A method and device for unit protection based on multiple fiber grating temperature measuring points

Abstract

The application relates to the technical field of unit protection, and discloses a unit protection method and device based on multiple fiber grating temperature measuring points, which comprises the following steps: arranging a plurality of key temperature measuring areas in all temperature measuring areas, and performing space-time parallel analysis on the temperature monitored by each temperature measuring area; when the temperature of a certain temperature measuring area is abnormal, determining an abnormal level according to the abnormal type; when the abnormal level is low, setting the temperature measuring area as a key temperature measuring area, and dynamically adjusting a threshold coefficient; when the abnormal level is medium, setting the temperature measuring area as a key temperature measuring area, reducing the unit load, and dynamically adjusting the threshold coefficient; and when the abnormal level is high, performing emergency shutdown. The application provides a unit protection method and device based on multiple fiber grating temperature measuring points, which can solve or at least alleviate the problem that the existing unit temperature protection system cannot adjust the position of a key area according to the running state of a measured device, and cannot be well applied to a complex unit system.

Description

Technical Field

[0001] This invention relates to the field of generator protection technology, and in particular to a generator protection method and device based on multiple fiber Bragg grating temperature measurement points. Background Technology

[0002] For equipment fault diagnosis, real-time monitoring of equipment status information is the most direct and effective method. Traditional temperature sensors can only be deployed at critical locations, failing to achieve large-scale continuous monitoring, and are easily affected by environmental factors, leading to signal distortion and inaccurate reflection of actual operating conditions. Fiber Bragg gratings, as a novel passive device, possess advantages such as strong anti-electromagnetic interference capability, high sensitivity, and fast response speed, and are widely used in health monitoring of various engineering structures and equipment. Fiber Bragg grating sensors can achieve distributed multi-point measurements, making them ideal for studying the distribution characteristics of temperature fields under high temperature and high pressure environments.

[0003] However, the turbine generator set is a large unit with numerous areas requiring temperature monitoring. Different temperature measurement areas target different components, and these components have varying impacts on the overall operation of the turbine generator set. Problems in critical areas require timely intervention, while issues in less critical areas may not immediately affect the unit's operation. However, these affected areas still require close monitoring to prevent further disruption. Applying the same judgment criteria to all monitoring data can lead to misinterpretations, causing unnecessary shutdowns or equipment damage.

[0004] Chinese patent application CN103398801B discloses a fiber Bragg grating temperature measurement device and method, including a broadband light source, an optical fiber coupler, a tunable optical fiber filter, an optical fiber, a fiber Bragg grating sensor, a photodetector circuit module, and a measurement control module. The output of the optical fiber coupler is split into two paths, which enter two cascaded sensing fiber Bragg grating arrays via optical fibers. One sensing fiber Bragg grating array includes multiple horizontally arranged fiber Bragg gratings, and the other sensing fiber Bragg grating array includes multiple vertically arranged fiber Bragg gratings. The horizontal and vertical fiber Bragg gratings overlap in space to form a gridded fiber Bragg grating. Utilizing a variable grid, the device can acquire the temperature change of any grid point within the region in real time. Furthermore, through parameter self-calibration comparison and averaging, as well as network node redundancy technology, the system reliability is improved. The measurement method also mentions that in non-critical temperature measurement areas, gratings are installed on only three or two sides of the grid. This reduces the number of gratings used while still achieving parameter measurement. To prevent damage to the gratings in critical areas, other fiber optic gratings can pass through those areas. This way, if a grating in one area fails, another fiber optic grating can replace it for temperature monitoring, ensuring no data loss. While this method establishes critical and non-critical areas to ensure monitoring reliability, the critical areas are fixed and cannot be adjusted according to the operating status of the equipment under test, making it unsuitable for complex unit systems. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art, solve or at least alleviate the problem that the existing unit temperature measurement and protection system cannot adjust the position of key areas according to the operating status of the tested equipment, and cannot be well applied to complex unit systems, and to provide a unit protection method and device based on multiple fiber optic grating temperature measurement points.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a unit protection method based on multiple fiber Bragg grating temperature measurement points, comprising:

[0007] The temperature measurement areas of the unit form a spatial sequence according to the temperature transfer order. Each temperature measurement area forms a time series based on historical temperature data. Several key temperature measurement areas are arranged in all temperature measurement areas. The sampling frequency of key temperature measurement areas is higher than that of ordinary temperature measurement areas. Spatiotemporal parallel analysis is performed on the temperature monitored in each temperature measurement area.

