Water-cooled wall thermal fluid-solid multi-field coupling simulation measured data assimilation method and system

Abstract

The application relates to the technical field of boiler heating surface state simulation and online monitoring, and discloses a water cooling wall thermal flow-solid multi-field coupling simulation actual measurement data assimilation method and system, which maps water cooling wall temperature, in-pipe flow and structure response actual measurement data to corresponding calculation nodes of a multi-physical field simulation model, outputs multi-field prediction values, calculates and standardizes multi-dimensional coupling residuals. The residuals are disassembled into state layer residuals and parameter layer residuals in combination with residual persistence and cross-physical field correlation, various parameter offset types are discriminated, and a residual-parameter attribution mapping matrix is constructed. The application corrects the multi-field real-time coupling state through the state layer residuals, updates the thermal boundary, heat exchange, flow resistance and material performance slow change parameters through the parameter layer residuals, limits and constrains the correction results in combination with the multi-physical field consistency rules, and effectively improves the precision and long-term operation stability of the water cooling wall multi-field coupling simulation.

Description

Technical Field

[0001] This invention relates to the field of boiler heating surface condition simulation and online monitoring technology, specifically a method and system for assimilating measured data from a multi-field coupled simulation of heat, fluid, and solid-state structures in a water-cooled wall. Background Technology

[0002] Large power plant boilers' water-cooled walls are subjected to the coupled effects of high-temperature radiation from the furnace, heat exchange from steam-water flow within the tubes, and the constraint of tube wall thermal expansion over long periods. During operation, not only do wall temperatures change dynamically, but the distribution of the working fluid within the tubes, local pressure drops, and the thermal strain and stress of the tube walls also change simultaneously, exhibiting typical thermal-fluid-solid multi-physics field coupled operating characteristics. Accurately understanding the multi-field coupled state of the water-cooled walls is a core element in ensuring the safe and stable operation of power plant boilers.

[0003] Current methods for analyzing the condition of water-cooled walls primarily rely on offline numerical simulations, wall temperature prediction models, and monitoring data from limited measurement points. These methods solve for the water-cooled wall temperature distribution by pre-setting furnace thermal boundaries, internal hydrodynamic conditions, and material parameters. While this approach can provide a preliminary characterization of the water-cooled wall's thermal state, the model boundary conditions and core parameters mostly utilize design values, empirical values, or results from a single calibration, failing to adapt to the dynamic changes during long-term unit operation. In actual operation, factors such as furnace combustion deviation, coking and ash accumulation on the water-cooled walls, decreased heat transfer capacity within the tubes, changes in flow resistance, and material property degradation can cause continuous long-term shifts in model parameters. Traditional methods cannot effectively capture these shifts, leading to significant discrepancies between simulation results and actual operating conditions.

[0004] In the relevant prior art, patent CN115374636A discloses a wall temperature prediction method that couples numerical simulation and performance calculation of a boiler. This method involves building a physical model of the boiler to conduct numerical simulation, obtaining the wall heat flux density and wall temperature distribution, using these as initial conditions for hydrodynamic calculations to solve for the tube wall temperature, and iteratively correcting the model's wall temperature boundary conditions until the calculation converges. This achieves the coupling of numerical simulation and hydrodynamic calculation, effectively improving the accuracy of wall temperature prediction under specific operating conditions.

[0005] However, this approach has significant technical flaws. Its optimization focuses solely on wall temperature prediction and hydrodynamic calculation correction, limiting the correction to wall temperature boundary conditions and hydrodynamic calculation results. This method cannot distinguish whether simulation deviations originate from short-term operating condition disturbances or slow-varying model parameter shifts. Furthermore, it fails to conduct coupled analysis using multi-dimensional residuals of temperature, flow, and structure, thus failing to achieve physical attribution of multi-dimensional shifts in thermal boundaries, flow heat transfer, and structural parameters.

