic data of this group of neutron detectors and the characteristic data of other groups of neutron detectors, combined with the random forest regression model;

[0059] The twin model of each group of neutron detectors takes the characteristic data of other groups of neutron detectors as input, takes the reconstructed characteristic data of the neutron detectors of this group as output and compares the reconstructed characteristic data of the neutron detectors of this group with the actual characteristic data;

[0060] If the difference between the reconstructed characteristic data and the actual characteristic data of this group of neutron detectors exceeds the signal anomaly detection threshold of this group of neutron detectors, the twin model of this group of neutron detectors determines that the signal of this group of neutron detectors is abnormal and issues an early warning.

[0061] Since the groups of neutron detectors distributed at different positions of the extra-core nuclear measurement system collect neutrons at the same time in a common environment, the current signals of each group of neutron detectors are independent of each other but still maintain a close cooperative relationship. After the current signals of each group of neutron detectors are converted into characteristic data, the characteristic data of one group of neutron detectors can be deduced from the characteristic data of other groups of neutron detectors around it.

[0062] The state relationship between groups of neutron detectors can be defined by the following triples:

[0063]

[0064] In the formula, A i Represents the set of characteristic data of each group of neutron detectors in the out-of-core nuclear measurement system; S i Represents the value range set of each group of neutron detector characteristic data of the ex-core nuclear measurement system; U i Represents the twin relationship set of characteristic data of each group of neutron detectors in the ex-core nuclear measurement system; U i (A) represents the twin relationship between the characteristic data of each group of neutron detectors in the extra-core nuclear measurement system, describing that the characteristic data of one group of neutron detectors is jointly represented by the characteristic data of other groups of neutron detectors.

[0065] According to the above theoretical analysis, based on the twin relationship between the characteristic data of each group of neutron detectors in the extra-core nuclear measurement system, a twin sub-model of each group of neutron detectors is built in combination with the random forest regression model to characterize the common connection between the characteristic data of each group of neutron detectors.

[0066] The random forest regression model is used as the basic framework for building the twin sub-models of each group of neutron detectors. The random forest regression model introduces randomness into the decision tree, takes the decision tree as the basic unit, selects the resampling bootstrap method to randomly obtain k different sample subsets from the original sample, and uses random subspace partitioning to construct a decision tree for each sample subset. The prediction results of all decision trees are voted by majority or averaged to obtain the reconstruction results of the random forest regression model.

[0067] In this embodiment, in step 103, the trained neutron detector twin model is obtained by the following steps:

[0068] Step 201, obtaining historical current signals of each group of neutron detectors of the out-of-core nuclear measurement system through monitoring equipment;

[0069] Step 202: Perform feature engineering on the historical current signals of each group of neutron detectors to obtain feature data of each group of neutron detectors; divide the feature data of each group of neutron detectors into training data and test data, the training data of each group of neutron detectors constitutes a training data set, and the test data of each group of neutron detectors constitutes a test data set;

[0070] Step 203: preprocess the data in the training data set and the test data set respectively;

[0071] Step 204: training the neutron detector twin model using the preprocessed data in the training data set to obtain a trained neutron detector twin model; the trained neutron detector twin model implements neutron detector signal anomaly detection;

[0072] Step 205: Test the trained neutron detector twin model using the preprocessed data in the test data set. If the test passes, the training of the neutron detector twin model is terminated; if the test fails, return to step 204.

[0073] In this embodiment, in step 203, the data in the training data set and the test data set are preprocessed respectively, including the following steps:

[0074] The characteristic data of each group of neutron detectors in the training data set and the test data set are divided into data patterns of each group of neutron detectors; the data pattern of each group of neutron detectors includes the characteristic data of other groups of neutron detectors at the same time and the characteristic data of the neutron detectors in this group.

[0075] In this embodiment, in step 204, the neutron detector twin model is trained by using the preprocessed data in the training data set to obtain the trained neutron detector twin model, including the following steps:

[0076] Step 301: input the characteristic data of each group of neutron detectors in the training data set at the same time as other groups of neutron detectors into the twin model of the current group of neutron detectors, and the twin model of the current group of neutron detectors outputs the reconstructed characteristic data of the current group of neutron detectors;

[0077] Step 302: The twin model of the neutron detectors in the present group determines the signal anomaly detection threshold of the neutron detectors in the present group according to the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in the present group, and obtains the trained twin model of the neutron detectors in the present group;

[0078] Step 303: The trained twin models of each group of neutron detectors are fused to form a trained neutron detector twin model;

[0079] The characteristic data of each group of neutron detectors in the training data set are the characteristic data of each group of neutron detectors in a normal state.

