Thermal power plant dust remover risk early warning method and system based on reduced-order model

Abstract

The invention relates to the technical field of structure risk early warning, and particularly discloses a thermal power plant dust remover risk early warning method and system based on a reduced-order model, and the method comprises the steps: deploying a multi-source sensor network at a key part of a dust remover to obtain multi-source heterogeneous data in real time, and constructing a dual-structure constraint mutual verification mechanism; physical consistency checking is carried out on multi-source heterogeneous data, a reduced-order mapping relation between monitoring point displacement and dangerous stress point stress is established in combination with a dust remover finite element model, and the structural stress state is rapidly reconstructed through a small amount of displacement data. According to the method, a stress time sequence is constructed based on theoretical stress, the structural stress evolution trend is evaluated, and a hierarchical and clear multi-level risk early warning mechanism is implemented in combination with structural safety redundancy calculation and mutual verification results, so that real-time monitoring and risk early warning of the structural state of the thermal power plant dust remover are realized, the monitoring reliability and early warning accuracy are greatly improved, and the safety of the thermal power plant dust remover is ensured. Powerful guarantee is provided for safe operation of the dust remover.

Description

Technical Field

[0001] This invention relates to the field of structural risk early warning technology, and more specifically, to a risk early warning method and system for dust collectors in thermal power plants based on a reduced-order model. Background Technology

[0002] Existing structural monitoring methods for dust collectors in thermal power plants mainly rely on single sensors or simple threshold judgments, lacking comprehensive analysis and verification mechanisms for multi-source data. This approach has significant limitations: first, single-sensor data is easily affected by environmental interference, leading to false alarms or missed alarms; second, the lack of physical constraint verification of sensor data makes it impossible to effectively identify spurious data; and third, structural stress state assessment typically relies on complex finite element analysis, failing to achieve real-time monitoring and early warning.

[0003] Current monitoring technologies typically process data from various sensors independently, ignoring the inherent correlations between different physical quantities, leading to insufficient reliability. Furthermore, traditional structural stress analysis requires establishing a complete finite element model and performing time-consuming calculations, which is insufficient for real-time monitoring and prevents timely detection and warning of structural risks.

[0004] Therefore, it is necessary to provide a risk early warning method and system for dust collectors in thermal power plants based on a reduced-order model to solve the above-mentioned technical problems. In order to solve the above problems, a technical solution is provided. Summary of the Invention

[0005] To overcome the aforementioned deficiencies in the prior art, this invention provides a risk early warning method and system for thermal power plant dust collectors based on a reduced-order model. This addresses the technical problems in existing thermal power plant dust collector structure monitoring, such as insufficient verification of sensor data reliability, inability to calculate structural stress in real time, and a single early warning mechanism.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A risk early warning method for dust collectors in thermal power plants based on a reduced-order model includes the following steps: Multi-source heterogeneous data is captured by deploying a multi-source sensor network in the dust collector of a thermal power plant, and the multi-source heterogeneous data is preprocessed. A dual structural constraint mutual verification mechanism is built based on multi-source heterogeneous data to obtain structural mutual verification results, and the true displacement vector of the dust collector structure is output according to the structural mutual verification results. By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Loads are applied to the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring point and the stress vector of the critical stress point is extracted. By performing matrix multiplication operations on the edge computing gateway, the actual displacement vector is mapped to the theoretical stress vector. Based on the theoretical stress vector, a stress time series is constructed and input into the structural stress risk evolution assessment model to assess whether there are any abnormalities in the structural stress evolution trend at the monitoring points. For monitoring points with normal structural stress evolution trends, the stress vector of dangerous stress points is extracted in real time to calculate the structural safety margin. Based on the structural mutual verification results and the structural safety margin, a multi-level early warning mechanism is implemented, and the early warning results are output.

[0007] As a further aspect of the present invention, a dual structural constraint mutual verification mechanism is built based on multi-source heterogeneous data to obtain structural mutual verification results, and the actual displacement vector of the dust collector structure is output according to the structural mutual verification results; wherein, the dual structural constraint mutual verification mechanism includes a primary structural constraint mutual verification model and a secondary structural constraint mutual verification model; the primary structural constraint mutual verification model is used to perform mutual verification of the tilt angle and displacement of the dust collector structure, and the secondary structural constraint mutual verification model is used to perform mutual verification of the settlement and cracks of the dust collector structure. Based on the results of the first and second structural constraint verifications, the true displacement vector is output.

[0008] As a further aspect of the present invention, a primary structural constraint mutual verification model is used to verify the tilt angle and displacement of the dust collector structure. The specific steps are as follows: Based on BeiDou differential positioning data, determine whether the monitoring point has undergone vertical relative displacement. If vertical relative displacement has occurred, use attitude sensors to measure the change in beam end rotation angle and calculate the corresponding theoretical displacement at the same moment. The displacement residual value is calculated based on the vertical relative displacement and theoretical displacement. Based on the displacement residual value, the change in beam end rotation angle, and the vertical relative displacement, it is determined whether the dust collector structure deformation has actually occurred. If the dust collector structure deformation has actually occurred, a structural constraint mutual verification is performed.

[0009] As a further aspect of the present invention, the determination of whether a dust collector structural deformation has actually occurred is based on a combination of displacement residual values, beam end rotation angle changes, and vertical relative displacement. If a dust collector structural deformation has indeed occurred, a structural constraint mutual verification is performed. The specific steps are as follows: If the displacement residual value is greater than the preset residual threshold, it is determined that a real dust collector structural deformation has occurred, and a structural constraint mutual verification is performed. If the vertical relative displacement is greater than the preset relative displacement threshold, and the change in beam end rotation angle... If so, it is determined that a false deformation of the dust collector structure has occurred and the first structural constraint mutual verification has not been passed; If the change in beam end rotation angle exceeds the preset rotation angle change threshold, and the vertical relative displacement... If the structure of the dust collector is falsely deformed, it is determined that the structural constraint mutual verification has failed.

