Abnormality detection system based on acceleration signals

Abstract

The application discloses an abnormality detection system based on acceleration signals, which is applied to a nuclear reactor. The system comprises: an acceleration sensor arranged in a pre-divided equipment space in the nuclear reactor, a digital processing module arranged at the equipment space, a centralized communication module arranged in the nuclear reactor, and a comprehensive diagnosis module arranged outside the nuclear reactor. The acceleration sensor is arranged at a nuclear reactor equipment in the equipment space, receives acceleration signals of the nuclear reactor equipment, and sends the acceleration signals to the digital processing module at the equipment space for abnormal feature identification. The abnormal features are sent to the centralized communication module, and then sent to the comprehensive diagnosis module outside the nuclear reactor for abnormal feature analysis, so as to obtain whether abnormal vibration signals, wear signals and impact signals exist in the nuclear reactor system. The system can timely find hidden faults in the nuclear reactor system, and improves the identification ability of the hidden faults in the nuclear reactor system.

Description

Technical Field

[0001] This invention relates to the field of nuclear reactor fault detection technology, and in particular to an anomaly detection system based on acceleration signals. Background Technology

[0002] With the increasing demand for electricity across society and the growing awareness of energy conservation, environmental protection, and emission reduction, countries worldwide are accelerating the development of nuclear power, which has become a crucial component of power generation systems. The nuclear reactor system within a nuclear power plant comprises numerous active devices, such as pumps, control rod drive mechanisms, and valves. Driven by motors, these devices undergo continuous rotation and reciprocating motion, inevitably leading to vibration and wear. Over long periods of operation, this can cause structural failures in critical components, resulting in equipment malfunctions and posing serious safety hazards to the nuclear reactor system.

[0003] Currently, both domestic and international nuclear reactor systems utilize vibration, wear, and impact-based anomaly detection systems to detect abnormalities in active equipment and identify potential faults. However, these systems all employ analog signal transmission. Analog signal transmission is costly due to the numerous long-distance signal cables and the need for a large number of electrical penetrations, resulting in complex structures and difficult installation and maintenance. Furthermore, the large number and dense arrangement of equipment and cables within a nuclear reactor system create a complex environment with numerous sources of interference. Analog signal transmission is also highly susceptible to environmental interference, leading to monitoring failures and a weak ability to identify potential faults in the nuclear reactor system. Summary of the Invention

[0004] In view of this, this application provides an anomaly detection system based on acceleration signals, the main purpose of which is to solve the technical problem that the existing nuclear reactor anomaly detection systems have a weak ability to identify potential faults in nuclear reactor systems.

[0005] According to a first aspect of the present invention, an anomaly detection system based on acceleration signals is provided for use in a nuclear reactor, the system comprising:

[0006] An accelerometer is installed at the nuclear reactor equipment within a pre-defined equipment space inside the nuclear reactor to collect acceleration signals from the nuclear reactor equipment.

[0007] A digital processing module is disposed in the equipment space and connected to the acceleration sensor. The digital processing module is used to receive the acceleration signal and generate abnormal feature data of the nuclear reactor equipment in the form of digital signals based on the acceleration signal.

[0008] A centralized communication module is installed inside the nuclear reactor. The centralized communication module is connected to the digital processing module via active Ethernet and is used to receive the abnormal feature data.

[0009] A comprehensive diagnostic module is located outside the nuclear reactor. The comprehensive diagnostic module is connected to the centralized communication module via active Ethernet to obtain the abnormal feature data from the digital processing module through the centralized communication module, and to detect the equipment status information of the nuclear reactor equipment based on the abnormal feature data. The equipment status information includes abnormal equipment status and normal equipment status.

[0010] This invention provides an anomaly detection system based on acceleration signals. The system pre-divides the space within a nuclear reactor into multiple equipment spaces, such as pressure vessel spaces and steam generator spaces. Accelerometers are installed at the nuclear reactor equipment within each space to collect acceleration signals. Digital processing modules are strategically placed within each equipment space according to the type of acceleration signal to process the signals and obtain anomaly characteristic data. This data is then transmitted digitally to a centralized communication module within the reactor, avoiding interference from complex environments during long-distance transmission of analog signals. Furthermore, the centralized communication module transmits the anomaly characteristic data generated by multiple digital processing modules to a comprehensive diagnostic module outside the reactor via a single cable. This architecture significantly reduces the number of through-cables and minimizes interference during transmission, improving signal transmission quality and preventing monitoring failures. It enables timely detection of potential faults in the nuclear reactor system and enhances the ability to identify such faults.