[0080] In this embodiment, in step 205, the trained neutron detector twin model is tested by using the preprocessed data in the test data set, including the following steps:

[0081] Step 401, input the characteristic data of each group of neutron detectors in the test data set at the same time as other groups of neutron detectors into the trained twin model of the group of neutron detectors;

[0082] The trained twin model of the group of neutron detectors outputs the reconstructed feature data of the group of neutron detectors and compares the actual feature data of the group of neutron detectors with the reconstructed feature data;

[0083] If the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in this group exceeds the signal anomaly detection threshold of the neutron detectors in this group, the trained twin model of the neutron detectors in this group determines that the signal of the neutron detectors in this group is abnormal;

[0084] Step 402: Calculate the mean square error of the signal anomaly detection result of the trained twin model of the neutron detectors in the group and the actual state of the neutron detectors in the group to obtain the signal anomaly detection error of the trained twin model of the neutron detectors in the group;

[0085] If the error of the signal anomaly detection of the neutron detectors in this group by the trained twin model exceeds the set threshold, it is determined that the trained twin model of the neutron detectors in this group has failed the test;

[0086] If the error of the trained twin model for detecting abnormal signals of the neutron detectors in this group is less than the set threshold, the trained twin model of the neutron detectors in this group is judged to have passed the test;

[0087] Step 403: If all the trained twin models of each group of neutron detectors pass the test, the trained neutron detector twin model passes the test, and the training of the neutron detector twin model ends;

[0088] If the trained twin model of any group of neutron detectors fails the test, the process returns to step 204 to train the twin model of the current group of neutron detectors.

[0089] The characteristic data of each group of neutron detectors in the test data set are characteristic data of known states of each group of neutron detectors.

[0090] refer to Figure 2 As an implementation of the above method, the present invention provides an embodiment of a device for detecting anomaly of neutron detector signals in an ex-core nuclear measurement system. The device embodiment corresponds to the embodiment of the above method for detecting anomaly of neutron detector signals in an ex-core nuclear measurement system. The device can be specifically applied to various electronic devices.

[0091] The device for detecting abnormality of neutron detector signals in the out-of-core nuclear testing system of this embodiment includes:

[0092] A data acquisition module, connected to the monitoring device data, for real-time acquisition of the current signal of the neutron detector and sending it to the data processing module;

[0093] A data processing module is used to receive the current signal of the neutron detector sent by the data acquisition module and perform feature engineering to obtain the feature data of the neutron detector; divide the feature data of the neutron detector into data modes and send them to the judgment module;

[0094] A judgment module, with an embedded trained neutron detector twin model, is used to receive the data pattern of the neutron detector sent by the data processing module and input it into the trained neutron detector twin model; the trained neutron detector twin model outputs the reconstructed feature data of the neutron detector and compares the reconstructed feature data of the neutron detector with the actual feature data; if the difference between the reconstructed feature data and the actual feature data of the neutron detector exceeds the signal anomaly detection threshold of the neutron detector, the trained neutron detector twin model determines that the neutron detector signal is abnormal and issues an early warning.

[0095] See also Figure 3 As an implementation of the above method, the present invention provides an embodiment of a computer device, which corresponds to the embodiment of the above-mentioned method for detecting abnormality of neutron detector signals in an ex-core nuclear measurement system.

[0096] The computer device described in this embodiment includes a memory, a processor and a network interface that are interconnected and communicated through a system bus. It should be noted that the figure only shows a computer device with a network interface, a storage and a processor, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to a microprocessor, an application-specific integrated circuit, a programmable gate array, a digital processor, an embedded device, etc.