[0006] In summary, existing technologies have two major shortcomings: First, the measured data utilizes only a single dimension, being used only to correct wall temperature and single boundary conditions, failing to constrain the multi-field coupling relationships of thermal, flow, and structural fields; second, the lack of a biased hierarchical discrimination and correction mechanism cannot distinguish between instantaneous state errors and long-term parameter shifts. Blindly updating model parameters with short-term disturbances will cause unreasonable model drift; while ignoring long-term parameter shifts and treating them only as instantaneous errors will significantly reduce the long-term predictive stability of the model. Summary of the Invention

[0007] In order to overcome the defects of the existing technology, the purpose of this invention is to provide a method and system for assimilating measured data from multi-field coupled simulation of water-cooled walls, so as to solve the technical problems of lack of multi-field coupling constraints, inability to attribute deviations in a hierarchical manner, and disordered parameter correction mechanism in existing boiler water-cooled wall simulation analysis technology.

[0008] This invention is achieved through the following technical solution: In a first aspect, the present invention provides a method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall, comprising: Collect on-site measured temperature data of the water-cooled wall, flow data inside the tube, and structural response data, and map the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model; Run the thermo-fluid-solid multiphysics coupled simulation model and output the temperature prediction, flow state prediction, and structural response prediction values ​​that match the measured data. The coupled residuals between the measured data and the corresponding predicted values, including temperature residuals, flow residuals, and structural response residuals, are calculated and standardized. The coupled residuals are then decomposed into state-level residuals and parameter-level residuals by combining the standardized residual persistence index and the cross-physics field correlation index. Based on the combined distribution characteristics of temperature residual, flow residual, and structural response residual, the offset type corresponding to the parameter layer residual is determined to be thermal boundary offset, flow heat transfer offset, or structural parameter offset, and a residual-parameter attribution mapping matrix is ​​constructed. The temperature field, flow field, and structural field of the model within the current assimilation cycle are corrected using the state layer residuals; based on the parameter layer residuals and combined with the residual-parameter attribution mapping matrix, the furnace side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material property parameters are corrected simultaneously. Based on the rules of heat transfer consistency, flow continuity within the pipe, and structural response consistency, the obtained field correction results and model parameter correction results are subject to amplitude limiting constraints; the coupled field state after amplitude limiting and the updated model parameters are used as the initial conditions for the next assimilation cycle simulation.

[0009] Preferably, the specific process of mapping the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model is as follows: divide the water-cooled furnace into independent pipe segment regions according to the orientation, pipe row number, and elevation, and associate each pipe segment region with the thermal field, flow field, and structural field calculation nodes respectively; match the temperature data to the thermal field calculation node, match the flow data inside the pipe to the flow field calculation node, and match the structural response data to the structural field calculation node.

[0010] Preferably, the coupling residual standardization process is as follows: the temperature residual, flow residual, and structural response residual are integrated into a coupling residual vector, and the coupling residual vector is standardized according to the normalization scale of each of the three types of residuals.

[0011] Preferably, the acquisition and stratification rules for the standardized residual persistence index include extracting the mean of standardized residuals from multiple consecutive assimilation periods and calculating the residual persistence index from the changes in standardized residuals in adjacent periods, and determining the proportion of coupled residuals allocated to the parameter layer residuals based on the residual persistence index.

[0012] Preferably, the cross-physical field correlation includes temperature-structure consistency, temperature-flow consistency, and spatial neighborhood consistency; temperature-structure consistency characterizes the matching relationship between temperature residuals and structural response residuals, temperature-flow consistency characterizes the matching relationship between temperature residuals and flow anomalies within the pipe, and spatial neighborhood consistency characterizes the continuity of residual distribution between adjacent pipe segments.

[0013] Furthermore, an effective residual threshold is set before the consistency determination; if the standardized residual does not reach the effective residual threshold, the weight of the corresponding consistency index on the parameter layer residual allocation is reduced.

[0014] Preferably, the step of determining whether the offset type corresponding to the parameter layer residual is thermal boundary offset, flow heat transfer offset, or structural parameter offset based on the combined distribution characteristics of temperature residual, flow residual, and structural response residual specifically includes: When the temperature residual is continuously distributed in the same direction along the same side of the furnace or adjacent pipe sections, and the flow residual does not reach the flow anomaly threshold, it is determined that there is a thermal boundary offset. When the temperature residual is consistently high, and the flow rate in the pipe is low or the pressure drop in the pipe is high, it is determined that there is a flow heat transfer deviation. When temperature and flow residuals cannot explain the recurring structural response residuals, it is determined that there is a structural parameter offset.

