Rod bundle channel flow instability early warning method based on disturbance wave modal recognition
By collecting and processing multi-source signals within the rod bundle channel, reconstructing the perturbation wave field, and performing mode decomposition, the shortcomings of existing early warning methods are addressed, enabling early identification and accurate warning of flow instability, thereby improving reactor safety.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NORTHEAST DIANLI UNIVERSITY
- Filing Date
- 2026-01-16
- Publication Date
- 2026-04-17
AI Technical Summary
Existing flow instability early warning methods can only effectively warn when the instability has fully developed. The warning time window is short, and they are easily affected by random noise and external disturbances in the system. Furthermore, it is difficult to accurately determine the type and stage of instability.
By collecting multi-source signals within the rod bundle channel, preprocessing them, and reconstructing the spatiotemporal evolution information of the disturbance wave field and two-phase flow characteristic parameters, modal decomposition is performed and modal characteristic parameters are calculated. Based on the early warning criteria, an evaluation is conducted, and graded early warnings are executed.
It enables advanced and accurate early warning of the early signs and development of flow instability, can distinguish the types of instability, and provides targeted operational guidance, thereby improving the safety control capabilities of the reactor.
Smart Images
Figure CN121884547A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of nuclear reactor thermal-hydraulic safety technology, and in particular to a method for early warning of flow instability in rod bundle channels based on disturbance wave mode identification. Background Technology
[0002] Within the core of a nuclear reactor (especially a light water reactor), fuel rods are typically arranged in bundles, forming complex flow channels. Under accident conditions in conventional pressurized water reactors or heavy water reactors, the two-phase flow within these bundles may experience flow instabilities, such as density wave oscillations (DWO) and pressure drop oscillations (PDO). These instabilities can also cause periodic oscillations in system parameters (such as pressure, flow rate, and void fraction), potentially leading to localized heat transfer deterioration, a sharp increase in fuel cladding temperature, or even fuel burnout, severely threatening reactor safety.
[0003] Existing methods for early warning of flow instability mainly rely on monitoring the time-domain thresholds or frequency-domain characteristics (such as the dominant frequency) of a single or a few key parameters (e.g., channel inlet / outlet pressure difference, outlet temperature, neutron noise). However, these methods have the following shortcomings: Effective alarms are typically only triggered when instability has fully developed and oscillations are significant. The warning window is short and easily affected by random system noise, external disturbances, or fluctuations in normal operating conditions, potentially leading to false alarms or missed alarms. Furthermore, the assessment of the type and stage of instability is not precise enough to provide deeper operational guidance. Summary of the Invention
[0004] The purpose of this invention is to provide a method for early warning of flow instability in rod bundle channels based on disturbance wave mode identification, aiming to solve or improve at least one of the above-mentioned technical problems.
[0005] To achieve the above objectives, the present invention provides the following solution: A method for early warning of flow instability in rod bundle channels based on perturbation wave mode identification includes: Multi-source signals of two-phase flow within the target rod bundle channel are acquired, and the multi-source signals are preprocessed to obtain multi-channel time-series data; the preprocessing includes bandpass filtering, detrending processing, and normalization processing; Based on the multi-channel time-series data, the spatiotemporal evolution information of the perturbation wavefield and the characteristic parameters of the two-phase flow is reconstructed; The spatiotemporal evolution information is subjected to modal decomposition, and the decomposed modal characteristic parameters are calculated; the modal characteristic parameters include modal spatial structure, oscillation frequency, energy ratio, and growth or decay rate; Based on a preset early warning criterion, the modal characteristic parameters calculated in real time are evaluated; the early warning criterion is related to the mapping relationship between different types of flow instability and specific disturbance wave modal characteristic parameters. Based on the assessment results, tiered early warnings are issued and relevant information is output.
[0006] Optionally, the method for acquiring the multi-source signals is as follows: At least three measurement sections are arranged along the axial direction on the target rod bundle channel. On each section, several sensors are arranged circumferentially according to the symmetry of the rod bundle sub-channels to form a sensor array. The sensor array is controlled by a unified clock source to synchronously acquire physical quantity signals from multiple spatial locations.
[0007] Optionally, the sensor uses a high-frequency dynamic pressure transmitter to collect pressure pulsation signals, and / or uses an equal neutron noise detector to collect neutron flux pulsation signals that are closely related to cavitation fraction fluctuations.