[0011] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0013] Figure 1 This illustration shows one of the structural schematic diagrams of an anomaly detection system based on acceleration signals provided in an embodiment of the present invention;

[0014] Figure 2 The second schematic diagram shows the structure of an anomaly detection system based on acceleration signals provided in an embodiment of the present invention;

[0015] Figure 3 The third schematic diagram shows the structure of an anomaly detection system based on acceleration signals provided in an embodiment of the present invention. Detailed Implementation

[0016] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0017] Currently, both domestic and international nuclear reactor systems utilize vibration, wear, and impact-based anomaly detection systems to detect abnormalities in active equipment and identify potential faults. However, these systems all employ analog signal transmission. Analog signal transmission is costly due to the numerous long-distance signal cables and the need for a large number of electrical penetrations, resulting in complex structures and difficult installation and maintenance. Furthermore, the large number and dense arrangement of equipment and cables within a nuclear reactor system create a complex environment with numerous sources of interference. Analog signal transmission is also highly susceptible to environmental interference, leading to monitoring failures and a weak ability to identify potential faults in the nuclear reactor system.

[0018] To address the above problems, in one embodiment, such as Figure 1 As shown, an anomaly detection system based on acceleration signals is provided. Taking the application of this system in a nuclear reactor as an example, it should be noted that this embodiment uses a nuclear reactor as a scenario for illustration, but other scenarios are also applicable to this embodiment. The anomaly detection system based on acceleration signals includes an acceleration sensor 110, a digital processing module 120, a centralized communication module 130, and a comprehensive diagnostic module 140.

[0019] Specifically, there can be multiple accelerometers 110, which are installed at the nuclear reactor equipment within pre-defined equipment spaces within the nuclear reactor to collect acceleration signals from the nuclear reactor equipment. These acceleration signals can include vibration identification signals, wear identification signals, and impact identification signals. The equipment spaces can be different pre-defined monitoring zones within the nuclear reactor, such as pressure vessel zones, steam generator zones, main pump zones, and control rod drive mechanism zones. The nuclear reactor equipment can be equipment such as steam generators, main pumps, and control rod drive mechanisms installed within these equipment spaces. Furthermore, the accelerometers 110 are broadband sensors. These sensors convert the vibrations, wear, and impacts generated by the nuclear reactor equipment into charge signals as acceleration signals. The technical specifications of the accelerometers 110 are as follows: measurement range: 500g, charge conversion coefficient: 10pC / g, frequency response: 1Hz-10kHz, operating temperature range: -40-400℃, maximum ambient humidity: 100%, γ integral dose: 100Mrad.

[0020] Furthermore, there can be multiple digital processing modules 120, each located within the equipment space. Each digital processing module 120 is connected to the accelerometer 110 and receives the acceleration signal. Based on the acceleration signal, it generates abnormal characteristic data of the nuclear reactor equipment in digital signal form. The abnormal characteristic data includes impact characteristic data and wear characteristic data. The impact characteristic data is used in subsequent processing to determine whether the nuclear reactor equipment with the accelerometer 110 has experienced an abnormal impact, thus identifying potential faults. The wear characteristic data is used in subsequent processing to determine whether the nuclear reactor equipment with the accelerometer 110 has experienced abnormal wear, thus identifying potential faults. Here, the digital processing module 120 can be installed near the accelerometer 110 it is connected to, generating abnormal characteristic data in digital signal form based on the acceleration signal, and then transmitting the abnormal characteristic data over a long distance, avoiding interference from complex environments during long-distance transmission of analog signals.

[0021] Furthermore, the digital processing module 120 adopts an industrial design, featuring U-shaped grooves for ample heat dissipation. The module has a cubic shape, with four M12 waterproof two-pin aviation connectors on the left side for sensor connection, and a waterproof switch and an M12 aviation network port on the right. The module's outer shell is made of aluminum alloy with multiple U-shaped grooves machined on the surface, significantly reducing weight. The material's inherent properties also provide excellent thermal conductivity, preventing overheating even during prolonged use. The black anodized surface enhances its oxidation resistance. Connection holes facilitate connection to cabinets and other equipment. The back cover of the module is also made of high-quality aluminum alloy, machined with a black anodized finish, ensuring color uniformity and flatness. Screws are stainless steel countersunk screws, embedded within the cover for easy installation. The internal processing circuit board of the digital processing module 120 is connected to the housing using high-quality brass and a thermal pad. Its excellent thermal conductivity allows for rapid heat dissipation, ensuring stable operation. Furthermore, to meet IP65 requirements, the module's structure and component selection utilize all waterproof and dustproof electrical components. A waterproof silicone gasket is used at the connection between the cover and the housing, providing excellent sealing and effectively preventing the intrusion of dust and water. The digital processing module 120 has the following dimensions: 230mm high × 100mm wide × 40mm deep. Each digital processing module 120 can connect to four accelerometer sensors 110, including analog and digital processing sections, providing vibration, wear, and impact signal feature analysis and transmitting data to the upper layer.

[0022] Furthermore, a centralized communication module 130 is installed within the nuclear reactor. This module is connected to the digital processing module 120 via active Ethernet and is used to receive the abnormal characteristic data in digital signal form. Specifically, the centralized communication module 130 transmits monitoring data from the digital processing module 120 to the integrated diagnostic module 140 outside the nuclear reactor, and simultaneously sends instructions from the integrated diagnostic module 140 to the digital processing module 120. This centralized communication module 130 is wall-mounted and has structural dimensions of 189mm (height) × 325mm (width) × 92mm (depth). It operates over a wide temperature range of -40℃ to 70℃, has an IP65 high protection rating to meet dustproof and waterproof requirements, and uses an M12 interface to ensure the sealing and robustness of the port connection. It is suitable for applications with strong vibration and impact, and overall meets the usage requirements of the nuclear reactor system. The centralized communication module 130 adopts active Ethernet technology, has a maximum POE power output of 240W, 16 100Mbps Fast Ethernet interfaces supporting POE, can drive 16 signal digitization processing modules, and has good scalability.