[0015] Preferably, the constructed residual-parameter attribution mapping matrix is ​​a four-row, three-column matrix; the four rows of the matrix correspond to the correction directions of furnace-side thermal boundary parameters, tube-side heat transfer parameters, flow resistance parameters, and material performance parameters, respectively, and the three columns of the matrix correspond to the temperature residual, flow residual, and structural response residual, respectively.

[0016] Preferably, the simultaneous correction of furnace-side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters based on parameter layer residuals combined with the residual-parameter attribution mapping matrix includes: When a thermal boundary offset is determined to exist, the mapping weight of the thermal boundary parameter pointing to the furnace side of the temperature residual is increased; When a flow heat transfer offset is determined to exist, the mapping weights of temperature, flow residual pointing to in-tube heat transfer parameters, and flow resistance parameters are increased. When a structural parameter offset is determined to exist, the mapping weight of the structural response residual to the material performance parameters is increased.

[0017] Secondly, the present invention also provides a system for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall, comprising: The measured data alignment unit is used to collect on-site measured temperature data of the water-cooled wall, flow data inside the tube, and structural response data, and to map the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model. The coupled simulation prediction unit is used to run the thermo-fluid-solid multiphysics coupled simulation model and output temperature prediction, flow state prediction, and structural response prediction values ​​that match the measured data. The residual decomposition unit is used to calculate the coupled residuals between the measured data and the corresponding predicted values, including temperature residuals, flow residuals, and structural response residuals, and to standardize the coupled residuals. Combining the standardized residual persistence index and the cross-physics correlation index, the coupled residuals are decomposed into state-level residuals and parameter-level residuals. The residual attribution unit is used to determine the offset type of the parameter layer residual as thermal boundary offset, flow heat transfer offset or structural parameter offset based on the combined distribution characteristics of temperature residual, flow residual and structural response residual, and to construct the residual-parameter attribution mapping matrix. The dual-layer assimilation unit is used to correct the model's temperature field, flow field, and structural field within the current assimilation cycle using the state layer residuals; and to simultaneously correct the furnace-side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters based on the parameter layer residuals combined with the residual-parameter attribution mapping matrix. The physical consistency constraint unit is used to limit the obtained field correction results and model parameter correction results based on the rules of heat transfer consistency, flow continuity in the pipe, and structural response consistency; the coupled field state after the limit and the updated model parameters are used as the initial conditions for the next assimilation period simulation.

[0018] Compared with the prior art, the present invention has the following beneficial technical effects: This invention provides a method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall. By standardizing the residuals from the thermal, fluid, and solid-state coupling, and combining the persistence of the residuals with cross-physics field correlations, it achieves hierarchical decomposition of the residuals, accurately distinguishing between state-level and parameter-level residuals. This effectively solves the technical problems of missing multi-field coupling constraints, inability to hierarchically attribute deviations, and disordered parameter correction mechanisms in existing technologies. Temporary operating condition deviations such as short-term load fluctuations, instantaneous combustion skew, and flow disturbances are attributed to state-level residuals, used only to correct the real-time states of the temperature, flow, and structural fields in the current cycle. This avoids erroneous updates to model parameters due to short-term disturbance data, fundamentally preventing unreasonable parameter drift in the simulation model. Simultaneously, parameter-level residuals that persist long-term and conform to the thermal-fluid-solid coupling law are accurately attributed to slow-varying parameter offsets such as furnace thermal boundaries, in-tube heat transfer, flow resistance, and material properties. This enables long-term iterative updates of model parameters, solving the problem of long-term parameter offsets remaining as state errors in traditional technologies, and significantly improving the simulation model's ability to track the actual operating state of the water-cooled wall.