[0010] As a further aspect of the present invention, a secondary structural constraint mutual verification model is used to perform mutual verification of settlement and cracks in the dust collector structure. The specific steps are as follows: Extracting structural variation data and vertical displacement data The structural anomaly data is compared with a preset anomaly threshold. If the structural anomaly data is greater than the preset anomaly threshold, and the vertical displacement data... If the deformation is false, it is determined that a false deformation of the dust collector structure has occurred and the secondary structural constraint mutual verification has not been passed; otherwise, it is determined that a true deformation of the dust collector structure has occurred and the secondary structural constraint mutual verification has been passed.

[0011] As a further aspect of the present invention, the monitoring points and critical stress points corresponding to the multi-source sensor network are determined by establishing a finite element model of the dust collector. A load is applied to the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring points and the stress vector of the critical stress points is extracted, as detailed below: By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Load conditions are applied to the finite element model of the dust collector. For each load condition, a static solution is performed on the finite element model to calculate the overall displacement and stress fields of the structure under the corresponding load conditions. After the calculation, the displacement response of each monitoring point under the load conditions is extracted to form the monitoring point displacement vector. At the same time, the stress response corresponding to each critical stress point is extracted to form the stress vector. ; A linear mapping relationship is constructed based on the displacement vector and stress vector of the monitoring points: In the formula: This is the stress influence matrix. This is a non-linear correction vector.

[0012] As a further aspect of the present invention, a stress time series is constructed based on the theoretical stress vector and input into the structural stress risk evolution assessment model to evaluate whether there are any anomalies in the structural stress evolution trend at the monitoring points. The specific steps are as follows: Stress time series constructed based on theoretical stress vectors ,in, Let be the theoretical stress vector at time t. For monitoring duration; Calculate the rate of stress change per unit time based on the stress time series. In the formula: Let be the rate of stress change at time t. Let be the theoretical stress vector at time t. Let be the theoretical stress vector at time t-1.

[0013] As a further aspect of the present invention, the stress change rate is compared with a preset reference rate range. If the stress change rate is within the preset reference rate range, the stress evolution trend of the monitoring point structure is normal; if the stress change rate exceeds the preset reference rate range, the stress evolution trend of the monitoring point structure is abnormal.

[0014] As a further aspect of the present invention, for monitoring points where the structural stress evolution trend is normal, the corresponding structural safety margin is calculated in real time. Based on the structural mutual verification results and the structural safety margin, a multi-level early warning mechanism is implemented, and the early warning results are output: Real-time acquisition of theoretical stress values ​​to calculate the structural safety margin at corresponding monitoring points In the formula: For structural safety margin, This is the theoretical stress value. The yield strength of the material. For safety factor; A multi-level early warning mechanism is implemented based on the structural mutual verification results and structural safety margin; the multi-level early warning mechanism includes the first early warning level, the second early warning level, the third early warning level, and the fourth early warning level.

[0015] A risk early warning system for dust collectors in thermal power plants based on a reduced-order model includes a multi-source heterogeneous data acquisition and processing module, a heterogeneous data dual mutual verification module, a finite element reduced-order mapping relationship construction module, a stress evolution assessment module, a structural safety margin calculation module, and a multi-level structural stress risk early warning module. The multi-source heterogeneous data acquisition and processing module is used to capture multi-source heterogeneous data by deploying a multi-source sensor network in the dust collector of a thermal power plant, and to preprocess the multi-source heterogeneous data. The heterogeneous data dual verification module is used to build a dual structural constraint verification mechanism based on multi-source heterogeneous data to obtain structural verification results and output the true displacement vector of the dust collector structure according to the structural verification results. The finite element order reduction mapping relationship construction module is used to determine the corresponding monitoring points and dangerous stress points of the multi-source sensor network by establishing a finite element model of the dust collector, apply loads to the finite element model of the dust collector for full calculation, and extract the correlation between the displacement vector of the monitoring point and the stress vector of the dangerous stress point. The stress evolution assessment module is used to map the real displacement vector to the theoretical stress vector by performing matrix multiplication operations on the edge computing gateway, construct a stress time series based on the theoretical stress vector, and input it into the structural stress risk evolution assessment model to assess whether there are any abnormalities in the structural stress evolution trend at the monitoring point. The structural safety margin calculation module is used to extract the stress vector of the dangerous stress point in real time and calculate the structural safety margin for monitoring points with normal stress evolution trends. The multi-level structural stress risk early warning module is used to implement a multi-level early warning mechanism based on the structural mutual inspection results and structural safety margin, and output the early warning results.

[0016] The technical effects and advantages of this invention, a risk early warning method and system for dust collectors in thermal power plants based on a reduced-order model, are as follows: This invention constructs a finite element model of the dust collector and extracts the linear mapping relationship between the displacement of monitoring points and the stress at dangerous stress points. This reduces the high-dimensional and complex full-scale finite element stress calculation process to matrix operations based on a small number of monitoring point displacement vectors, achieving rapid reconstruction of the structural stress state and significantly reducing computational complexity and power requirements. Furthermore, it introduces a dual structural constraint mutual verification mechanism based on multi-source heterogeneous data. This mechanism performs physical consistency verification of multi-source monitoring information, effectively distinguishing between real structural deformation and false anomalies caused by sensor noise, environmental disturbances, or installation errors. This improves the authenticity and reliability of displacement input data from the source. By constructing real displacement vectors only for data that passes mutual verification, it avoids the amplification of abnormal data. By incorporating stress mapping and risk assessment processes, the risk of false alarms and missed alarms is significantly reduced. A stress time series is constructed based on theoretical stress vectors, and the structural stress changes are dynamically assessed from the perspective of stress evolution rate. This not only identifies whether limits are exceeded but also captures abnormal signs of structural stress change trends in advance, helping to identify potential risks of dust collectors under conditions such as ash load fluctuations, temperature changes, or uneven foundation settlement. Combining structural safety margin calculations and structural verification results, a multi-level early warning mechanism with clear layers and explicit criteria is constructed. Structural risk states are subdivided into different levels, and differentiated prompts, warnings, or emergency response suggestions are output for different risk levels. This effectively avoids the problem of "frequent alarms but difficulty in decision-making" in traditional single-threshold alarm methods, improving the understandability of early warning information and the efficiency of handling by maintenance personnel.