[0023] Furthermore, the integrated diagnostic module 140 is located outside the nuclear reactor. The integrated diagnostic module 140 is connected to the centralized communication module 130 via active Ethernet to acquire the abnormal feature data from the digital processing module 120 through the centralized communication module 130, and to detect the equipment status information of the nuclear reactor equipment based on the abnormal feature data. The equipment status information includes abnormal equipment status and normal equipment status. Specifically, the integrated diagnostic module 140 can combine data from all vibration, wear, and impact signals on-site to perform vibration signal anomaly identification and diagnosis, wear signal anomaly identification and diagnosis, and impact signal identification and fault area location. Furthermore, the integrated diagnostic module 140 can be connected to the centralized communication module 130 inside the nuclear reactor via a single Ethernet cable. The structural dimensions of the integrated diagnostic module 140 can be 255mm high × 237mm wide × 177mm deep.

[0024] This embodiment provides an anomaly detection system based on acceleration signals. Employing an active Ethernet architecture, the system pre-divides the space within the nuclear reactor into multiple equipment spaces, such as pressure vessel spaces and steam generator spaces. Accelerometers are placed at the nuclear reactor equipment within each space to collect acceleration signals. Digital processing modules are strategically placed within each equipment space according to the type of acceleration signal to process the signals and obtain anomaly characteristic data. This data is then transmitted digitally to a centralized communication module within the reactor, avoiding interference from complex environments during long-distance transmission of analog signals. Furthermore, the centralized communication module transmits the anomaly characteristic data generated by multiple digital processing modules to a comprehensive diagnostic module outside the reactor via a single cable. This architecture significantly reduces the number of through-cables and minimizes interference during transmission, improving signal transmission quality and preventing monitoring failures. It enables timely detection of potential faults in the nuclear reactor system and enhances the ability to identify such faults.

[0025] In one embodiment, the digitization module is further configured to determine the impact identification signal waveform of the nuclear reactor device based on the impact identification signal, and store the impact identification signal waveform locally. Specifically, the digitization module can generate the impact identification signal waveform of the device where the accelerometer is located based on the impact identification signal received from the accelerometer. In the embodiments provided in this application, the digitization module can read the impact identification signal waveform and determine which channel of the impact identification signal waveform corresponds to the impact event generated by the device, complete the channel-level feature analysis of the impact identification signal, and store the data for subsequent impact fault analysis.

[0026] In one embodiment, such as Figure 2As shown, the anomaly detection system based on acceleration signals also includes a power supply module 150, which is connected to the centralized communication module 130. The power supply module 150 obtains external power and supplies power to the centralized communication module 130, enabling the centralized communication module 130 to supply power to the digital processing module 120 via active Ethernet. The power supply module 150 can be installed outside the nuclear reactor and connected to the centralized communication module 130 via a single wire. Its structural dimensions are 380mm high × 155mm wide × 210mm deep. It converts 220VAC into DC power to provide operating power for the system, allowing the centralized communication module 130 to achieve data communication and power supply for all digital processing modules 120 via active Ethernet. The embodiments provided in this application can supply power to the centralized communication module and digital processing module in the reactor via a single wire, significantly reducing the number of through-wire cables, reducing electrical connection requirements, lowering deployment costs, reducing structural complexity, and simplifying installation and maintenance.

[0027] In one embodiment, such as Figure 3 As shown, the digital processing module 120 includes an accelerometer adapter circuit 121, a signal amplitude adjustment circuit 122, a signal frequency adjustment circuit 123, an analog-to-digital conversion circuit 124, and a digital processing unit 125.

[0028] The accelerometer adapter circuit 121 is connected to the accelerometer 110 and is used to receive the acceleration signal and perform impedance conversion processing on the acceleration signal to obtain the impedance-converted acceleration signal. Specifically, the accelerometer adapter circuit 121 converts the high output impedance of the accelerometer 110 into a low output impedance and converts the impact charge signal detected by the accelerometer 110 into an impact voltage signal usable by the back-end circuit. The conversion ratio can be 10 pC / g.

[0029] Furthermore, the signal amplitude adjustment circuit 122 is connected to the accelerometer adapter circuit 121 to receive the acceleration signal after impedance conversion processing and amplify the acceleration signal to obtain the amplified acceleration signal. Specifically, the signal amplitude adjustment circuit 122 needs to monitor an impact acceleration of 100g, which can be achieved by adjusting the amplitude of the impact recognition signal. Since its voltage monitoring range is 10V, the charge conversion coefficient of the accelerometer 110 is 10pC / g, and the charge-voltage conversion coefficient of the accelerometer adapter circuit 121 is 5mV / pC, the signal amplitude adjustment circuit 122 can amplify the impact recognition signal by 2 times.