[0019] Furthermore, this invention constructs a residual-parameter attribution mapping matrix to match the correspondence between residuals and parameter offsets, and relies on the multiphysics field physical consistency rule to limit the amplitude of the correction results, thereby ensuring the physical rationality and continuity of the parameter correction and field correction process, and significantly improving the accuracy and long-term stability of the simulation prediction results of the water-cooled wall's full-field temperature, in-tube flow and thermal stress state. Attached Figure Description

[0020] Figure 1 This is a flowchart of the method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall, as described in this embodiment of the invention. Figure 2 This is a schematic diagram of the water-cooled wall tube section area, measuring point arrangement, and spatial neighborhood consistency in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the relationship between parameter layer residual attribution and parameter mapping in an embodiment of the present invention; Figure 4 This is a schematic diagram of the measured data assimilation cycle process in an embodiment of the present invention; Figure 5 This is a schematic diagram of the water-cooled wall thermal-fluid-solid multi-field coupling simulation and measured data assimilation system in an embodiment of the present invention. Detailed Implementation

[0021] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0022] The purpose of this invention is to provide a method and system for assimilating measured data from multi-field coupled simulation of water-cooled walls, in order to solve the technical problems of lack of multi-field coupling constraints, inability to attribute deviations hierarchically, and disordered parameter correction mechanisms in existing boiler water-cooled wall simulation analysis technology.

[0023] The present invention will now be described in further detail with reference to the accompanying drawings: Example 1 See Figure 1 In one embodiment of the present invention, a method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall is provided, comprising: Step 1: Map the measured data to the corresponding calculation nodes within the thermal-fluid-solid multiphysics coupling simulation model; Step 2: Run the thermal-fluid-solid multiphysics coupling simulation model and output the predicted temperature, flow state, and structural response values ​​that match the measured data. Step 3: Calculate the coupling residual between the measured data and the corresponding predicted value, including the temperature residual, flow residual, and structural response residual, and standardize the coupling residual; combine the standardized residual persistence index and the cross-physics correlation index to decompose the coupling residual into state layer residual and parameter layer residual. Step 4: Based on the combined distribution characteristics of temperature residual, flow residual, and structural response residual, determine whether the offset type corresponding to the parameter layer residual is thermal boundary offset, flow heat transfer offset, or structural parameter offset, and construct the residual-parameter attribution mapping matrix. Step 5: Use state layer residuals to correct the model temperature field, flow field, and structural field within the current assimilation cycle; based on parameter layer residuals combined with the residual-parameter attribution mapping matrix, simultaneously correct furnace side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters. Step 6: Based on the rules of heat transfer consistency, flow continuity in the pipe, and structural response consistency, limit the amplitude of the obtained field correction results and model parameter correction results; use the limited coupled field state and the updated model parameters as the initial conditions for the next assimilation cycle simulation.

[0024] In this embodiment, the water-cooled wall is divided into multiple pipe segment regions according to the furnace orientation, tube bank number, and elevation position. Each pipe segment region corresponds to a thermal field calculation node, a flow field calculation node, and a structural field calculation node.

[0025] Wall temperature or distributed fiber optic temperature data corresponds to the pipe wall thermal field node, pipe pressure, inlet flow rate and pipe section pressure drop data correspond to the flow field node, and local strain or expansion displacement data correspond to the structural field node.

[0026] For simulation nodes without direct measurement points, the measured data alignment unit determines its observation correlation range based on the relationship between adjacent measurement points in the same pipe row, adjacent elevation measurement points, and the heat load distribution of the furnace on the same side, rather than simply spreading the measured values ​​to all simulation nodes.

[0027] In this embodiment, let the first... Within the first assimilation cycle, the first The coupled simulation prediction state of each pipe segment region is as follows: The coupled simulation prediction state is determined by the local thermal field state. Local state of the flow field and local state of structural field constitute:

[0028] in, Including the Temperature status of one or more thermal field nodes within a pipe section area This includes the flow rate, pressure, or pressure drop status of the corresponding pipe section. This includes the thermal strain, expansion displacement, or equivalent thermal stress state of the corresponding structural nodes.

[0029] The predicted observations corresponding to the measured data are extracted from the coupled simulation prediction state and denoted as the predicted pipe wall temperature values. Predicted values ​​of flow state inside the pipe and structural response prediction values .

[0030] The corresponding observation state is obtained after aligning the measured data. It includes temperature observations Flow observation and structural response observations .