[0097] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0098] The memory includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory, random access memory, static random access memory, read-only memory, electrically erasable programmable read-only memory, programmable read-only memory, magnetic memory, disk, optical disk, etc. In some embodiments, the memory may be an internal storage unit of the computer device, such as a hard disk or memory of the computer device. In other embodiments, the memory may also be an external storage device of the computer device, such as a plug-in hard disk, smart memory card, secure digital card, flash memory card, etc. equipped on the computer device. Of course, the memory may also include both the internal storage unit of the computer device and its external storage device. In this embodiment, the memory is generally used to store the operating system and various application software installed on the computer device, such as the computer-readable instructions of the above-mentioned method for detecting abnormal neutron detector signals in the out-of-core nuclear measurement system. In addition, the memory may also be used to temporarily store various types of data that have been output or are to be output.

[0099] In some embodiments, the processor may be a central processing unit, a controller, a microcontroller, a microprocessor, or other data processing chip. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run the computer-readable instructions stored in the memory or process data, such as running the computer-readable instructions of the above-mentioned method for detecting abnormal sub-detector signals in the ex-core nuclear measurement system.

[0100] The network interface may include a wireless network interface or a wired network interface, which is generally used to establish a communication connection between the computer device and other electronic devices.

[0101] As an implementation of the above method, the present invention provides an embodiment of a computer-readable storage medium, and the embodiment of the computer-readable storage medium corresponds to the embodiment of the above-mentioned method for detecting abnormality of neutron detector signals in an ex-core nuclear measurement system.

[0102] The computer-readable storage medium described in this embodiment stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned method for detecting abnormality of sub-detector signals in the extra-core nuclear measurement system.

[0103] The method, device, computer equipment and storage medium for detecting abnormal signals of neutron detectors in an ex-core nuclear measurement system of this embodiment are based on the twin relationship between the characteristic data of neutron detectors in the ex-core nuclear measurement system, and use a random forest regression model with high regression prediction accuracy to build a twin model of neutron detectors, so as to achieve 100% accuracy in detecting abnormal signals of neutron detectors in the ex-core nuclear measurement system, and provide auxiliary judgment basis for status monitoring of relevant core equipment or components in nuclear power plants.

[0104] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for detecting abnormality of neutron detector signals in an out-of-core nuclear measurement system, characterized in that: The following steps are involved: Step 101, obtaining a current signal of a neutron detector in real time; Step 102: Perform feature engineering on the current signal of the neutron detector to obtain feature data of the neutron detector; Step 103, dividing the characteristic data of the neutron detector into data patterns and inputting the data patterns into the trained neutron detector twin model; Step 104: The trained neutron detector twin model outputs the reconstructed feature data of the neutron detector and performs neutron detector signal anomaly detection.

2. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 1, characterized in that: In step 103, the trained neutron detector twin model is obtained by the following steps: Step 201, obtaining historical current signals of each group of neutron detectors of the out-of-core nuclear measurement system through monitoring equipment; Step 202: Perform feature engineering on the historical current signals of each group of neutron detectors to obtain feature data of each group of neutron detectors; divide the feature data of each group of neutron detectors into training data and test data, the training data of each group of neutron detectors constitutes a training data set, and the test data of each group of neutron detectors constitutes a test data set; Step 203: preprocess the data in the training data set and the test data set respectively; Step 204: training the neutron detector twin model using the preprocessed data in the training data set to obtain a trained neutron detector twin model; the trained neutron detector twin model implements neutron detector signal anomaly detection; Step 205: Test the trained neutron detector twin model using the preprocessed data in the test data set. If the test passes, the training of the neutron detector twin model is terminated; if the test fails, return to step 204.

3. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 2, characterized in that: In step 103, the twin model of the neutron detector is formed by integrating the twin models of each group of neutron detectors in the extra-core nuclear measurement system; the twin model of each group of neutron detectors is built based on the twin relationship between the characteristic data of the neutron detectors in this group and the characteristic data of other groups of neutron detectors, combined with the random forest regression model.

4. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 3, characterized in that: The twin model of each group of neutron detectors takes the characteristic data of other groups of neutron detectors as input, takes the reconstructed characteristic data of the neutron detectors of this group as output and compares the reconstructed characteristic data of the neutron detectors of this group with the actual characteristic data; If the difference between the reconstructed characteristic data and the actual characteristic data of this group of neutron detectors exceeds the signal anomaly detection threshold of this group of neutron detectors, the twin model of this group of neutron detectors determines that the signal of this group of neutron detectors is abnormal and issues an early warning.

5. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 3, characterized in that: In step 203, the data in the training data set and the test data set are preprocessed respectively, including the following steps: The characteristic data of each group of neutron detectors in the training data set and the test data set are divided into data patterns of each group of neutron detectors; the data pattern of each group of neutron detectors includes the characteristic data of other groups of neutron detectors at the same time and the characteristic data of the neutron detectors in this group.

6. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 3, characterized in that: In step 204, the neutron detector twin model is trained by using the preprocessed data in the training data set to obtain the trained neutron detector twin model, including the following steps: Step 301: input the characteristic data of each group of neutron detectors in the training data set at the same time as other groups of neutron detectors into the twin model of the current group of neutron detectors, and the twin model of the current group of neutron detectors outputs the reconstructed characteristic data of the current group of neutron detectors; Step 302: The twin model of the neutron detectors in the present group determines the signal anomaly detection threshold of the neutron detectors in the present group according to the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in the present group, and obtains the trained twin model of the neutron detectors in the present group; Step 303: The trained twin models of each group of neutron detectors are fused to form a trained neutron detector twin model; The characteristic data of each group of neutron detectors in the training data set are the characteristic data of each group of neutron detectors in a normal state.

7. The method for detecting abnormality of neutron detector signals in an out-of-core nuclear testing system according to claim 3, characterized in that: In step 205, the trained neutron detector twin model is tested using the preprocessed data in the test data set, including the following steps: Step 401, input the characteristic data of each group of neutron detectors in the test data set at the same time as other groups of neutron detectors into the trained twin model of the group of neutron detectors; The trained twin model of the group of neutron detectors outputs the reconstructed feature data of the group of neutron detectors and compares the actual feature data of the group of neutron detectors with the reconstructed feature data; If the difference between the actual characteristic data and the reconstructed characteristic data of the neutron detectors in this group exceeds the signal anomaly detection threshold of the neutron detectors in this group, the trained twin model of the neutron detectors in this group determines that the signal of the neutron detectors in this group is abnormal; Step 402: Calculate the mean square error of the signal anomaly detection result of the trained twin model of the neutron detectors in the group and the actual state of the neutron detectors in the group to obtain the signal anomaly detection error of the trained twin model of the neutron detectors in the group; If the error of the signal anomaly detection of the neutron detectors in this group by the trained twin model exceeds the set threshold, it is determined that the trained twin model of the neutron detectors in this group has failed the test; If the error of the trained twin model for detecting abnormal signals of the neutron detectors in this group is less than the set threshold, the trained twin model of the neutron detectors in this group is judged to have passed the test; Step 403: If all the trained twin models of each group of neutron detectors pass the test, the trained neutron detector twin model passes the test, and the training of the neutron detector twin model ends; If the trained twin model of any group of neutron detectors fails the test, the process returns to step 204 to train the twin model of the current group of neutron detectors. The characteristic data of each group of neutron detectors in the test data set are characteristic data of known states of each group of neutron detectors.

8. A neutron detector signal anomaly detection device in an out-of-core nuclear measurement system, characterized in that: include: A data acquisition module, used to acquire the current signal of the neutron detector in real time and send it to a data processing module; A data processing module is used to receive the current signal of the neutron detector sent by the data acquisition module and perform feature engineering to obtain the feature data of the neutron detector; divide the feature data of the neutron detector into data modes and send them to the judgment module; The judgment module has an embedded trained neutron detector twin model, which is used to receive the data pattern of the neutron detector sent by the data processing module and input it into the trained neutron detector twin model; the trained neutron detector twin model outputs the characteristic data reconstructed by the neutron detector and performs neutron detector signal anomaly detection.

9. A computer device comprising a memory and a processor, wherein the memory stores computer-readable instructions, characterized in that: When the processor executes the computer-readable instructions, the steps of the method for detecting abnormality of sub-detector signals in an out-of-core nuclear measurement system as described in any one of claims 1-7 are implemented.

10. A computer-readable storage medium having computer-readable instructions stored thereon, characterized in that: When the computer-readable instructions are executed, the steps of the method for detecting abnormality of sub-detector signals in an out-of-core nuclear measurement system as described in any one of claims xxx-xxx are implemented.

Images