[0008] Optionally, the spatiotemporal evolution information of the perturbation wavefield and two-phase flow characteristic parameters is reconstructed based on the multi-channel time-series data, specifically including: The multi-channel time-series data is used to reconstruct a discrete spatiotemporal perturbation field at discrete time points based on the three-dimensional spatial coordinates of each sensor; the three-dimensional spatial coordinates consist of axial coordinates, circumferential coordinates, and radial coordinates. Cross-correlation analysis was used to quantitatively analyze the spatiotemporal disturbance field and the characteristic parameters of two-phase flow in three dimensions to determine the direction and velocity distribution information of disturbance propagation.
[0009] Optionally, modal decomposition is performed on the spatiotemporal evolution information, and the decomposed modal feature parameters are calculated, specifically including: The spatiotemporal evolution information is modally decomposed using an intrinsic orthogonal decomposition algorithm. Dominant perturbation wave modes are extracted from each decomposed mode, and corresponding modal characteristic parameters are calculated based on the dominant perturbation wave modes.
[0010] Optionally, based on preset warning criteria, the modal feature parameters calculated in real time are evaluated, specifically including: Based on historical experimental data, high-fidelity simulation results, or scaled-down bench tests, a "modal characteristic-instability state" mapping library is established, and early warning criteria are constructed based on the mapping library. The dominant modal features within the current sliding time window are calculated in real time. The key feature parameters are compared with the thresholds in the warning criteria, and the comparison results are output as the evaluation results.
[0011] Optionally, based on the assessment results, a tiered early warning system can be implemented and relevant information can be output, specifically including: Based on the assessment results, the current warning level is determined, and the spatial structure animation of the dominant disturbance wave mode, the historical trend curves of mode energy ratio and mode growth / attenuation rate, and the suspected instability types are dynamically displayed.
[0012] Optionally, the warning levels include: Level 1 warning indicates potential risks to stability and emerging trends, suggesting that "a potentially unstable mode has been detected and its evolution is being closely monitored." Level 2 warning indicates clear risks in growth and development, suggesting adjustments to operating parameters. A Level 3 warning indicates an imminent risk of instability and high risk, suggesting immediate action.
[0013] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: This invention discloses a method for early warning of flow instability in rod bundle channels based on perturbation wave mode identification. The method includes acquiring multi-source signals of two-phase flow within a target rod bundle channel and preprocessing them to obtain multi-channel time-series data; reconstructing the spatiotemporal evolution information of the perturbation wave field and two-phase flow characteristic parameters based on the multi-channel time-series data; performing modal decomposition on the spatiotemporal evolution information and calculating the decomposed modal characteristic parameters; evaluating the real-time calculated modal characteristic parameters based on a preset early warning criterion; the early warning criterion is related to the mapping relationship between different flow instability types and specific perturbation wave modal characteristic parameters; and executing graded early warnings and outputting relevant information based on the evaluation results. This invention can achieve advanced and accurate early warning of the nascent and early development of flow instability in conventional pressurized water reactors or heavy water reactors during accidents by identifying and analyzing the modal characteristics of specific perturbation waves in the flow channel in real time, and can distinguish the instability types. Attached Figure Description
[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0015] Figure 1 This is a flowchart illustrating the early warning method for flow instability in rod bundle channels based on disturbance wave mode identification according to the present invention. Figure 2 This is a schematic diagram of the overall prediction framework in this embodiment. Detailed Implementation
[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0017] The purpose of this invention is to provide a method for early warning of flow instability in rod bundle channels based on disturbance wave mode identification, aiming to solve or improve at least one of the above-mentioned technical problems.
[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0019] like Figures 1-2 As shown, this invention provides a method for early warning of flow instability in rod bundle channels based on perturbation wave mode identification, comprising: Step 1: Acquire multi-source signals of the two-phase flow within the target rod bundle channel, and preprocess the multi-source signals to obtain multi-channel time-series data; the preprocessing includes bandpass filtering, detrending processing, and normalization processing; the acquisition method of the multi-source signals is as follows: At least three measurement sections are arranged axially along the target rod bundle channel. On each section, several sensors are arranged circumferentially according to the symmetry of the rod bundle sub-channels, forming a sensor array. A unified clock source controls the sensor array to synchronously acquire physical quantity signals from multiple spatial locations. The sensors employ high-frequency dynamic pressure transmitters to acquire pressure pulsation signals and / or neutron noise detectors such as miniature fission chambers to acquire neutron flux pulsation signals closely related to cavitation fraction fluctuations.