[0031] The residual decomposition unit first calculates the... The pipe section area in the first The original coupled residual vector within each assimilation period:

[0032] Since the temperature residual, flow residual, and structural response residual have different dimensions, the residual decomposition unit standardizes the original coupled residual vector before performing residual persistence judgment and cross-physics correlation judgment to obtain a standardized residual vector. :

[0033] in, , and These represent the normalized scales of the temperature residual, flow residual, and structural response residual, respectively. The normalized scales are determined based on the permissible deviation of the corresponding measurement, the standard deviation of the historical residual, or the calibration error of the simulation model.

[0034] In this embodiment, the three components of the standardized residual vector are denoted as follows: , and .

[0035] The standardized residual vector is used to calculate the residual persistence index, cross-physics correlation coefficient, and parameter layer allocation coefficient. The original coupling residual vector is used to perform state-level assimilation and parameter-level assimilation after the allocation ratio is determined, so that the assimilation correction amount retains the actual dimensions of the corresponding physical quantity.

[0036] State layer residuals are used to represent instantaneous deviations caused by load fluctuations, combustion skew, feedwater disturbances, or short-term flow distribution changes. The parameter layer residual is used to represent the persistent deviation caused by long-term shift of the furnace side thermal boundary, changes in heat transfer capacity inside the tube, changes in flow resistance, changes in tube wall material properties, or changes in the equivalent thickness of the tube wall.

[0037] The residual decomposition unit establishes a residual persistence index for each pipe segment area. It is used to determine whether the residual has a long-term stable trend.

[0038] The residual persistence index is based on standardized residual calculation:

[0039] in, Indicates the first The section area was recently Standardized residual mean over one assimilation period This represents the standardized residual vector of the previous assimilation period. To prevent extremely small constants with a denominator of zero.

[0040] In this embodiment, to avoid Excessive scaling leads to excessive residuals entering the parameter layer. The residual decomposition unit further adjusts the residuals based on a persistence normalization threshold. Calculate the normalized persistence index :

[0041] in, Determined based on the historical residual persistence distribution or model calibration conditions.

[0042] when When the value is low, it indicates that the residual changes rapidly and is not stable enough. The system will mainly classify the corresponding residual into the state layer residual. when When the value is high, it indicates that the residual persists and changes slowly over multiple assimilation cycles, and the system will classify a portion of the corresponding residual into the parameter layer residual.

[0043] In this embodiment, to avoid misjudging sensor drift, single-point noise, or local measurement point anomalies as model parameter shifts based solely on residual persistence, the residual attribution unit further calculates cross-physics correlation coefficients. .

[0044] The cross-physics correlation coefficient is used to describe whether the temperature residual, flow residual, and structural response residual conform to the thermo-fluid-solid coupling mechanism of the water-cooled wall.

[0045] Cross-physics correlation coefficients include temperature-structure consistency components. Temperature-flow consistency component and spatial neighborhood consistency components :

[0046] in, , and These are the weights of the corresponding components, and .

[0047] In this embodiment, the temperature-structure consistency component Used to determine whether the structural strain residual or equivalent thermal stress residual has a synchronous increasing trend when the temperature residual or temperature gradient residual increases; The temperature-flow consistency component Used to determine whether changes in wall temperature residual are accompanied by reasonable responses to flow rate, pressure, or pressure drop residuals; The spatial neighborhood consistency component It is used to determine whether the residual occurs continuously along the same tube bank, adjacent elevation, or same side of the furnace area.

[0048] The temperature-structure consistency component Determined based on the consistency of the directions of the temperature residual and the structural response residual.

[0049] In this embodiment, to avoid misjudgment caused by small noise, the residual attribution unit sets an effective temperature residual threshold. and structural effective residual threshold .

[0050] when or If this is the case, it indicates that the corresponding residuals are insufficient to support the temperature-structure correlation judgment. Take the preset low value; When both reach the effective residual threshold, the following formula is used to determine:

[0051] in, It is a symbolic function.

[0052] When the temperature residual and the structural response residual change in the same direction, it indicates that a higher temperature or a larger temperature gradient will be reflected in the structural response. Take the higher value; When the temperature residual and the structural response residual are in opposite directions, it indicates that they lack a consistent response. Take the lower value.