[0008] When an abnormal temperature is detected in a certain temperature measurement area, the abnormality level is determined based on the type of abnormality.

[0009] When the anomaly level is low, the temperature measurement area is set as a key temperature measurement area, and the threshold coefficient is dynamically adjusted.

[0010] When the anomaly level is medium, the temperature measurement area is set as a key temperature measurement area, the unit load is reduced, and the threshold coefficient is dynamically adjusted.

[0011] When the anomaly level is high, an emergency shutdown is initiated.

[0012] To further realize the present invention, the following technical solutions may be preferred:

[0013] Preferably, each temperature measurement area is provided with multiple temperature measurement points, with all temperature measurement points in key temperature measurement areas being in working condition, and at least one temperature measurement point in ordinary temperature measurement areas being in working condition.

[0014] Preferably, when the anomaly type is a time anomaly, the anomaly level is determined by comparing it with historical temperatures and comparing it with the temperature changes in adjacent temperature measurement areas;

[0015] When the anomaly type is spatial anomaly, the anomaly level is determined by comparing it with historical temperatures and the position of the temperature measurement area in the spatial sequence.

[0016] Preferably, in the case of temporal anomalies, the absolute temperature deviation ΔT and the temperature rise rate dT / dt are calculated; in the case of spatial anomalies, the spatial temperature gradient is calculated. The calculation formula is as follows:

[0017] ΔT = T_current - T_historical baseline

[0018] dT / dt=(T_t-T_{t-Δt}) / Δt

[0019]

[0020] Where T_current is the detected temperature, T_historical baseline is the historical temperature, Δt is the change time, T_t is the temperature at time t, T_{t-Δt} is the temperature at time t-Δt, T_i and T_j are the temperatures of temperature measurement area i and temperature measurement area j, and d_ij is the distance between temperature measurement area i and temperature measurement area j.

[0021] Preferably, a weighted anomaly index W is also calculated for both temporal and spatial anomalies, and its calculation formula is as follows:

[0022]

[0023] Among them, ΔT_max, dT_max and These are the maximum allowable absolute temperature deviation, the maximum allowable temperature rise rate, and the maximum allowable spatial temperature gradient, respectively.

[0024] α, β, and γ are all weight coefficients, α+β+γ=1.

[0025] Preferably, each temperature measurement zone in the spatial sequence is set with a different γ according to its position in the unit. The higher the temperature of the component monitored by the temperature measurement zone and the greater its effect in the unit, the larger the γ will be, and the ratio of α to β will be a constant.

[0026] Preferably, the method for determining the anomaly level is as follows:

[0027] A condition is classified as a low-level anomaly if it meets any of the following criteria:

[0028] ΔT ∈ [10%λT_max, 30%λT_max) and lasts for >5 minutes;

[0029] dT / dt∈[2σ, 3σ);

[0030] W∈[0.4λ, 0.6λ);

[0031] A condition is classified as intermediate-level abnormal if it meets any two of the following criteria;

[0032] ΔT∈[30%λT_max, 50%λT_max);

[0033] dT / dt∈[3σ, 5σ);

[0034]

[0035] W∈[0.6λ, 0.8λ);

[0036] A condition is classified as a high-level anomaly if it meets any of the following criteria;

[0037] ΔT≥50%T_max;

[0038] dT / dt≥5σ;

[0039] Three adjacent temperature measurement areas simultaneously triggered a medium-level anomaly;

[0040] W≥0.8λ;

[0041] Where T_max is the highest allowable operating temperature in the temperature measurement area, σ is the standard deviation of the historical temperature rise rate, and λ is the threshold coefficient, with an initial value of 1.