[0020] As a specific implementation, in preprocessing, bandpass filtering is used to design and apply digital bandpass filters based on the frequency range of instability of interest (e.g., 0.1-10 Hz for density wave oscillations) to preserve the characteristic frequency band signal and suppress high-frequency noise and extremely low-frequency drift. Detrending is used to remove linear or slowly changing trend terms from the signal, ensuring that the analysis object is a pulsating component. Normalization is used to subtract the mean of the signal within its time window for each channel and divide by its standard deviation to obtain a preprocessed signal with zero mean and unit variance, facilitating subsequent modal analysis and eliminating the influence of dimensions and absolute amplitude.
[0021] Step 2: Reconstruct the spatiotemporal evolution information of the perturbation wavefield and the characteristic parameters of the two-phase flow within the channels based on the multi-channel time-series data; specific steps include: The multi-channel time-series data is used to reconstruct a discrete spatiotemporal disturbance field at discrete time points based on the three-dimensional spatial coordinates of each sensor; the three-dimensional spatial coordinates consist of axial, circumferential, and radial coordinates. Then, cross-correlation analysis is used to quantitatively analyze the spatiotemporal disturbance field and the disturbance propagation characteristics of the two-phase flow characteristic parameters in the three-dimensional direction, and to determine the direction and velocity distribution information of the disturbance propagation.
[0022] As a specific implementation method, taking axial propagation as an example, a sensor signal at a reference section (such as the inlet) is selected as the reference signal, and its cross-correlation function with the sensor signals at downstream sections is calculated. The time delay τ corresponding to the peak value of the cross-correlation function is then identified. max The axial propagation speed of the disturbance from the reference point to the downstream point is V. ax = Δz / τ max , where Δz is the axial distance. By analyzing the cross-correlation between different sensor pairs, the direction and velocity distribution of disturbance propagation can be inferred.
[0023] Step 3: Perform modal decomposition on the spatiotemporal evolution information and calculate the decomposed modal characteristic parameters; the modal characteristic parameters include modal spatial structure, oscillation frequency, energy proportion, and growth or decay rate; the specific process includes: The spatiotemporal evolution information is modally decomposed using an intrinsic orthogonal decomposition algorithm. Dominant perturbation wave modes are extracted from each decomposed mode, and corresponding modal characteristic parameters are calculated based on the dominant perturbation wave modes.
[0024] As a specific implementation method, the modal space structure feature vector is directly given, and its distribution pattern on the cross-section of the rod bundle channel can be observed visually (such as overall in-phase oscillation, out-of-phase oscillation of adjacent sub-channels, etc.). The modal oscillation frequency is determined by performing a Fourier transform on the time coefficient sequence and taking the frequency corresponding to the peak value of its power spectrum. The formula for calculating the modal energy ratio is: E i = λ i / Σ(λ j ) 100%, of which λ i Let i be the eigenvalue corresponding to this mode, and i ≠ j.
[0025] Modal growth / decay rate is a key early warning parameter, and it is fitted to the envelope of the time coefficient sequence (obtainable through Hilbert transform) within a sliding time window. Assume the amplitude change is A(t) ≈ A0. exp(σ i t), through logarithmic transformation and linear fitting, its slope is the growth rate σ. i σ i>0 indicates that the modal energy is increasing, and the system is becoming unstable; σ i <0 indicates attenuation and system stability.
[0026] Step 4: Based on preset early warning criteria, evaluate the modal characteristic parameters calculated in real time; the early warning criteria are related to the mapping relationship between different flow instability types and specific disturbance wave modal characteristic parameters; the specific process includes: Based on historical experimental data, high-fidelity simulation results, or scaled-down bench tests, a "modal feature-instability state" mapping library is established, and an early warning criterion is constructed based on the mapping library. Then, the dominant modal features within the current sliding time window are calculated in real time, and the key feature parameters are compared with the thresholds in the early warning criterion. The comparison results are output as the evaluation results.
[0027] The warning criteria specifically include: Criterion 1: Potential risk identification criteria (corresponding to Level 1 / Monitoring Level early warning).
[0028] This criterion is used to identify the earliest signs of instability. The trigger condition is met when the system identifies a dominant mode's spatial structure as having a low but clear similarity to the characteristic modes of known instability types (such as density wave oscillations or pressure drop oscillations), while the mode's growth rate fluctuates slightly near zero (neither decaying rapidly nor increasing significantly), but its energy share has stably exceeded the background noise level. At this point, the system determines that the disturbance structure is a potential risk source, having emerged from random noise and persisting, but not yet entering the development stage. The system will trigger a Level 1 (monitoring) warning, alerting operators to the evolution of this specific mode, but not requiring immediate action.