[0017] This invention achieves real-time monitoring and risk warning of the structural status of dust collectors in thermal power plants through multi-source data spatiotemporal alignment, dual structural constraint mutual verification mechanism, reduced-order influence matrix stress inversion, and multi-level risk warning mechanism. This significantly improves the reliability of monitoring and the accuracy of warning, providing strong protection for the safe operation of dust collectors. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a risk early warning method for dust collectors in thermal power plants based on a reduced-order model, provided as an embodiment of the present invention; Figure 2 This is a system block diagram of a risk early warning system for dust collectors in thermal power plants based on a reduced-order model, provided as an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of this invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described technical solutions are only a part of this invention, and not all of it. All other technical solutions obtained by those skilled in the art based on the technical solutions of this invention without inventive effort are within the scope of protection of this invention.

[0020] like Figure 1 The diagram shown is a flowchart of a risk early warning method for dust collectors in thermal power plants based on a reduced-order model, provided by an embodiment of the present invention. Figure 1 The execution entity of the method shown can be a software and / or hardware device. The execution entity of this application can include, but is not limited to, at least one of the following: user equipment, network equipment, etc. User equipment can include, but is not limited to, computers, smartphones, personal digital assistants (PDAs), and the aforementioned electronic devices. Network equipment can include, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of computers or network servers. Cloud computing is a type of distributed computing, consisting of a super virtual computer composed of a group of loosely coupled computers. This embodiment does not limit this. Steps S1 to S5 are detailed as follows: Step S1: Capture multi-source heterogeneous data by deploying a multi-source sensor network in the dust collector of the thermal power plant, and preprocess the multi-source heterogeneous data. Step S2: Based on multi-source heterogeneous data, a dual structural constraint mutual verification mechanism is built to obtain the structural mutual verification results, and the true displacement vector of the dust collector structure is output according to the structural mutual verification results. Step S3: By establishing a finite element model of the dust collector, the corresponding monitoring points and dangerous stress points of the multi-source sensor network are determined. Loads are applied in the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring point and the stress vector of the dangerous stress point is extracted. Step S4: The actual displacement vector is mapped to the theoretical stress vector by performing matrix multiplication operation on the edge computing gateway. The stress time series is constructed based on the theoretical stress vector and input into the structural stress risk evolution assessment model to assess whether there are any abnormalities in the structural stress evolution trend at the monitoring point. Step S5: For monitoring points with normal structural stress evolution trends, extract the stress vector of the dangerous stress point in real time to calculate the structural safety margin. Based on the structural mutual verification results and the structural safety margin, implement a multi-level early warning mechanism and output the early warning results.

[0021] It should be noted that the multi-source heterogeneous data includes BeiDou differential positioning data, attitude perception data, vertical displacement data, and structural variation data.

[0022] The multi-source sensor network includes BeiDou differential terminals, attitude sensing devices, vertical displacement devices, and structural anomaly detection devices; multiple sensors are deployed at key locations in the dust collector of the thermal power plant to construct a comprehensive monitoring network. The collected input vectors include: BeiDou differential positioning data acquired by the BeiDou differential terminals. The sampling rate is relatively low (approximately 1 Hz); the three-axis Euler angles acquired by the attitude sensing device are used as attitude sensing data. The sampling rate is relatively high (approximately 5Hz); the relative settlement obtained by the vertical displacement device is used as the vertical displacement data. The structural variation data obtained by the structural variation device .

[0023] Preprocessing of multi-source heterogeneous data includes temporal alignment, spatial alignment, and normalization. Temporal alignment specifically involves unifying the temporal dimension of data collected by the multi-source sensor network to ensure temporal consistency. Using the BeiDou PPS (pulse-per-second) signal as the reference time, high-frequency data (such as attitude sensing data) undergoes sliding window averaging to reduce noise and align with the reference time; low-frequency data (such as BeiDou differential data) is interpolated to increase data density. Through these processes, multi-source heterogeneous data are unified to a standard timestamp, ensuring consistency and comparability of different physical quantities in the temporal dimension. This time-aligned data is a prerequisite for subsequent heterogeneous data cross-validation; only when multi-source heterogeneous data is aligned in time can effective physical geometric constraint verification be performed.

[0024] Spatial alignment specifically involves constructing a local coordinate system for the dust collector structure and unifying multi-source heterogeneous data under this coordinate system. Using a pre-defined rotation and translation matrix, the position data acquired by the BeiDou differential terminal in the WGS84 global coordinate system is converted into position data in the structural local coordinate system. Similarly, data from the attitude sensing device, vertical displacement device, and structural deformation device are also converted to the same coordinate system. This spatial alignment ensures that all physical quantities are described in the same reference frame, creating the necessary conditions for cross-verification of heterogeneous data based on physical geometric constraints.

[0025] Preferably, a dual structural constraint mutual verification mechanism is built based on multi-source heterogeneous data to obtain structural mutual verification results, and the actual displacement vector of the dust collector structure is output according to the structural mutual verification results; wherein, the dual structural constraint mutual verification mechanism includes a primary structural constraint mutual verification model and a secondary structural constraint mutual verification model; the primary structural constraint mutual verification model is used to verify the tilt angle and displacement of the dust collector structure, and the secondary structural constraint mutual verification model is used to verify the settlement and cracks of the dust collector structure. Based on the results of the first and second structural constraint verifications, the true displacement vector is output.

[0026] It should be noted that the actual displacement vector is fused with the vertical displacement data from the BeiDou differential positioning data.