[0042] Preferably, in the case of low-level anomalies, the adjustment formula for the threshold coefficient is λ_new = λ_original * (1 + 0.2W);

[0043] During intermediate-level anomalies, the formula for calculating the reduction in unit load is P_reduction = P_max * (1 - 0.8 * W), and the formula for adjusting the threshold coefficient is λ_new = λ_original * P_reduction / P_max.

[0044] In the event of a high-level anomaly, an emergency shutdown is initiated, and the spatial coordinates of the anomaly are recorded.

[0045] Preferably, the spatiotemporal parallel analysis involves constructing a spatiotemporal joint feature matrix after performing spatial and temporal dimension analysis on the temperature measurement area;

[0046] Among them, spatial dimension analysis includes spatial sequence anomaly detection and cross-verification of data from adjacent temperature measurement areas. Spatial sequence anomaly detection uses Kalman filtering to predict thermal diffusion path anomalies based on the temperature field propagation model. Cross-verification of data from adjacent temperature measurement areas judges anomalies based on temperature changes in adjacent temperature measurement areas.

[0047] Time-dimensional analysis includes time series anomaly detection and Kalman filter trend prediction. Time series anomaly detection identifies anomalies by comparing the measured temperature with the time series of the temperature measurement area. Kalman filter trend prediction calculates future predicted values ​​based on the measured temperature and the rate of temperature change, and identifies anomalies based on the predicted values.

[0048] A unit protection device based on multiple fiber Bragg grating temperature measurement points is provided, and a unit protection method based on multiple fiber Bragg grating temperature measurement points is employed.

[0049] The beneficial effects of this invention are:

[0050] 1. This invention performs spatiotemporal parallel analysis of all temperature measurement areas of the unit through spatial and time series, breaking through the limitations of single time series or spatial distribution. It can simultaneously capture the dynamic evolution law and spatial propagation characteristics of the equipment temperature field, and integrate historical data and real-time spatial distribution to construct a spatiotemporal joint prediction model, thereby improving the accuracy of temperature gradient prediction to over 92%.

[0051] 2. This invention achieves rapid location of abnormal hotspots through spatial sequence topology analysis. At the same time, this invention also sets up a key temperature measurement area that can change according to the abnormal situation. The location of the key temperature measurement area can be automatically concentrated on the newly emerging high temperature area, avoiding the "blind spot" problem of traditional fixed monitoring points. It can also dynamically adjust the data acquisition frequency, so that computing resources are tilted towards high-risk areas, and resource allocation is optimized.

[0052] 3. This invention enables the method to adapt to the environment by dynamically adjusting the threshold coefficient, integrates multiple parameters such as temperature gradient and load rate, realizes the linkage correction between the threshold and the health status of the equipment, and optimizes the graded triggering mechanism (early warning / load reduction / shutdown) to balance safety and operational continuity.

[0053] 4. This invention also introduces a weighted anomaly index to participate in the assessment of anomaly level, giving higher weight to data in important locations, quickly identifying sudden temperature changes, and linking the threshold coefficient with the weighted anomaly index for dynamic adjustment of the shutdown threshold, reducing the false alarm rate from 15% to below 3%. Attached Figure Description

[0054] Figure 1This is a flowchart of the unit protection method of the present invention.

[0055] Figure 2 This is a flowchart of the spatiotemporal parallel analysis of the present invention. Detailed Implementation

[0056] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0057] 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.

[0058] Example 1

[0059] The turbine generator unit is a large piece of equipment with numerous areas requiring temperature monitoring. Different temperature measurement areas target different components, and these components have varying impacts on the overall operation of the turbine generator unit. Problems in critical areas require timely intervention, while problems in less critical areas may not immediately affect the unit's operation. However, these affected areas still require close monitoring to prevent further impact. Applying the same judgment criteria to all monitoring data can lead to misinterpretations, causing unnecessary shutdowns or equipment damage.

[0060] Reference Figure 1 This embodiment discloses a unit protection method based on multiple fiber Bragg grating temperature measurement points, including:

[0061] Step S101: The temperature measurement areas of the unit form a spatial sequence according to the temperature transfer order. Each temperature measurement area forms a time series based on historical temperature data. Several key temperature measurement areas are arranged in all temperature measurement areas. The sampling frequency of key temperature measurement areas is higher than that of ordinary temperature measurement areas. Spatiotemporal parallel analysis is performed on the temperature monitored in each temperature measurement area.