[0029] Criterion 2: Clearly define growth and development criteria (corresponding to Level 2 / Alert level warning).
[0030] This criterion is used to confirm that instability has entered an active development state. The triggering condition is met when the spatial structure of a specific mode is highly similar to the target instability type, its oscillation frequency falls within the characteristic frequency band of that type, and key parameters—the growth rate consistently exceeding a clear positive threshold and the energy proportion exceeding a set baseline—remain stable for a period of time. Physically, this indicates that the perturbation mode corresponding to this instability mechanism is being continuously excited by the system, its energy is accumulating steadily, and the system is clearly deviating from a stable state. The system will trigger a Level 2 (alert) warning, clearly indicating the instability type and its development trend, and recommending that operators prepare for parameter adjustments.
[0031] Criterion 3: Emergency instability Criterion (corresponding to Level 3 / Action-level Warning).
[0032] This criterion is used to determine whether the system is on the verge of instability or in the early stages of instability, requiring immediate intervention. It is triggered when a highly defined instability mode has risen to a dominant energy proportion and simultaneously meets either of the following acceleration conditions: first, its growth rate increases sharply, exceeding a higher danger threshold; second, the rate of increase in its energy proportion (time derivative) exceeds a critical value. This indicates that the unstable disturbance is accelerating and is about to, or has already, triggered violent oscillations in the system's macroscopic parameters. The system will trigger the highest level, Level 3 (action-level) warning, issuing a strong alarm and directly or automatically providing preset intervention recommendations (such as increasing the main pump speed or rapidly reducing power).
[0033] Criterion 4: Multimodal conflict arbitration criteria.
[0034] This criterion is used to handle complex situations where multiple modalities are active simultaneously or where a single modality's characteristics are ambiguous, ensuring the clarity and rationality of the early warning conclusions. When the system simultaneously assesses multiple instability risks based on the aforementioned criterion, it will select the type with the highest risk as the primary early warning output based on a comprehensive indicator of "the product of modal similarity, growth rate, and energy proportion." If a single modality's characteristics partially match multiple instability types but none fully satisfy the higher-level criteria, the system will output a "mixed feature risk" warning, listing all possibilities and their confidence levels, while maintaining or adopting a higher-level conservative early warning strategy to ensure safety.
[0035] Criterion 5: False alarm suppression and state recovery criteria.
[0036] This criterion aims to enhance the system's anti-interference capabilities and intelligence, preventing continuous false alarms caused by momentary disturbances. For triggered Level 1 or Level 2 warnings, the system will continuously monitor the growth rate of the relevant risk modes. If this growth rate consistently falls below the growth threshold over a prolonged period, the system determines that the disturbance growth trend has ceased or the system has stabilized, thereby automatically lowering the warning level or canceling the warning, and providing a status message indicating that "disturbance growth has stopped." This allows the warning system to dynamically reflect changes in the actual safety status.
[0037] Step 5: Based on the assessment results, implement tiered early warning systems and output relevant information. Specific steps include: Based on the assessment results, the current warning level is determined, and a corresponding color indicator is displayed. The system also dynamically displays an animation of the spatial structure of the dominant perturbation wave mode, historical trend curves of mode energy percentage and mode growth / attenuation rate, and the identified suspected instability types. The warning levels include: Level 1 warning indicates the presence of potential risks to stability and emerging conditions, suggesting that "a potential unstable mode has been detected and its evolution is being closely monitored"; Level 2 warning indicates the presence of clear growth and development risks, suggesting that adjustments to operating parameters be prepared; Level 3 warning indicates the presence of emergency instability and high risk, suggesting that immediate action be taken.
[0038] Based on the above technical solution, taking the early warning of density wave oscillation (DWO) in a pressurized water reactor rod bundle channel as an example, the implementation process is as follows.
[0039] This example illustrates a pressurized water reactor (PWR) where a main pump malfunction causes a sudden drop in coolant flow, triggering the safety protection system and activating the Emergency Core Cooling System (ECCS) to inject cryogenic emergency cooling water into the core. During the injection process, the cold water undergoes intense thermo-hydraulic interactions with the high-temperature coolant within the reactor, potentially leading to temperature stratification and density-difference-driven flow oscillations (such as density wave oscillations and thermal shock-induced oscillations), threatening the uniformity and stability of core cooling.