[0027] In this embodiment of the invention, multi-source heterogeneous sensors, including BeiDou differential positioning, attitude sensors, and crack monitoring devices, are deployed at key load-bearing components, support nodes, and foundation parts of the dust collector to continuously collect multi-dimensional operational data such as structural tilt angle, vertical displacement, settlement, and crack propagation. Based on this, a dual structural constraint mutual verification mechanism is constructed to perform physical consistency and structural constraint verification on the multi-source heterogeneous data, thereby effectively distinguishing between actual structural deformation and sensor noise or environmental interference. The dual structural constraint mutual verification mechanism includes a primary structural constraint mutual verification model and a secondary structural constraint mutual verification model. The primary structural constraint mutual verification model focuses on the geometric constraint relationship between the dust collector's structural tilt angle and displacement, while the secondary structural constraint mutual verification model focuses on the mechanical consistency constraint between structural settlement and crack evolution.

[0028] Specifically, in the first structural constraint verification process, vertical displacement data obtained using BeiDou differential positioning is jointly analyzed with beam end rotation angle changes measured by attitude sensors. By comparing the measured vertical displacement with the theoretical displacement calculated from the structural geometry, it is determined whether the tilt angle change and displacement change are physically matched. When both meet the preset residual constraint conditions, it is determined that a real structural deformation has occurred, and the first structural constraint verification is passed. Subsequently, in the second structural constraint verification process, structural anomaly data obtained from a crack monitoring device is further introduced and cross-checked with vertical settlement data at the same monitoring location or related components. When crack propagation and structural settlement have consistent characteristics in time and amplitude, it is confirmed that the structure has undergone real stress deformation, and the second structural constraint verification is passed. If the two lack a mutual corroboration relationship, it is determined to be a false deformation, and abnormal data is suppressed.

[0029] In the dual structural constraint mutual verification mechanism, only data that passes both the primary and secondary structural constraint mutual verifications are retained and fused as a reliable structural response to ultimately output the true displacement vector of the dust collector structure. It should be noted that this true displacement vector uses the vertical displacement obtained from BeiDou differential positioning data as the core reference quantity, and integrates the tilt angle mutual verification correction results and settlement mutual verification results to perform error correction and reliability weighting on the original vertical displacement. Specifically, when both the primary and secondary mutual verifications pass, the BeiDou differential vertical displacement data is given a higher weight, and the displacement components obtained from tilt angle inversion are combined to perform consistency correction, thereby constructing a displacement vector containing the true vertical displacement components of each monitoring point. When any mutual verification fails, the displacement data of the corresponding monitoring point is reduced or removed to prevent abnormal data from entering the subsequent stress mapping and risk assessment process. Through this method, the true displacement vector can accurately reflect the actual deformation state of the dust collector structure under complex working conditions, providing a reliable data input basis for subsequent finite element model mapping, structural stress assessment, and multi-level early warning.

[0030] Preferably, a primary structural constraint verification model is used to verify the tilt angle and displacement of the dust collector structure. The specific steps are as follows: Based on BeiDou differential positioning data, it is determined whether vertical relative displacement has occurred at the monitoring point. If vertical relative displacement has occurred, the change in beam end rotation angle is measured using attitude sensors, and the corresponding theoretical displacement at the same moment is calculated. In the formula: For theoretical displacement, The length of the lever arm from the measuring point to the fulcrum. This represents the change in beam end rotation angle. For the second-order correction term generated by beam deflection; The displacement residual value is calculated based on the vertical relative displacement and the theoretical displacement. In the formula: This represents the displacement residual value. This is a vertical relative displacement. This is the theoretical displacement; The determination of whether the dust collector structure deformation has actually occurred is based on the displacement residual value, the change in beam end rotation angle, and the vertical relative displacement. If the dust collector structure deformation has actually occurred, a structural constraint mutual verification is performed.

[0031] Preferably, the determination of whether the dust collector structure deformation has actually occurred is based on a combination of displacement residual values, beam end rotation angle changes, and vertical relative displacement. If the dust collector structure deformation has actually occurred, a structural constraint mutual verification is performed. The specific steps are as follows: If the displacement residual value is greater than the preset residual threshold, it is determined that a real dust collector structural deformation has occurred, and a structural constraint mutual verification is performed. If the vertical relative displacement is greater than the preset relative displacement threshold, and the change in beam end rotation angle... If a false dust collector structural deformation occurs and fails the first structural constraint mutual verification, it is determined that there is Beidou differential multipath interference or tropospheric error. The alarm is automatically suppressed and the data is marked as suspicious. If the change in beam end rotation angle exceeds the preset rotation angle change threshold, and the vertical relative displacement... If a false deformation of the dust collector structure occurs and the first structural constraint mutual verification is not passed, it is determined that the attitude sensor is interfered with by wind load vibration or the support is loose.

[0032] In one embodiment of the present invention, in order to verify the authenticity of the dust collector structure monitoring data and prevent misjudgment caused by the anomaly of a single sensor, a primary structural constraint cross-verification model is constructed to cross-verify the tilt angle change and vertical displacement of the dust collector structure.

[0033] Specifically, firstly, based on the BeiDou differential positioning devices deployed at key parts of the dust collector structure, the vertical relative displacement data of the corresponding monitoring points is acquired in real time to determine whether there is a change in vertical displacement at the monitoring point. When a vertical relative displacement is detected at the monitoring point, the attitude sensor installed at the end of the corresponding structural beam is further invoked to obtain the change in the beam end rotation angle.

[0034] After obtaining the change in beam end rotation angle, the theoretical vertical displacement of the monitoring point at that moment is calculated based on the geometric relationship of the dust collector structure. Subsequently, the vertical relative displacement obtained by Beidou differential positioning is compared with the theoretical displacement to calculate the displacement residual value. The displacement residual value is used to characterize the degree of consistency between the measured displacement and the displacement calculated by the structural geometry.

[0035] Based on this, the residual displacement value, the change in beam end rotation angle, and the vertical relative displacement are used to determine whether the dust collector structure has actually undergone structural deformation. If the structure is determined to have actually deformed, the structural constraint mutual verification is considered to have passed.