[0062] Step S102: When the temperature of a certain temperature measurement area is abnormal, the abnormality level is determined according to the abnormality type;

[0063] Step S103: When the abnormality level is low, set the temperature measurement area as a key temperature measurement area and dynamically adjust the threshold coefficient.

[0064] Step S104: When the abnormality level is medium, set the temperature measurement area as the key temperature measurement area, reduce the unit load, and dynamically adjust the threshold coefficient.

[0065] Step S105: When the anomaly level is high, perform an emergency shutdown.

[0066] The technical solution provided in this application embodiment can realize comprehensive monitoring of the unit, improve monitoring efficiency and accuracy, reduce false alarm rate and false alarm rate, reduce labor cost input, realize intelligent management, improve operation and maintenance level, and ensure the safe and stable operation of equipment.

[0067] The spatial sequence mentioned in the embodiments of this application refers to a temperature measurement region sequence arranged in the order of temperature transfer, such as from upstream to downstream or from low temperature region to high temperature region.

[0068] In this embodiment, each temperature measurement area is equipped with multiple temperature measurement points. All temperature measurement points in the key temperature measurement area are in working condition, and at least one temperature measurement point in the ordinary temperature measurement area is in working condition. The data of multiple temperature measurement points in the key temperature measurement area are compared with each other, which can ensure the monitoring effect and verify the validity of the data. The ordinary temperature measurement area has fewer temperature measurement points, which reduces the use of computing resources and optimizes resource allocation while ensuring normal monitoring.

[0069] In addition, key temperature monitoring areas refer to those locations most prone to problems during unit operation, while ordinary temperature monitoring areas refer to relatively safer locations. By conducting high-frequency monitoring of key temperature monitoring areas and low-frequency monitoring of ordinary temperature monitoring areas, we can more effectively identify potential safety hazards.

[0070] Furthermore, the key temperature monitoring areas and ordinary temperature monitoring areas are not fixed but change according to the unit's operating status. For example, when an anomaly occurs at a certain location, it will be upgraded from an ordinary temperature monitoring area to a key temperature monitoring area for more stringent monitoring and control. It can automatically concentrate monitoring resources on newly emerging high-temperature areas, avoiding the "blind spot" problem of traditional fixed monitoring points; it can also dynamically adjust the data acquisition frequency, tilting computing resources towards high-risk areas for optimized resource allocation.

[0071] The "spatiotemporal parallel analysis" mentioned in this application is a method for data analysis that comprehensively considers both time and space dimensions. This method can help us better understand the correlation and regularity between data, thereby improving the accuracy and reliability of data analysis.

[0072] The "anomaly level" mentioned in this application embodiment is an indicator used to measure the severity of an anomaly. It can be flexibly set according to the actual situation and is not fixed. Generally, the level of anomaly is closely related to the corresponding handling measures.

[0073] This method enables environmental adaptability by dynamically adjusting the threshold coefficient, integrating multiple parameters such as temperature gradient and load rate to achieve linkage correction between the threshold and equipment health status, and optimizing the graded triggering mechanism (early warning / load reduction / shutdown) to balance safety and operational continuity.

[0074] Example 2

[0075] The technical solution in Example 1 only performs parallel analysis of spatial and time series, which can achieve mutual verification of data, but cannot effectively predict future temperature changes.

[0076] Reference Figure 2 In this embodiment, the spatiotemporal parallel analysis involves constructing a spatiotemporal joint feature matrix after performing spatial and temporal dimension analysis on the temperature measurement area.

[0077] Among them, spatial dimension analysis includes spatial sequence anomaly detection and cross-verification of data from adjacent temperature measurement areas. Spatial sequence anomaly detection uses Kalman filtering to predict thermal diffusion path anomalies based on the temperature field propagation model. Cross-verification of data from adjacent temperature measurement areas judges anomalies based on temperature changes in adjacent temperature measurement areas.