[0040] In the above-mentioned accident scenarios, the implementation process of this method in accident handling includes: Step 1: Sensor deployment and data acquisition.
[0041] Five measurement sections (Z1 near the inlet to Z5 near the outlet) are arranged axially along the target rod bundle channel. Each section contains four high-frequency dynamic pressure sensors and four fast-response thermocouples arranged circumferentially, forming an array of 40 sensors. The sensor array synchronously acquires pressure and temperature pulsation signals via a unified clock source at a sampling frequency of 500 Hz. An additional monitoring section is added near the emergency cooling water injection point to improve the spatial resolution of this area.
[0042] Step 2: Data preprocessing.
[0043] The original signal is bandpass filtered at 1–20 Hz (covering the oscillation frequency band that may be caused by cold water injection); detrending processing is performed to remove slow trends caused by power changes or overall system pressure decrease; the signals of each channel are normalized within a sliding time window (60 seconds) to obtain a multi-channel time series data matrix with zero mean and unit variance.
[0044] Step 3: Reconstruct the spatiotemporal evolution information of the perturbation wave field and the characteristic parameters of the two-phase flow.
[0045] Based on the sensor's three-dimensional coordinates, the spatiotemporal disturbance field coupled with pressure and temperature is reconstructed. Cross-correlation analysis is used to identify the propagation direction and velocity of the disturbance wave and two-phase flow after cold water injection. For example, a significant axial temperature gradient disturbance wave appears near the injection point, with a propagation velocity of approximately 1.5 m / s; the circumferential sensor pair shows asymmetric temperature and pressure oscillations, suggesting the possible existence of local flow deviations or vortex structures.
[0046] Step 4: Mode decomposition and feature parameter extraction.
[0047] The coupled perturbation field was decomposed using intrinsic orthogonal decomposition (POD); dominant modes (e.g., energy percentage > 50%) were extracted: Modal space structure: showing axial temperature stratification and pressure wave in-phase oscillation; oscillation frequency: dominant frequency approximately 1.2 Hz, consistent with density wave oscillation characteristics in cold water mixing process; energy percentage: first mode accounts for 58%; growth rate: calculated as σ1 = +0.03 s. -1 (showing an increasing trend).
[0048] Step 5: Real-time assessment based on early warning criteria.
[0049] The system has a built-in mapping library of "emergency cooling condition modal characteristics - instability states", including: thermal shock oscillations caused by cold water injection, mixed oscillations driven by density difference, and pressure oscillations caused by local vortex shedding. At this time, the modal spatial structure has a similarity of 0.82 with the density difference-driven oscillations; the growth rate exceeds the threshold for three consecutive windows (>+0.01 s). -1 The energy ratio continued to rise; the system triggered a level-two warning, indicating that "density wave oscillations caused by cold water injection have been detected, and it is recommended to adjust the injection strategy or start auxiliary mixing measures."
[0050] Step 6: Tiered early warning and operational guidance.
[0051] After a Level 2 warning (yellow) is triggered, the operation interface will dynamically display: a modal structure animation showing the temperature fluctuation propagation in the mixing zone of cold water and high-temperature coolant; a trend curve showing the modal growth rate and energy percentage continuously increasing; and a text prompt suggesting "gradually adjust the injection flow rate to avoid concentrated impact of cold water and start the circulation pump to assist mixing".
[0052] During the emergency response assessment, based on the warning prompts, the operator reduced the injection flow rate from 100% to 70% and started the auxiliary circulation pump. Data monitoring was then conducted. Within 10 minutes of the injection adjustment, the dominant mode growth rate increased from +0.03 s. -1 Gradually decreased to -0.005 s -1 The energy percentage decreased from 58% to 25%, while the energy of the second mode (stable mixed mode) increased; the fluctuation amplitude of the core outlet temperature decreased by about 60%.
[0053] At this point, the warning level is adaptively adjusted: based on criterion 5 (false alarm suppression and state recovery criterion), the system automatically lowers the warning level from level two to level one, and finally cancels the warning, indicating "flow tends to stabilize and disturbance growth has stopped".
[0054] This embodiment demonstrates that the method described in this invention, even in complex transient conditions such as emergency cooling water injection, can still achieve early warning of the nascent stage of flow instability through multi-source signal acquisition, spatiotemporal field reconstruction, and modal recognition, and provide targeted operational guidance based on intelligent early warning criteria. Through real-time monitoring of modal characteristic evolution, the effectiveness of emergency response can be quantitatively evaluated, providing closed-loop decision support for accident handling and significantly improving the safety and control capabilities of nuclear reactors under extreme conditions.