[0036] In a preferred embodiment, the process for determining the actual structural deformation specifically includes the following scenarios: When the displacement residual value is greater than the preset residual threshold, it indicates that there is a significant deviation between the measured displacement and the theoretical displacement. Combined with the change of the beam end rotation angle, it is determined that the dust collector structure has actually deformed, and the structural constraint mutual verification is performed.

[0037] When the vertical relative displacement is greater than the preset relative displacement threshold, and the change in beam end rotation angle is approximately zero, it is determined that the displacement change does not conform to the structural geometric deformation characteristics and belongs to false structural deformation. At this time, the first structural constraint mutual verification fails. This anomaly is usually caused by the multipath effect of Beidou differential positioning or tropospheric error. The structure alarm is automatically suppressed and the corresponding data is marked as suspicious data.

[0038] When the change in beam end rotation angle exceeds the preset rotation angle change threshold, while the vertical relative displacement is approximately zero, the anomaly is determined to be caused by external interference such as wind load vibration or support loosening of the attitude sensor. This also does not conform to the characteristics of real structural deformation, and the first structural constraint mutual verification is deemed to have failed.

[0039] The above-mentioned structural constraint mutual verification model realizes the verification of the physical consistency between tilt angle data and displacement data, effectively distinguishes between real structural deformation and sensor anomalies, and provides a reliable data foundation for subsequent output of real displacement vectors and structural stress risk assessment.

[0040] Preferably, the secondary structural constraint mutual verification model is used to verify the settlement and cracks of the dust collector structure. The specific steps are as follows: Extracting structural variation data and vertical displacement data The structural anomaly data is compared with a preset anomaly threshold. If the structural anomaly data is greater than the preset anomaly threshold, and the vertical displacement data... If the deformation is false, it is determined that a false deformation of the dust collector structure has occurred and the secondary structural constraint mutual verification has not been passed; otherwise, it is determined that a true deformation of the dust collector structure has occurred and the secondary structural constraint mutual verification has been passed.

[0041] In one embodiment of the present invention, in the steel structure of a baghouse dust collector in a thermal power plant, crack monitoring sensors and high-precision settlement monitoring sensors are deployed in the connection area between the ash hopper and the main support frame to simultaneously acquire structural anomaly data and vertical displacement data. During operation, if the crack sensor detects structural anomalies exceeding a preset anomaly threshold, but the settlement monitoring results show that the vertical displacement at the corresponding monitoring point is essentially zero, it is generally determined that the anomaly may originate from local false responses caused by loose sensor installation, electromagnetic interference, or temperature stress, rather than actual settlement or cracking caused by overall structural stress. In this case, it is determined that the secondary structural constraint verification has not been passed, and the abnormal data is shielded to avoid falsely triggering structural safety warnings. Conversely, when the crack anomaly data exceeds the threshold and the settlement monitoring sensor simultaneously detects a significant change in vertical displacement, it indicates that the structure has undergone actual stress deformation under its own weight, ash load, or uneven foundation settlement. Crack propagation and structural settlement corroborate each other in terms of physical mechanism, thus determining that actual dust collector structural deformation has occurred. Through secondary structural constraint verification, a reliable data basis is provided for subsequent structural stress evolution analysis and safety margin assessment.

[0042] Preferably, by establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Loads are applied to the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring points and the stress vector of the critical stress points is extracted, as follows: By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Load conditions are applied to the finite element model of the dust collector. For each load condition, a static solution is performed on the finite element model to calculate the overall displacement and stress fields of the structure under the corresponding load conditions. After the calculation, the displacement response of each monitoring point under the load conditions is extracted to form the monitoring point displacement vector. At the same time, the stress response corresponding to each critical stress point is extracted to form the stress vector. ; A linear mapping relationship is constructed based on the displacement vector and stress vector of the monitoring points: In the formula: This is the stress influence matrix. This is a non-linear correction vector.

[0043] In this embodiment of the invention, firstly, based on the structural design drawings and actual installation parameters of the dust collector, a three-dimensional finite element model consistent with the physical object is established. The monitoring point positions corresponding to the multi-source sensor network are then marked in the model. Simultaneously, based on the structural stress analysis results, locations prone to stress concentration, such as the base of the supporting columns, the welded joints of the ash hopper, and the span of the crossbeams, are selected as critical stress points. Subsequently, various typical load conditions are applied to the finite element model, including the equipment's self-weight, ash accumulation load, wind load, and temperature effects. Static solutions are performed for each load condition to obtain the corresponding overall displacement field and stress field. After the calculation is completed, the monitoring points at each point are extracted from the displacement field. The displacement response under the combined load conditions constitutes the displacement vector of the monitoring point. Simultaneously, the stress response at each critical stress point is extracted from the stress field to form a stress vector. Based on displacement and stress vector samples obtained under multiple load conditions, a linear mapping relationship between the two is established through regression analysis. The stress influence matrix A nonlinear correction vector is used to characterize the sensitivity of displacement changes at monitoring points to stress response at critical stress points. It is used to compensate for deviations introduced by factors such as finite element simplification and idealized boundary conditions. During the actual operation of the dust collector, the edge computing gateway receives the real displacement vector obtained from the fusion of multiple source sensors in real time. It also performs matrix multiplication locally to quickly calculate the theoretical stress value at the critical stress point. This enables online estimation of the stress state at key critical stress points without requiring real-time full finite element calculations, providing an efficient and reliable data foundation for structural stress time series construction and risk evolution assessment.

[0044] Preferably, a stress time series is constructed based on the theoretical stress vector and input into the structural stress risk evolution assessment model to evaluate whether there are any anomalies in the stress evolution trend of the monitoring points. The specific steps are as follows: Stress time series constructed based on theoretical stress vectors ,in, Let be the theoretical stress vector at time t. For monitoring duration; Calculate the rate of stress change per unit time based on the stress time series. In the formula: Let be the rate of stress change at time t. Let be the theoretical stress vector at time t. Let be the theoretical stress vector at time t-1; The stress change rate is compared with a preset reference rate range. If the stress change rate is within the preset reference rate range, the stress evolution trend of the monitoring point structure is normal; if the stress change rate exceeds the preset reference rate range, the stress evolution trend of the monitoring point structure is abnormal.