[0053] The temperature-flow consistency component Determined based on the consistency between the temperature residual and the direction of the flow anomaly.

[0054] The residual attribution unit sets a threshold for effective residual flow. ,when or If the corresponding residual is insufficient to support the temperature-flow correlation judgment, Take the preset low value; When both reach the effective residual threshold, the residual attribution unit sets the flow residual direction conversion function. This is used to convert different types of flow residuals into flow anomalies that are aligned with the direction of wall temperature rise.

[0055] Among them, when When representing the flow rate within the pipe, This is used to indicate the direction of wall temperature increase when the measured flow rate is lower than the simulated flow rate; when When indicating the pressure drop in a pipe section, This is used to indicate the direction of increase in flow resistance when the measured pressure drop is higher than the simulated pressure drop.

[0056] In this embodiment, the flow residual and pressure drop residual can be uniformly involved in the temperature-flow consistency judgment through the flow residual direction conversion function.

[0057] The spatial neighborhood consistency component According to the Determining the continuity of the residual direction between a single pipe segment and adjacent pipe segments:

[0058] in, Indicates the relationship with the first A collection of adjacent pipe sections located in the same pipe bank, at adjacent elevations, or on the same side of the furnace area. Indicates the number of adjacent pipe segments. This is an indicator function that takes the value when the condition is true. Otherwise take .

[0059] when At that time, the spatial neighborhood consistency component does not participate in the parameter layer enhancement judgment.

[0060] If the temperature residual in a certain pipe section is distributed in the same direction as the surrounding pipe sections, it indicates that the residual is more likely to be caused by the furnace heat load distribution, local heat transfer deterioration, or changes in structural constraint state. If only a single measurement point shows an isolated mutation A lower value is chosen to reduce the likelihood of it entering the parameter layer assimilation.

[0061] The residual attribution unit is based on the normalized persistence index. and cross-physics correlation coefficient Determine the parameter layer allocation coefficients :

[0062] This yields the standardized state layer residuals. and standardized parameter layer residuals :

[0063]

[0064] To preserve the original physical dimensions during state and parameter assimilation, the residual decomposition unit further constructs a scale reduction matrix. :

[0065] The standardized state-level residuals and standardized parameter-level residuals are then restored to the state-level residuals in their original dimensions. and parameter layer residuals :

[0066]

[0067] When the residual changes rapidly, the correlation across physical fields is weak, or the spatial neighborhood is discontinuous... The residuals are relatively small, and they mainly enter the state layer. When the residual persists, has strong cross-physical field correlation, and has continuity between adjacent pipe sections... The larger the residual, the more of it enters the parameter layer.

[0068] In this embodiment, through this decomposition method, the system distinguishes between short-term operating condition disturbances and long-term model offsets, avoiding writing instantaneous load fluctuations into model parameters and avoiding retaining long-term boundary offsets as state errors for a long time.

[0069] The state layer assimilation unit is used to quickly correct the temperature field, flow field, and structure field within the current assimilation cycle based on the state layer residual.

[0070] State assimilation does not change the long-term parameters of the model, but only modifies the state variables of the current period, so that the simulation results can follow the changes in actual operating conditions.

[0071] The state assimilation unit determines the state correction amount based on the measurement point reliability, residual amplitude, and physical sensitivity of the pipe section region:

[0072] in, Indicates the first State layer correction amount for each pipe section area This is the state assimilation mapping matrix, used to map the temperature residual, flow residual, and structural response residual to the state corrections of the corresponding thermal field nodes, flow field nodes, and structural field nodes.

[0073] When the The coupling state dimension of each pipe segment region is: hour, for A matrix with three rows and three columns.

[0074] The state assimilation mapping matrix is ​​determined based on the reliability of the measurement points and the coupling sensitivity of the pipe section area; When the temperature residual in this region will significantly cause changes in thermal stress, or the flow residual will significantly affect the heat transfer of the pipe wall, the corresponding mapping weight in the state assimilation mapping matrix will increase. When a measurement point experiences a jump, packet loss, or significant inconsistency with adjacent measurement points, the corresponding mapping weight in the state assimilation mapping matrix is ​​reduced.