[0055] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.
[0056] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.
Claims
1. A rod bundle channel flow instability early warning method based on perturbation wave mode identification, characterized in that, include: Multi-source signals of two-phase flow within the target rod bundle channel are acquired, and the multi-source signals are preprocessed to obtain multi-channel time-series data. The preprocessing includes bandpass filtering, detrending processing, and normalization processing; Based on the multi-channel time-series data, the spatiotemporal evolution information of the perturbation wavefield and the characteristic parameters of the two-phase flow is reconstructed; The spatiotemporal evolution information is subjected to modal decomposition, and the decomposed modal characteristic parameters are calculated; the modal characteristic parameters include modal spatial structure, oscillation frequency, energy ratio, and growth or decay rate; Based on preset warning criteria, the modal characteristic parameters calculated in real time are evaluated; the warning criteria are related to the mapping relationship between different types of flow instability and specific disturbance wave modal characteristic parameters. Based on the assessment results, tiered early warnings are issued and relevant information is output.
2. The rod bundle channel flow instability warning method based on perturbation wave mode identification according to claim 1, characterized in that, The method for acquiring the multi-source signals is as follows: At least three measurement sections are arranged along the axial direction on the target rod bundle channel. On each section, several sensors are arranged circumferentially according to the symmetry of the rod bundle sub-channels to form a sensor array. The sensor array is controlled by a unified clock source to synchronously acquire physical quantity signals from multiple spatial locations.
3. The method for early warning of flow instability in rod bundle channels based on disturbance wave mode identification according to claim 2, characterized in that, The sensor uses a high-frequency dynamic pressure transmitter to collect pressure pulsation signals and / or uses a neutron noise detector to collect neutron flux pulsation signals that are closely related to cavitation fraction fluctuations.
4. The method for early warning of flow instability in rod bundle channels based on disturbance wave mode identification according to claim 1, characterized in that, Based on the multi-channel time-series data, the spatiotemporal evolution information of the perturbation wavefield and two-phase flow characteristic parameters is reconstructed, specifically including: The multi-channel time-series data is used to reconstruct a discrete spatiotemporal perturbation field at discrete time points based on the three-dimensional spatial coordinates of each sensor; the three-dimensional spatial coordinates consist of axial coordinates, circumferential coordinates, and radial coordinates. Cross-correlation analysis was used to quantitatively analyze the spatiotemporal disturbance field and the characteristic parameters of two-phase flow in three dimensions to determine the direction and velocity distribution information of disturbance propagation.
5. The rod bundle channel flow instability warning method based on perturbation wave mode identification according to claim 1, characterized in that, The spatiotemporal evolution information is subjected to modal decomposition, and the decomposed modal feature parameters are calculated, specifically including: The spatiotemporal evolution information is modally decomposed using an intrinsic orthogonal decomposition algorithm. The dominant perturbation wave mode is extracted from each decomposed mode, and the corresponding modal characteristic parameters are calculated based on the dominant perturbation wave mode.
6. The rod bundle channel flow instability warning method based on perturbation wave mode identification according to claim 1, characterized in that, Based on preset early warning criteria, the modal characteristic parameters calculated in real time are evaluated, specifically including: Based on historical experimental data, high-fidelity simulation results, or scaled-down bench tests, a "modal characteristic-instability state" mapping library is established, and early warning criteria are constructed based on the mapping library. The dominant modal features within the current sliding time window are calculated in real time. The key feature parameters are compared with the thresholds in the warning criteria, and the comparison results are output as the evaluation results.
7. The rod bundle channel flow instability warning method based on perturbed wave mode identification according to claim 1, characterized in that, Based on the assessment results, a tiered early warning system will be implemented and relevant information will be output, including: Based on the assessment results, the current warning level is determined, and the spatial structure animation of the dominant disturbance wave mode, the historical trend curves of mode energy ratio and mode growth / attenuation rate, and the suspected instability types are dynamically displayed.
8. The rod bundle channel flow instability warning method based on perturbed wave mode identification according to claim 7, characterized in that, The warning levels include: Level 1 warning indicates potential risks to stability and emerging trends, suggesting that "a potentially unstable mode has been detected and its evolution is being closely monitored." Level 2 warning indicates clear risks in growth and development, suggesting adjustments to operating parameters. A Level 3 warning indicates an imminent risk of instability and high risk, suggesting immediate action.