[0030] Furthermore, the signal frequency adjustment circuit 123 is connected to the signal amplitude adjustment circuit 122, and is used to receive the amplified acceleration signal and filter out the interference signals in the amplified acceleration signal to obtain a simulated acceleration signal. Specifically, the signal frequency adjustment circuit 123 performs frequency adjustment according to the characteristics of vibration, wear, and impact identification signals. The effective frequency range of the vibration identification signal is 5Hz-1kHz, and the effective frequency range of the wear and impact identification signals is 1kHz-10kHz. At the same time, it filters out interference frequencies outside the effective frequency range, retains the effective frequency of the vibration, wear, and impact identification signals, and filters out interference frequencies outside the effective frequency range.

[0031] Furthermore, the analog-to-digital conversion circuit 124 is connected to the signal frequency adjustment circuit 123 to receive analog acceleration signals and perform analog-to-digital conversion on the analog acceleration signals to obtain digital acceleration signals. Specifically, the analog-to-digital conversion circuit 124 converts vibration, wear, and impact analog signals into digital signals for the subsequent digital processing of vibration, wear, and impact identification signals. The analog-to-digital conversion accuracy is 16 bits, and the analog-to-digital conversion rate is not less than 50 kHz. Furthermore, a digital processing module 120 may include multiple accelerometer adapter circuits 121, signal amplitude adjustment circuits 122, signal frequency adjustment circuits 123, and analog-to-digital conversion circuits 124 as signal processing channels to connect to different accelerometers 110.

[0032] Furthermore, the digitization processing unit 125 is connected to the analog-to-digital conversion circuit 124, and is used to receive digital acceleration signals and generate abnormal characteristic data of the nuclear reactor equipment based on the digital acceleration signals. For example, Figure 3 As shown, the digitization processing unit 125 may include a field-programmable gate array (FPGA), an ARM chip A, and a data storage chip S. The FPGA controls the signal digitization processing module, the ARM chip A performs feature analysis of vibration, wear, and impact identification signals and Ethernet communication, and the data storage chip S performs data storage in the digitization processing unit 125. Furthermore, the digitization processing module 120 also includes a PoE power supply circuit P and an active Ethernet interface C. The active Ethernet interface C is used for data interaction with the centralized communication module 130. The PoE power supply circuit converts the DC power provided by the active Ethernet into various DC currents required for the operation of the signal digitization processing module. The embodiments provided in this application can process acceleration signals collected by accelerometers based on the digitization processing module and generate abnormal feature data in digital signal form, which is then sent to the centralized communication module. This reduces signal interference during data transmission and improves the ability to identify potential faults in nuclear reactors.

[0033] In one embodiment, the abnormal feature data includes impact feature data; impact identification signal feature analysis can be performed by the digital processing unit in the digital processing module. Each digital processing unit processing impact identification signals can process four channels simultaneously, and more than four channels can be processed by multiple digital processing units working together. The implementation method of the digital processing module generating abnormal feature data of the nuclear reactor equipment based on the acceleration signal can be as follows: acquiring a long effective signal strength value and determining the product value of the long effective signal strength value and a preset peak factor; wherein the long effective signal strength value is the background noise signal value at the location of the nuclear reactor equipment within a preset long effective time period.

[0034] Specifically, the values ​​of the long effective time period Tl and the product value K can be predetermined. For example, the long effective time period Tl can be 1 to 10 seconds, the product value K can be 5, and the long effective signal strength value can be the effective value of the background noise signal strength. Further, using a preset short effective time period as the acquisition cycle, the short effective signal strength value of the acceleration signal is continuously acquired within each short effective time period, and the short effective signal strength values ​​are combined into short effective signal groups according to the acquisition time from front to back. The short effective time period Ts can be a preset time period, such as 1ms to 10ms. Specifically, the effective value of the acceleration signal acquired by the accelerometer within each short effective time period can be continuously acquired, and the short effective signal strength values ​​acquired in each short effective time period are arranged according to time order and added to the short effective signal group. Further, a count value is set and set to zero, where the count value Na can be determined according to the actual situation. Furthermore, a target signal strength value is obtained from the short effective signal group; wherein, the target signal strength value is the short effective signal strength value collected within the short effective time period corresponding to the current time point in the short effective signal group. Specifically, a short effective time period corresponding to the current time point can be determined, and the short effective signal strength value collected within this short effective time period can be determined as the target signal strength value.

[0035] Further, the first loop process is executed until the first preset condition is met. First, the target signal strength value is compared with the product value. Specifically, the product value between the peak factor and the long effective signal strength value can be determined, and the relationship between the target signal strength value and the product value can be compared. Then, when the target signal strength value is less than the product value, the count value is set to zero, the target signal strength value is determined as the previous short effective signal strength value, and the short effective signal strength value arranged one position after the previous short effective signal strength value in the short effective signal group is determined as the target signal strength value, and the step of comparing the target signal strength value with the product value is executed. Specifically, when the target signal strength value is less than the product value, the count value is set to zero, and the target signal strength value is determined as the previous short effective signal strength value. The short effective signal strength value that is one position after the previous short effective signal strength value in the short effective signal group is determined as the target signal strength value. For example, if the previous short effective signal strength value is the value corresponding to the first short effective time period in the short effective signal group, then the value corresponding to the second short effective time period in the short effective signal group is determined as the target signal strength value. Further, the step of comparing the target signal strength value with the product value is performed repeatedly in a loop.