[0045] In this embodiment of the invention, during long-term online monitoring of dust collectors in thermal power plants, the edge computing gateway calculates the theoretical stress vector at each sampling period based on the actual displacement vector using a stress influence matrix. This data is then stored and updated chronologically to gradually construct a stress time series for critical stress points. Based on this stress time series, at each new moment, the stress change rate between adjacent moments is automatically calculated to characterize the intensity and stability of structural stress evolution over time. Subsequently, the obtained stress change rate is compared with a pre-defined benchmark rate range based on historical stable operating conditions and finite element simulation results. When the stress change rate consistently falls within the preset benchmark rate range, the stress evolution trend of the monitoring point is determined to be normal, and the structural stress changes match environmental and operating condition disturbances. When the stress change rate exceeds the benchmark rate range, the stress evolution trend of the monitoring point is determined to be abnormal, indicating an abnormal change in the structural stress state. This provides a key criterion for subsequent safety margin calculations and multi-level early warning mechanism triggering.

[0046] Preferably, for monitoring points with normal structural stress evolution trends, the corresponding structural safety margin is calculated in real time. Based on the structural mutual inspection results and the structural safety margin, a multi-level early warning mechanism is implemented, and the early warning results are output. Real-time acquisition of theoretical stress values ​​to calculate the structural safety margin at corresponding monitoring points In the formula: For structural safety margin, This is the theoretical stress value. The yield strength of the material. For safety factor; A multi-level early warning mechanism is implemented based on the structural mutual verification results and structural safety margin; the multi-level early warning mechanism includes the first early warning level, the second early warning level, the third early warning level, and the fourth early warning level.

[0047] It should be noted that the specific implementation steps of the multi-level early warning mechanism are as follows: The first warning level is defined as follows: when the structural safety margin is within the preset normal range, but the rate of stress change shows a continuous monotonically increasing trend, the structural condition is determined to enter the first warning level. At this time, no risk alarm is triggered; only a structural condition change notification is output to the monitoring platform to remind maintenance personnel to pay attention to the structural deformation trend and increase the frequency of subsequent monitoring.

[0048] The second warning level is as follows: when any multi-source heterogeneous data exceeds the corresponding preset threshold but fails to pass the dual structure constraint mutual verification mechanism, it is judged to be in a warning state.

[0049] The third warning level is specifically defined as follows: when the cross-verification results of multi-source heterogeneous numbers pass the dual structural constraint cross-verification mechanism, and the structural safety margin is satisfied. At this point, a severe warning is triggered. Under this warning level, a severe warning message is sent to the monitoring platform, indicating that the structure is approaching the design safety boundary, and recommending corresponding operational adjustments or structural reinforcement measures to prevent further deterioration of the structural condition.

[0050] The fourth warning level is specifically defined as: when the structural safety margin... If the structural anomaly detection device detects a step change in structural anomaly data, it indicates a critical warning state. At this warning level, a critical warning message is issued, prompting the need for emergency response measures, including but not limited to shutdown for inspection, isolation of the risk area, or emergency reinforcement, to prevent structural failure or safety accidents.

[0051] In this embodiment of the invention, after the structural stress risk evolution assessment model determines that the stress evolution trend of each monitoring point is in a normal state, the structural safety margin of the corresponding monitoring point is automatically calculated based on the theoretical stress value output in real time by the edge computing gateway, thereby quantifying the safety margin between the current structural bearing capacity and the actual stress level. Subsequently, by combining the structural safety margin calculation results with the judgment results of the dual structural constraint mutual verification mechanism, a multi-level early warning mechanism with hierarchical classification is implemented to achieve refined management of the structural safety status of the dust collector.

[0052] Specifically, when the structural safety margin remains within the preset normal range, but stress time series analysis shows a continuous monotonically increasing stress change rate, the structural condition is determined to enter the first warning level. At this stage, no explicit risk alarm is triggered; only structural condition change alerts are output to the monitoring platform to remind maintenance personnel to pay attention to the structural deformation trend and to suggest appropriately increasing the subsequent monitoring frequency to identify potential risk developments early. When any multi-source heterogeneous monitoring data exceeds the corresponding preset threshold but fails to pass the dual structural constraint mutual verification mechanism, it is determined to be at the second warning level. At this time, the anomaly is more likely to originate from sensor interference or local non-structural factors. The platform outputs a warning level alert to guide maintenance personnel to prioritize checking the equipment status and sensor operational reliability.

[0053] When multi-source heterogeneous data is confirmed by a dual structural constraint mutual verification mechanism to accurately reflect the structural stress changes, and the calculated structural safety margin meets the requirements... At this point, the system enters the third warning level and triggers a severe warning. The monitoring platform then issues a severe warning to relevant personnel, clearly indicating that the structure is approaching its design safety boundary and recommending timely intervention measures such as adjusting operating conditions, reinforcing local structures, or reducing load to prevent further deterioration of the structure.

[0054] If further monitoring reveals that the structural safety margin has decreased to If the structural anomaly detection device detects a step change in structural anomaly data such as crack length, it is determined to enter the fourth warning level, namely the critical warning state. A critical warning message is immediately issued, and it is required to take emergency measures, including shutdown for inspection, isolation of high-risk areas, or implementation of emergency reinforcement, so as to minimize the risk of dust collector structural failure and safety accidents.

[0055] A risk early warning system for dust collectors in thermal power plants based on a reduced-order model includes a multi-source heterogeneous data acquisition and processing module, a heterogeneous data dual verification module, a finite element reduced-order mapping relationship construction module, a stress evolution assessment module, a structural safety margin calculation module, and a multi-level structural stress risk early warning module. The multi-source heterogeneous data acquisition and processing module is connected to the heterogeneous data dual verification module, which is connected to the finite element reduced-order mapping relationship construction module. The finite element reduced-order mapping relationship construction module is connected to the stress evolution assessment module, which is connected to the structural safety margin calculation module, and the structural safety margin is connected to the multi-level structural stress risk early warning module.