[0078] Time-dimensional analysis includes time series anomaly detection and Kalman filter trend prediction. Time series anomaly detection identifies anomalies by comparing the measured temperature with the time series of the temperature measurement area. Kalman filter trend prediction calculates future predicted values ​​based on the measured temperature and the rate of temperature change, and identifies anomalies based on the predicted values.

[0079] By performing spatiotemporal parallel analysis on all temperature measurement areas of the unit, the limitations of a single time series or spatial distribution are overcome. It can simultaneously capture the dynamic evolution law and spatial propagation characteristics of the equipment temperature field. By integrating historical data and real-time spatial distribution, a spatiotemporal joint prediction model is constructed, which improves the accuracy of temperature gradient prediction to over 92%.

[0080] Example 3

[0081] In the technical solution of Embodiment 1, when the temperature of the temperature measurement area is abnormal, the abnormality level is determined according to the type of abnormality. In order to determine the abnormality level more accurately.

[0082] When the anomaly type is time anomaly, the anomaly level is determined by comparing it with historical temperatures and the temperature changes in nearby temperature measurement areas.

[0083] When the anomaly type is spatial anomaly, the anomaly level is determined by comparing it with historical temperatures and the position of the temperature measurement area in the spatial sequence.

[0084] Furthermore, for time anomalies, the absolute temperature deviation ΔT and the temperature rise rate dT / dt are calculated; for spatial anomalies, the spatial temperature gradient is calculated. The calculation formula is as follows:

[0085] ΔT = T_current - T_historical baseline

[0086] dT / dt=(T_t-T_{t-Δt}) / Δt

[0087]

[0088] Where T_current is the detected temperature, T_historical baseline is the historical temperature, Δt is the change time, T_t is the temperature at time t, T_{t-Δt} is the temperature at time t-Δt, T_i and T_j are the temperatures of temperature measurement area i and temperature measurement area j, and d_ij is the distance between temperature measurement area i and temperature measurement area j.

[0089] Using temperature deviation ΔT, temperature rise rate dT / dt, and spatial temperature gradient By focusing on key parameters of abnormal temperature changes, it becomes easier to accurately determine the level of abnormality.

[0090] Example 4

[0091] Although the key parameters of temperature change were accurately calculated in Example 3, the impact of different temperature measurement areas on the normal operation of the unit is different. Therefore, relying solely on the key parameters of temperature change in each temperature measurement area cannot reflect the operating status of the entire unit and is prone to misjudgment.

[0092] For both temporal and spatial anomalies, a weighted anomaly index W is also calculated, and its calculation formula is as follows:

[0093]

[0094] Among them, ΔT_max, dT_max and These are the maximum allowable absolute temperature deviation, the maximum allowable temperature rise rate, and the maximum allowable spatial temperature gradient, respectively.

[0095] a, β, and γ are all weighting coefficients, and α + β + γ = 1.

[0096] In the spatial sequence, each temperature measurement zone is set with a different γ according to its position in the unit. The higher the temperature of the component monitored by the temperature measurement zone and the greater its effect in the unit, the larger the γ will be, and the ratio of α to β is a constant.

[0097] Based on key parameters of temperature change and the weighted anomaly index W, the method for determining the anomaly level is as follows:

[0098] A condition is classified as a low-level anomaly if it meets any of the following criteria:

[0099] ΔT ∈ [10%λT_max, 30%λT_max) and lasts for >5 minutes;

[0100] dT / dt∈[2σ, 3σ);

[0101] W∈[0.4λ, 0.6λ);

[0102] A condition is classified as intermediate-level abnormal if it meets any two of the following criteria;

[0103] ΔT∈[30%λT_max, 50%λT_max);

[0104] dT / dt∈[3σ, 5σ);

[0105]

[0106] W∈[0.6λ, 0.8λ);

[0107] A condition is classified as a high-level anomaly if it meets any of the following criteria;

[0108] ΔT≥50%T_max;

[0109] dT / dt≥5σ;

[0110] Three adjacent temperature measurement areas simultaneously triggered a medium-level anomaly;

[0111] W≥0.8λ;

[0112] Where T_max is the highest allowable operating temperature in the temperature measurement area, σ is the standard deviation of the historical temperature rise rate, and λ is the threshold coefficient, with an initial value of 1.