[0036] Conversely, when the target signal strength value is greater than or equal to the product value, the count value is incremented by one, the target signal strength value is determined as the previous short effective signal strength value, the short effective signal strength value in the short effective signal group that is one position after the previous short effective signal strength value is determined as the target signal strength value, and the step of comparing the target signal strength value with the product value is performed repeatedly.

[0037] Furthermore, the first preset condition is that the count value is equal to a preset upper limit of the count value. The upper limit of the count value can be obtained by dividing the duration value Td by the short effective time period value Ts. Furthermore, both the duration value Td and the short effective time period value Ts are preset, and the duration of each impact in a specific reactor area can be determined in advance through experiments or other methods, and Td is determined based on the duration of each impact.

[0038] Furthermore, the impact feature data is generated and sent to the integrated diagnostic module. The impact feature data may include information about the occurrence of a count value reaching its upper limit within a unit of time. This information is sent to the integrated diagnostic module to enable it to perform system-level integrated diagnostics of the impact identification signal. Further, the count value is set to zero, and after a preset interval, the step of obtaining the target signal strength value in the short effective signal group is executed. The interval value Te can be preset, and can be determined in advance through experiments or other methods to determine the interval time of impacts occurring in a specific reactor area, and the interval time Te is determined based on the impact interval time.

[0039] Furthermore, if the digitization module has multiple channels, while performing channel-level feature analysis of the impact signal, the data from each channel is stored cyclically in the "LP_Block1" data block of the data storage chip. The data length of each channel depends on the sampling frequency and pre-sampling time. Once the channel-level feature analysis of the impact signal is complete, the data is stored in the "LP_Block2" data block of the data storage chip, and the data processing module is notified to read the current impact waveform, continuing the cyclic operation according to the above logic. Simultaneously, the design of the acceleration signal-based anomaly detection system should at least ensure that, within the specified interval, when the channel-level feature analysis of an impact signal occurs in one or all channels, the digitization unit can read the impact waveforms of all channels. The embodiments provided in this application can determine whether there is a potential impact-type fault in the nuclear reactor equipment based on the number of times the effective value of the acceleration signal is greater than the product of the effective value of the background noise and the peak factor, thus improving the ability to identify potential faults in the nuclear reactor.

[0040] In one embodiment, the digital processing module determines the impact identification signal waveform of the nuclear reactor device based on the impact identification signal and stores the impact identification signal waveform locally. This can be achieved by: first, monitoring the short effective signal strength value and determining whether the short effective signal strength value is greater than or equal to the product value; then, if the short effective signal strength value is greater than or equal to the product value, determining a target time point before the current time point at a preset first time length. Here, the time point where the short effective signal strength value is greater than or equal to the product value may not be the starting point of the impact identification signal waveform, requiring time backtracking. For example, if the current time point is Tn and the first time length is Tx, then the target time point is the time point Tx before Tn. Finally, starting from the target time point, determining the time period after a preset second time length as the target time period, determining the impact identification signal waveform based on the impact identification signal within the target time period, and storing the impact identification signal waveform locally. As an example, if the second time length is Ty and the target time point is Tn-Tx, then the target time period is [Tn-Tx to Ty], and the impact identification signal waveform is determined based on the impact identification signal within [Tn-Tx to Ty]. The embodiments provided in this application can acquire the complete impact identification signal waveform and save it for subsequent impact fault analysis.

[0041] In one embodiment, the digital processing module generates impact feature data based on the impact identification signal collected by the accelerometer, and sets a first channel identifier corresponding to the accelerometer. The abnormal equipment state includes an abnormal impact state. Further, the integrated diagnostic module can perform system-level integrated diagnostics of the impact identification signal. The method for detecting the equipment status information of the nuclear reactor equipment based on the abnormal feature data can be as follows: First, upon receiving the impact feature data, the impact feature data is identified as target impact feature data. Further, based on the first channel identifier of the target impact feature data, the accelerometer corresponding to the target impact feature data is identified, and the equipment space where the accelerometer is located is identified as the first target equipment space. Specifically, the integrated diagnostic module can monitor in real time whether impact feature data is obtained from the centralized communication module. When the impact feature data is received, the accelerometer corresponding to the first channel identifier can be identified based on the first channel identifier of the impact feature data, and the equipment space where the accelerometer is located can be identified as the first target equipment space. As an example, if the accelerometer corresponding to the target impact feature data is installed in the pressure vessel equipment space, the pressure vessel equipment space can be identified as the first target equipment space.