[0056] The multi-source heterogeneous data acquisition and processing module is used to capture multi-source heterogeneous data by deploying a multi-source sensor network in the dust collector of a thermal power plant, and to preprocess the multi-source heterogeneous data. The heterogeneous data dual verification module is used to build a dual structural constraint verification mechanism based on multi-source heterogeneous data to obtain structural verification results and output the true displacement vector of the dust collector structure according to the structural verification results. The finite element order reduction mapping relationship construction module is used to determine the corresponding monitoring points and dangerous stress points of the multi-source sensor network by establishing a finite element model of the dust collector, apply loads to the finite element model of the dust collector for full calculation, and extract the correlation between the displacement vector of the monitoring point and the stress vector of the dangerous stress point. The stress evolution assessment module is used to map the real displacement vector to the theoretical stress vector by performing matrix multiplication operations on the edge computing gateway, construct a stress time series based on the theoretical stress vector, and input it into the structural stress risk evolution assessment model to assess whether there are any abnormalities in the structural stress evolution trend at the monitoring point. The structural safety margin calculation module is used to extract the stress vector of the dangerous stress point in real time and calculate the structural safety margin for monitoring points with normal stress evolution trends. The multi-level structural stress risk early warning module is used to implement a multi-level early warning mechanism based on the structural mutual inspection results and structural safety margin, and output the early warning results.

[0057] like Figure 2 The diagram shown is a system block diagram of a risk early warning system for dust collectors in thermal power plants based on a reduced-order model, according to an embodiment of the present invention. This system can be used to execute... Figure 1 The steps in the method embodiments shown are implemented in a similar manner and have similar technical effects, and will not be repeated here.

[0058] Through the above embodiments, this invention, by constructing a finite element model of a dust collector and extracting the linear mapping relationship between the displacement of monitoring points and the stress at critical stress points, reduces the high-dimensional and complex full-scale finite element stress calculation process to matrix operations based on a small number of monitoring point displacement vectors. This achieves rapid reconstruction of the structural stress state, significantly reducing computational complexity and computational power requirements. Furthermore, by introducing a dual structural constraint mutual verification mechanism based on multi-source heterogeneous data, the invention performs physical consistency verification on multi-source monitoring information, effectively distinguishing between real structural deformation and false anomalies caused by sensor noise, environmental disturbances, or installation errors. This improves the authenticity and reliability of displacement input data from the source. By constructing true displacement vectors only for data that passes mutual verification, it avoids amplifying abnormal data and allowing it to enter the stress mapping and risk assessment process. This significantly reduces the risk of false alarms and missed alarms. By constructing a stress time series based on theoretical stress vectors and dynamically assessing structural stress changes from the perspective of stress evolution rate, it can not only identify whether limits are exceeded but also capture abnormal signs of structural stress change trends in advance. This helps to identify potential risks of dust collectors under conditions such as ash load fluctuations, temperature changes, or uneven foundation settlement. Combining structural safety margin calculations with structural verification results, a multi-level early warning mechanism with clear layers and explicit criteria is constructed. Structural risk states are subdivided into different levels, and differentiated prompts, warnings, or emergency response suggestions are output for different risk levels. This effectively avoids the problem of "frequent alarms but difficulty in decision-making" in traditional single-threshold alarm methods, improving the understandability of early warning information and the efficiency of handling by maintenance personnel.

[0059] This invention achieves real-time monitoring and risk warning of the structural status of dust collectors in thermal power plants through multi-source data spatiotemporal alignment, dual structural constraint mutual verification mechanism, reduced-order influence matrix stress inversion, and multi-level risk warning mechanism. This significantly improves the reliability of monitoring and the accuracy of warning, providing strong protection for the safe operation of dust collectors.

[0060] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

[0061] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A risk early warning method for dust collectors in thermal power plants based on a reduced-order model, characterized in that, Includes the following steps: Multi-source heterogeneous data is captured by deploying a multi-source sensor network in the dust collector of a thermal power plant, and the multi-source heterogeneous data is preprocessed. A dual structural constraint mutual verification mechanism is built based on multi-source heterogeneous data to obtain structural mutual verification results, and the true displacement vector of the dust collector structure is output according to the structural mutual verification results. By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Loads are applied in the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring point and the stress vector of the critical stress point is extracted. By performing matrix multiplication operations on the edge computing gateway, the actual displacement vector is mapped to the theoretical stress vector. Based on the theoretical stress vector, a stress time series is constructed and input into the structural stress risk evolution assessment model to assess whether there are any anomalies in the structural stress evolution trend at the monitoring points. For monitoring points with normal structural stress evolution trends, the stress vector of dangerous stress points is extracted in real time to calculate the structural safety margin. Based on the structural mutual verification results and the structural safety margin, a multi-level early warning mechanism is implemented, and the early warning results are output.

2. The risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 1, characterized in that, A dual structural constraint mutual verification mechanism is built based on multi-source heterogeneous data to obtain structural mutual verification results, and the actual displacement vector of the dust collector structure is output according to the structural mutual verification results. The dual structural constraint mutual verification mechanism includes a primary structural constraint mutual verification model and a secondary structural constraint mutual verification model. The primary structural constraint mutual verification model is used to verify the tilt angle and displacement of the dust collector structure, and the secondary structural constraint mutual verification model is used to verify the settlement and cracks of the dust collector structure. Based on the results of the first and second structural constraint verifications, the true displacement vector is output.

3. The risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 2, characterized in that, A structural constraint cross-verification model is used to verify the tilt angle and displacement of the dust collector structure. The specific steps are as follows: Based on BeiDou differential positioning data, determine whether the monitoring point has undergone vertical relative displacement. If vertical relative displacement has occurred, use attitude sensors to measure the change in beam end rotation angle and calculate the corresponding theoretical displacement at the same moment. The displacement residual value is calculated based on the vertical relative displacement and theoretical displacement. Based on the displacement residual value, the change in beam end rotation angle, and the vertical relative displacement, it is determined whether the dust collector structure deformation has actually occurred. If the dust collector structure deformation has actually occurred, a structural constraint mutual verification is performed.