[0113] The threshold coefficient is dynamically adjusted according to the weighted anomaly index W at different anomaly levels. In the case of intermediate anomalies, the reduction in unit load is precisely adjusted according to the weighted anomaly index W, so that the unit can maintain a high load while meeting the requirements of normal operation.

[0114] For low-level anomalies, the adjustment formula for the threshold coefficient is λ_new = λ_original * (1 + 0.2W);

[0115] During intermediate-level anomalies, the formula for calculating the reduction in unit load is P_reduction = P_max * (1 - 0.8 * W), and the formula for adjusting the threshold coefficient is λ_new = λ_original * P_reduction / P_max;

[0116] In the event of a high-level anomaly, an emergency shutdown is initiated, and the spatial coordinates of the anomaly are recorded.

[0117] To more accurately assess the overall operating status of the unit, this embodiment introduces a weighted anomaly index W, which assigns higher weight to data at important locations, quickly identifies sudden temperature changes, and links the threshold coefficient with the weighted anomaly index for dynamic adjustment of the shutdown threshold.

[0118] Example 5

[0119] This application provides a unit protection device based on multiple fiber optic temperature measurement points. Using the aforementioned method, the unit protection device includes:

[0120] The monitoring unit is used to form a spatial sequence of the temperature measurement areas of the unit according to the temperature transmission order. Each temperature measurement area forms a time series based on historical temperature data. Several key temperature measurement areas are set up in all temperature measurement areas. The sampling frequency of key temperature measurement areas is higher than that of ordinary temperature measurement areas. Spatiotemporal parallel analysis is performed on the temperature monitored in each temperature measurement area.

[0121] The determination unit is used to determine the abnormality level based on the type of abnormality when an abnormality occurs in a certain temperature measurement area.

[0122] The processing unit is used to set the temperature measurement area as a key temperature measurement area and dynamically adjust the threshold coefficient when the anomaly level is low; when the anomaly level is medium, it sets the temperature measurement area as a key temperature measurement area, reduces the unit load, and dynamically adjusts the threshold coefficient; when the anomaly level is high, it performs an emergency shutdown.

[0123] The technical solution provided in this application enables comprehensive monitoring of the unit, improves monitoring efficiency and accuracy, reduces false alarm and missed alarm rates, reduces labor costs, achieves intelligent management, improves operation and maintenance levels, and ensures the safe and stable operation of the equipment.

[0124] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A unit protection method based on multiple fiber Bragg grating temperature measurement points, characterized in that, include: The temperature measurement areas of the unit form a spatial sequence according to the temperature transfer order. Each temperature measurement area forms a time series based on historical temperature data. Several key temperature measurement areas are arranged in all temperature measurement areas. The sampling frequency of key temperature measurement areas is higher than that of ordinary temperature measurement areas. Spatiotemporal parallel analysis is performed on the temperature monitored in each temperature measurement area. When an abnormal temperature is detected in a certain temperature measurement area, the abnormality level is determined based on the type of abnormality. When the anomaly level is low, the temperature measurement area is set as a key temperature measurement area, and the threshold coefficient is dynamically adjusted. When the anomaly level is medium, the temperature measurement area is set as a key temperature measurement area, the unit load is reduced, and the threshold coefficient is dynamically adjusted. When the anomaly level is high, an emergency shutdown is required. The spatiotemporal parallel analysis involves constructing a spatiotemporal joint feature matrix after performing spatial and temporal dimension analysis on the temperature measurement area. Spatial dimension analysis includes spatial sequence anomaly detection and cross-verification of data from adjacent temperature measurement areas. Spatial sequence anomaly detection uses Kalman filtering based on the temperature field propagation model to predict anomalies in the heat diffusion path. Cross-verification of data from adjacent temperature measurement areas judges anomalies based on temperature changes in adjacent temperature measurement areas. Time-dimensional analysis includes time series anomaly detection and Kalman filter trend prediction. Time series anomaly detection identifies anomalies by comparing the measured temperature with the time series of the temperature measurement area. Kalman filter trend prediction calculates future predicted values ​​based on the measured temperature and the rate of temperature change, and identifies anomalies based on the predicted values.