[0042] Furthermore, all the acceleration sensors located within the first target device space are identified as first target acceleration sensors, and the first channel identifier in each impact feature data received within a first preset time period after the target impact feature data is received is determined. The length of the first preset time period can be determined based on actual conditions. Specifically, all acceleration sensors within the first target device space can be identified as first target acceleration sensors, and impact feature data acquired from the centralized communication module continues to be monitored within the first preset time period. Each acquired impact feature data is analyzed and processed to extract the first channel identifier located at the impact feature data location. Here, the first channel identifiers at all impact feature data locations acquired within the first preset time period can be grouped into a first identifier queue for convenient subsequent processing. Then, the number of first sensors among all the first target acceleration sensors that correspond to the first channel identifier is determined, and it is judged whether the number of first sensors is greater than or equal to a preset first quantity threshold. The value of the first quantity threshold can be determined based on actual conditions. Specifically, based on the first identifier queue, the number of first target acceleration sensors that correspond to the first channel identifiers contained in the first identifier queue can be determined as the number of first sensors.

[0043] Furthermore, if the number of the first sensors is greater than or equal to the first quantity threshold, it is determined whether there is any equipment operation in the first target device space. This equipment operation can be a system action in the target device space that might interfere with the initial analysis of the impact recognition signal, such as valve operation, pump start / stop, or control rod drive mechanism operation. For example, if the first quantity threshold is 2 and the number of the first sensors is 2, it is determined whether there is any equipment operation in the first target device space. Finally, if there is no equipment operation in the first target device space, the device status information is determined to be the abnormal impact state. The embodiments provided in this application can determine whether any other channel in the device space generating impact feature data has reached a specified number of initial analysis passes of the impact recognition signal within a specified time, thereby generating impact feature data. When there is another channel in the device space generating impact feature data that has reached a specified number of initial analysis passes of the impact recognition signal within a specified time, an impact event alarm can be provided, along with alarm device space information.

[0044] In one embodiment, the abnormal feature data further includes wear feature data; the acceleration signal further includes a wear identification signal; the digital processing module can generate wear feature data of the nuclear reactor equipment based on the acceleration signal in the following manner: First, the wear identification signal is received and filtered to obtain the original wear identification signal waveform. Specifically, the wear identification signal is detected, and frequency components below 1kHz containing rotational vibration, shell vibration, and support structure vibration are filtered out. The high-speed sampled acceleration signal is bandpass filtered to retain frequency components in the frequency range of 1kHz to 10kHz to obtain the original wear identification signal waveform. Then, the effective value of the original wear identification signal waveform is extracted to obtain acceleration level data. Specifically, the effective value of the original wear identification signal waveform can be extracted at certain time intervals to obtain a data set of acceleration signal waveforms to form a vibration acceleration sampling sequence, and the acceleration level La of the wear feature is obtained as acceleration level data based on Formula 1:

[0045]

[0046] Where a0 is the reference value for the vibration acceleration level, which can be preset to a0 = 1 μm / s², and a1 to a n A vibration acceleration sampling sequence can be used. Further, the acceleration level data is compared with a preset acceleration level threshold. If the acceleration level data is greater than or equal to the acceleration level threshold, the wear characteristic data is generated for subsequent wear fault analysis. The magnitude of the acceleration level threshold can be determined based on actual conditions. The embodiments provided in this application can determine whether there are potential wear-related faults in nuclear reactor equipment based on the acceleration level of the acceleration signal, improving the ability to identify potential faults in nuclear reactors.

[0047] In one embodiment, the digital processing module generates wear feature data based on the wear identification signal collected by the accelerometer, and sets a second channel identifier corresponding to the accelerometer. The abnormal equipment state also includes abnormal wear state. Further, the integrated diagnostic module can perform the detection of equipment status information of the nuclear reactor equipment based on the abnormal feature data in the following way: First, upon receiving the wear feature data, the wear feature data is identified as target wear feature data. Specifically, the integrated diagnostic module can monitor in real time whether wear feature data is obtained from the centralized communication module. When wear feature data is obtained, it is identified as target wear feature data. Further, based on the second channel identifier of the target wear feature data, the accelerometer corresponding to the target wear feature data is identified, and the equipment space where the accelerometer is located is identified as the second target equipment space. Specifically, when the wear feature data is received, the accelerometer corresponding to the second channel identifier can be identified based on the second channel identifier of the wear feature data, and the device space where the accelerometer is located can be determined as the second target device space. For example, if the device space where the accelerometer is located is the evaporator space, then the evaporator space can be determined as the second target device space.

[0048] Furthermore, all the acceleration sensors located within the second target device space are identified as second target acceleration sensors, and the second channel identifier is determined for each wear feature data received within a second preset time period after the target wear feature data is received. The length of the second preset time period can be determined based on actual conditions. Specifically, all acceleration sensors within the second target device space can be identified as second target acceleration sensors, and wear feature data acquired from the centralized communication module continues to be monitored within the second preset time period. Each piece of wear feature data acquired within the second preset time period is analyzed and processed to extract the second channel identifier located at the wear feature data location. Here, the second channel identifiers at all wear feature data locations acquired within the second preset time period can be grouped into a second identifier queue for convenient subsequent processing. Further, the number of second sensors corresponding to the second channel identifier among all the second target acceleration sensors is determined, and it is judged whether the number of second sensors is greater than or equal to a preset second quantity threshold. The value of the second quantity threshold can be determined based on actual conditions. Specifically, based on the second identifier queue, the number of second target acceleration sensors that correspond to the second channel identifiers contained in the second identifier queue is determined as the number of second sensors.