4. The risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 3, characterized in that, Based on the displacement residual value, the change in beam end rotation angle, and the vertical relative displacement, it is determined whether the dust collector structure deformation has actually occurred. If the dust collector structure deformation has actually occurred, a structural constraint mutual verification is performed. The specific steps are as follows: If the displacement residual value is greater than the preset residual threshold, it is determined that a real dust collector structural deformation has occurred, and a structural constraint mutual verification is performed. If the vertical relative displacement is greater than the preset relative displacement threshold, and the change in beam end rotation angle... If so, it is determined that a false deformation of the dust collector structure has occurred and the first structural constraint mutual verification has not been passed; If the change in beam end rotation angle is greater than the preset rotation angle change threshold, and the vertical relative displacement... If the structure of the dust collector is falsely deformed, it is determined that the structural constraint mutual verification has failed.

5. A risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 3, characterized in that, The secondary structural constraint cross-verification model is used to cross-verify the settlement and cracks of the dust collector structure. The specific steps are as follows: Extracting structural variation data and vertical displacement data The structural anomaly data is compared with a preset anomaly threshold. If the structural anomaly data is greater than the preset anomaly threshold, and the vertical displacement data... If so, it is determined that a false deformation of the dust collector structure has occurred and the secondary structural constraint mutual verification has not been passed; Conversely, if no deformation occurs, it is determined that a real deformation of the dust collector structure has occurred, and the deformation is verified through secondary structural constraints.

6. The risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 1, characterized in that, By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Loads are applied to the finite element model of the dust collector for full calculation, and the correlation between the displacement vector of the monitoring points and the stress vector of the critical stress points is extracted, as detailed below: By establishing a finite element model of the dust collector, the corresponding monitoring points and critical stress points of the multi-source sensor network are determined. Load conditions are applied to the finite element model of the dust collector. For each load condition, a static solution is performed on the finite element model to calculate the overall displacement and stress fields of the structure under the corresponding load conditions. After the calculation, the displacement response of each monitoring point under the load conditions is extracted to form the monitoring point displacement vector. At the same time, the stress response corresponding to each critical stress point is extracted to form the stress vector. ; A linear mapping relationship is constructed based on the displacement vector and stress vector of the monitoring points: In the formula: This is the stress influence matrix. This is a non-linear correction vector.

7. A risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 1, characterized in that, A stress time series is constructed based on the theoretical stress vector and input into the structural stress risk evolution assessment model to evaluate whether there are any anomalies in the structural stress evolution trend at the monitoring points. The specific steps are as follows: Stress time series constructed based on theoretical stress vectors ,in, Let be the theoretical stress vector at time t. For monitoring duration; Calculate the rate of stress change per unit time based on the stress time series. In the formula: Let be the rate of stress change at time t. Let be the theoretical stress vector at time t. Let be the theoretical stress vector at time t-1.

8. A risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 7, characterized in that, The stress change rate is compared with the preset reference rate range. If the stress change rate is within the preset reference rate range, the stress evolution trend of the monitoring point structure is normal. If the rate of stress change exceeds the preset reference rate range, then the stress evolution trend of the monitoring point structure is abnormal.

9. A risk early warning method for dust collectors in thermal power plants based on a reduced-order model according to claim 1, characterized in that, For monitoring points with normal structural stress evolution trends, the corresponding structural safety margin is calculated in real time. Based on the structural mutual inspection results and the structural safety margin, a multi-level early warning mechanism is implemented, and the early warning results are output. Real-time acquisition of theoretical stress values ​​to calculate the structural safety margin at corresponding monitoring points In the formula: For structural safety margin, This is the theoretical stress value. The yield strength of the material. For safety factor; A multi-level early warning mechanism is implemented based on the results of structural mutual verification and structural safety margin. The multi-level early warning mechanism includes a first warning level, a second warning level, a third warning level, and a fourth warning level.

10. A risk early warning system for dust collectors in thermal power plants based on a reduced-order model, applied to the risk early warning method for dust collectors in thermal power plants based on a reduced-order model as described in any one of claims 1-9, characterized in that, It includes a multi-source heterogeneous data acquisition and processing module, a heterogeneous data dual mutual verification module, a finite element order reduction mapping relationship construction module, a stress evolution assessment module, a structural safety margin calculation module, and a multi-level structural stress risk early warning module. The multi-source heterogeneous data acquisition and processing module is used to capture multi-source heterogeneous data by deploying a multi-source sensor network in the dust collector of a thermal power plant, and to preprocess the multi-source heterogeneous data. The heterogeneous data dual verification module is used to build a dual structural constraint verification mechanism based on multi-source heterogeneous data to obtain structural verification results and output the true displacement vector of the dust collector structure according to the structural verification results. The finite element order reduction mapping relationship construction module is used to determine the corresponding monitoring points and dangerous stress points of the multi-source sensor network by establishing a finite element model of the dust collector, apply loads to the finite element model of the dust collector for full calculation, and extract the correlation between the displacement vector of the monitoring point and the stress vector of the dangerous stress point. The stress evolution assessment module is used to map the real displacement vector to the theoretical stress vector by performing matrix multiplication operations on the edge computing gateway, construct a stress time series based on the theoretical stress vector, and input it into the structural stress risk evolution assessment model to assess whether there are any abnormalities in the structural stress evolution trend at the monitoring point. The structural safety margin calculation module is used to extract the stress vector of the dangerous stress point in real time and calculate the structural safety margin for monitoring points with normal stress evolution trends. The multi-level structural stress risk early warning module is used to implement a multi-level early warning mechanism based on the structural mutual inspection results and structural safety margin, and output the early warning results.

Images