2. The unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 1, characterized in that, Each temperature measurement area is equipped with multiple temperature measurement points. In key temperature measurement areas, all temperature measurement points are in working condition, while in ordinary temperature measurement areas, at least one temperature measurement point is in working condition.

3. The unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 1, characterized in that, When the anomaly type is time anomaly, the anomaly level is determined by comparing it with historical temperatures and the temperature changes in nearby temperature measurement areas. When the anomaly type is spatial anomaly, the anomaly level is determined by comparing it with historical temperatures and the position of the temperature measurement area in the spatial sequence.

4. The unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 3, characterized in that, For time anomalies, calculate the absolute temperature deviation ΔT and the temperature rise rate dT / dt; for spatial anomalies, calculate the spatial temperature gradient. T is calculated using the following formula: ΔT = T_current - T_historical baseline; dT / dt = (T_t - T_{t-Δt}) / Δt; T = max(T_i - T_j) / d_ij; Where T_current is the detected temperature, T_historical baseline is the historical temperature, Δt is the change time, T_t is the temperature at time t, T_{t-Δt} is the temperature at time t-Δt, T_i and T_j are the temperatures of temperature measurement area i and temperature measurement area j, and d_ij is the distance between temperature measurement area i and temperature measurement area j.

5. A unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 4, characterized in that, For both temporal and spatial anomalies, a weighted anomaly index W is also calculated, and its calculation formula is as follows: W=α*(ΔT / ΔT_max)+β*(dT / dt / dT_max)+γ*( T / T_max); Among them, ΔT_max, dT_max and T_max represents the maximum allowable absolute temperature deviation, the maximum allowable temperature rise rate, and the maximum allowable spatial temperature gradient, respectively. α, β, and γ are all weight coefficients, α+β+γ=1.

6. The unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 5, characterized in that, In the spatial sequence, each temperature measurement zone is set with a different γ according to its position in the unit. The higher the temperature of the component monitored by the temperature measurement zone and the greater its effect in the unit, the larger the γ will be, and the ratio of α to β is a constant.

7. A unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 5, characterized in that, The method for determining the anomaly level is as follows: A condition is classified as a low-level anomaly if it meets any of the following criteria: ΔT ∈ [10%λT_max, 30%λT_max) and lasts for >5 minutes; dT / dt ∈ [2σ, 3σ); W∈ [0.4λ, 0.6λ); A condition is classified as intermediate-level abnormal if it meets any two of the following criteria; ΔT ∈ [30%λT_max, 50%λT_max); dT / dt ∈ [3σ, 5σ); T > 150% of the design allowable value; W∈ [0.6λ, 0.8λ); A condition is classified as a high-level anomaly if it meets any of the following criteria; ΔT ≥ 50%T_max; dT / dt ≥ 5σ; Three adjacent temperature measurement areas simultaneously triggered a medium-level anomaly; W ≥ 0.8λ; Where T_max is the highest allowable operating temperature in the temperature measurement area, σ is the standard deviation of the historical temperature rise rate, and λ is the threshold coefficient, with an initial value of 1.

8. A unit protection method based on multiple fiber Bragg grating temperature measurement points according to claim 7, characterized in that, For low-level anomalies, the adjustment formula for the threshold coefficient is λ_new = λ_original * (1 + 0.2W). During intermediate-level anomalies, the formula for calculating the reduction in unit load is P_reduction = P_max * (1 - 0.8 * W), and the formula for adjusting the threshold coefficient is λ_new = λ_original * P_reduction / P_max; In the event of a high-level anomaly, an emergency shutdown is initiated, and the spatial coordinates of the anomaly are recorded.

9. A unit protection device based on multiple fiber Bragg grating temperature measurement points, characterized in that, The unit protection method based on multiple fiber Bragg grating temperature measurement points as described in any one of claims 1-8 is adopted.

Images