[0049] Furthermore, if the number of the second sensors is greater than or equal to the second quantity threshold, it is determined whether there is any device operation in the second target device space. For example, if the second quantity threshold is 3 and the number of the second sensors is 3, it is determined whether there is any device operation in the second target device space. Finally, if there is no device operation in the second target device space, the device status information is determined to be the abnormal wear state. The embodiments provided in this application can determine whether any other channel in the device space that generates wear feature data has reached a specified number of initial analysis passes of the wear identification signal within a specified time, thereby generating wear feature data. When there is another channel in the device space that generates wear feature data having reached a specified number of initial analysis passes of the wear identification signal within a specified time, a wear event alarm can be provided, along with information about the device space that triggered the alarm.

[0050] In one embodiment, the integrated diagnostic module is further configured to: issue an alarm notification when the equipment status information indicates an abnormal wear state or an abnormal impact state. Specifically, when the equipment status information indicates an abnormal wear state or an abnormal impact state, the alarm information containing the equipment status information, along with the identification information of the equipment space that triggered the alarm and information about the equipment involved, can be sent to the terminal of an alarm center or other department for early warning. The embodiments provided in this application can provide timely early warning of potential faults, enabling relevant personnel to take further action.

[0051] The acceleration signal-based anomaly detection system provided in this embodiment employs a distributed architecture based on active Ethernet. It digitizes vibration, wear, and impact signals collected by acceleration sensors via a local digitization module. This system offers advantages such as small size, fewer cables, and strong anti-interference capabilities, making it suitable for the confined space of nuclear reactors and addressing application challenges such as limited cable quantity and susceptibility to interference from complex environmental signals. Furthermore, the acceleration signal-based anomaly detection system integrates vibration, wear, and impact signal characteristic analysis with system-level comprehensive diagnostics, effectively improving the accuracy and reliability of vibration, wear, and impact monitoring in nuclear reactor systems. It can promptly detect abnormal vibration, wear, and impact signals in nuclear reactor systems, preventing equipment failures due to malfunctions and providing an effective guarantee for the safe and reliable operation of nuclear reactor systems.

[0052] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.

[0053] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.

Claims

1. An anomaly detection system based on acceleration signals, applied in a nuclear reactor, characterized in that, The system includes: An accelerometer is installed at the nuclear reactor equipment within a pre-defined equipment space inside the nuclear reactor to collect acceleration signals from the nuclear reactor equipment. A digital processing module is disposed in the equipment space and connected to the acceleration sensor. The digital processing module is used to receive the acceleration signal and generate abnormal feature data of the nuclear reactor equipment in the form of digital signals based on the acceleration signal. A centralized communication module is installed inside the nuclear reactor. The centralized communication module is connected to the digital processing module via active Ethernet and is used to receive the abnormal feature data. A comprehensive diagnostic module is located outside the nuclear reactor. The comprehensive diagnostic module is connected to the centralized communication module via active Ethernet to obtain the abnormal feature data from the digital processing module through the centralized communication module, and to detect the equipment status information of the nuclear reactor equipment based on the abnormal feature data. The equipment status information includes abnormal equipment status and normal equipment status. The abnormal feature data includes impact feature data; the generation of abnormal feature data for the nuclear reactor equipment based on the acceleration signal includes: The long effective signal strength value is obtained, and the product value of the long effective signal strength value and the preset peak factor is determined, wherein the long effective signal strength value is the background noise signal value of the location of the nuclear reactor equipment within the preset long effective time period; Using a preset short effective time period as the acquisition cycle, the short effective signal intensity value of the acceleration signal is continuously acquired within each short effective time period, and the short effective signal intensity values ​​are combined into short effective signal groups according to the acquisition time from front to back. Set a count value, and set the count value to zero; Obtain the target signal strength value in the short effective signal group, wherein the target signal strength value is the short effective signal strength value collected in the short effective time period corresponding to the current time point in the short effective signal group; Execute the first loop process until the first preset condition is met; Compare the target signal strength value with the product value; When the target signal strength value is less than the product value, the count value is set to zero, the target signal strength value is determined as the previous short effective signal strength value, and the short effective signal strength value arranged one position after the previous short effective signal strength value in the short effective signal group is determined as the target signal strength value, and the step of comparing the target signal strength value with the product value is performed. When the target signal strength value is greater than or equal to the product value, the count value is incremented by one, the target signal strength value is determined as the previous short effective signal strength value, the short effective signal strength value in the short effective signal group that is one position after the previous short effective signal strength value is determined as the target signal strength value, and the step of comparing the target signal strength value with the product value is performed. Wherein, the first preset condition is that the count value is equal to a preset upper limit of the count value; The impact feature data is generated and sent to the integrated diagnostic module; The count value is set to zero, and after a preset interval, the step of obtaining the target signal strength value in the short effective signal group is executed.

2. The system according to claim 1, characterized in that, The acceleration signal includes an impact identification signal; the digitization module is further used for: Based on the impact identification signal, the impact identification signal waveform of the nuclear reactor equipment is determined and stored locally.

3. The system according to claim 2, characterized in that, The system also includes: A power module, connected to the centralized communication module, is used to obtain external power and supply power to the centralized communication module, so that the centralized communication module can supply power to the digital processing module based on active Ethernet.

4. The system according to claim 1, characterized in that, The digitization processing module includes: An accelerometer adapter circuit is connected to the accelerometer and is used to receive the acceleration signal and perform impedance conversion processing on the acceleration signal to obtain an impedance-converted acceleration signal. A signal amplitude adjustment circuit is connected to the acceleration sensor adapter circuit. It is used to receive the acceleration signal after impedance conversion and to amplify the acceleration signal after impedance conversion to obtain an amplified acceleration signal. A signal frequency adjustment circuit, which is connected to the signal amplitude adjustment circuit, is used to receive the amplified acceleration signal and filter out interference signals in the amplified acceleration signal to obtain a simulated acceleration signal. An analog-to-digital converter circuit, connected to the signal frequency adjustment circuit, is used to receive an analog acceleration signal and perform analog-to-digital conversion on the analog acceleration signal to obtain a digital acceleration signal. A digital processing unit, connected to the analog-to-digital converter circuit, is used to receive digital acceleration signals and generate abnormal characteristic data of the nuclear reactor equipment based on the digital acceleration signals.

5. The system according to claim 2, characterized in that, The step of determining the impact identification signal waveform of the nuclear reactor equipment based on the impact identification signal and storing the impact identification signal waveform locally includes: Monitor the short effective signal strength value and determine whether the short effective signal strength value is greater than or equal to the product value; When the short effective signal strength value is greater than or equal to the product value, a target time point with a preset first time length is determined before the current time point; Starting from the target time point, a time period of a preset second time length is determined as the target time period. The impact recognition signal waveform is determined based on the impact recognition signal within the target time period, and the impact recognition signal waveform is stored locally.

6. The system according to claim 2, characterized in that, The digital processing module generates impact feature data based on the impact identification signal collected by the accelerometer, and sets a first channel identifier that corresponds to the accelerometer in the data. The abnormal equipment states include abnormal impact states; the detection of equipment status information of the nuclear reactor equipment based on the abnormal feature data includes: Upon receiving the impact feature data, the impact feature data is identified as the target impact feature data; Based on the first channel identifier of the target impact feature data, the acceleration sensor corresponding to the target impact feature data is determined, and the device space where the acceleration sensor is located is determined as the first target device space; All the acceleration sensors set in the space of the first target device are identified as the first target acceleration sensors, and the first channel identifier in each of the impact feature data received within a first preset time period after the target impact feature data is received is identified. Determine the number of first sensors that correspond to the first channel identifier among all the first target acceleration sensors, and determine whether the number of first sensors is greater than or equal to a preset first number threshold. If the number of the first sensors is greater than or equal to the first number threshold, then it is determined whether there is any device operation in the first target device space; If there is no device operation in the first target device space, the device status information is determined to be the abnormal impact state.

7. The system according to claim 1, characterized in that, The abnormal feature data also includes wear feature data; the acceleration signal also includes a wear identification signal; generating wear feature data of the nuclear reactor equipment based on the acceleration signal includes: The wear identification signal is received and filtered to obtain the original wear identification signal waveform; The effective value of the original wear identification signal waveform is extracted to obtain acceleration level data; The acceleration level data is compared with a preset acceleration level threshold. If the acceleration level data is greater than or equal to the acceleration level threshold, the wear characteristic data is generated.

8. The system according to claim 7, characterized in that, The digital processing module generates wear feature data based on the wear identification signal collected by the accelerometer, and a second channel identifier corresponding to the accelerometer is provided at the wear feature data location; The abnormal equipment condition also includes abnormal wear condition; the detection of equipment status information of the nuclear reactor equipment based on the abnormal feature data includes: Upon receiving the wear characteristic data, the wear characteristic data is identified as the target wear characteristic data; Based on the second channel identifier of the target wear feature data, the acceleration sensor corresponding to the target wear feature data is determined, and the device space where the acceleration sensor is located is determined as the second target device space; All the acceleration sensors set in the second target device space are identified as the second target acceleration sensors, and the second channel identifier in each wear feature data received within a second preset time period after the target wear feature data is received is identified. Determine the number of second sensors that correspond to the second channel identifier among all the second target acceleration sensors, and determine whether the number of second sensors is greater than or equal to a preset second number threshold. If the number of the second sensors is greater than or equal to the second number threshold, then it is determined whether there is any device operation in the second target device space; If there is no equipment operation in the second target equipment space, the equipment status information is determined to be the abnormal wear state.

9. The system according to claim 1, characterized in that, The comprehensive diagnostic module is also used for: When the device status information indicates abnormal wear or abnormal impact, an alarm message will be issued.

Images