[0075] The candidate coupled states after state layer assimilation are updated as follows:

[0076] in, This represents the current candidate coupling state after state layer assimilation.

[0077] For the temperature field, state layer assimilation mainly corrects the current pipe wall temperature distribution and local temperature gradient; For the flow field, state layer assimilation mainly corrects the current flow distribution and pressure distribution within the pipe; For structural fields, state layer assimilation primarily corrects the current thermal strain and equivalent thermal stress response.

[0078] The corrections for the three states are not performed in isolation. After the temperature state is updated, it is synchronously transferred to the thermal expansion calculation. After the flow state is updated, it is synchronously transferred to the heat transfer boundary inside the pipe. After the structural state is updated, it is used to verify whether the stress concentration corresponding to the temperature gradient is reasonable.

[0079] The parameter layer assimilation unit is used to update the parameters of the slowly varying model based on the parameter layer residuals.

[0080] The slowly varying model parameters include at least the furnace-side thermal boundary correction coefficient. Heat transfer correction factor inside the tube Flow resistance correction factor and material property correction factor .

[0081] in, Used to correct the The deviation of the actual heat intensity on the furnace side of each pipe section from the design heat load. Used to correct deviations in the heat transfer capacity of the tube from the design value. Used to correct for pressure drop or resistance characteristic deviations within the pipe. Used to correct the effects of changes in the thermal conductivity, elastic response, or equivalent wall thickness of the pipe wall material on thermal stress calculations.

[0082] Let the first The first section of the pipeline area The parameter vector for each assimilation period is :

[0083] The parameter layer assimilation adopts a slow update method to avoid accidentally writing short-term fluctuations into the model parameters.

[0084] The residual attribution unit determines the parameter correction direction based on the combined characteristics of temperature residual, flow residual, and structural response residual.

[0085] When the When the temperature residual in a pipe section is consistently positive, and adjacent pipe sections exhibit temperature residuals in the same direction, while the flow residual and pressure drop residual are insignificant, it indicates that the simulation model underestimates the furnace-side heating intensity in this region. The parameter layer assimilation unit should prioritize increasing the furnace-side thermal boundary correction coefficient. .

[0086] When the temperature residual is consistently positive, while the flow residual is negative or the pressure drop residual is positive, it indicates that the flow or heat transfer capacity within the pipe in that region has shifted relative to the simulation model. The parameter layer assimilation unit will preferentially reduce the heat transfer correction factor within the pipe. Or increase the flow resistance correction factor .

[0087] When the temperature residual is not significant but the structural response residual persists, or when the measured strain is higher than the simulated strain under the same temperature gradient, it indicates that the structural response parameters do not match the field conditions. In this case, the parameter layer assimilation unit should prioritize correcting the material property correction factor. .

[0088] When there is no physical correspondence between the temperature residual, flow residual, and structural response residual as described above, the system does not perform parameter layer updates, or only performs state layer updates after limiting.

[0089] In the specific attribution process, the residual attribution unit establishes thermal boundary offset criteria, flow heat transfer offset criteria, and structural parameter offset criteria.

[0090] The thermal boundary offset criterion is used to identify the offset of the furnace side heat load distribution relative to the simulation set value. The conditions that must be met are: the temperature residual remains in the same direction in multiple consecutive assimilation cycles, is continuously distributed in space along the same side furnace area or adjacent pipe sections, and the flow residual does not reach the flow anomaly threshold.

[0091] The flow heat transfer offset criterion is used to identify offsets in pipe flow distribution, pressure drop, or heat transfer capacity. It meets the following conditions: the temperature residual is consistently high, while the pipe flow is low, the pressure drop is high, or the outlet steam-water parameters deviate from the simulation value.

[0092] The structural parameter offset criterion is used to identify material properties, equivalent wall thickness, or constraint state offsets. It satisfies the following conditions: the temperature residual and flow residual are insufficient to explain the structural response residual, and the structural response residual repeats under the same temperature gradient.

[0093] The residual attribution unit generates the first [value] based on the above criteria. Parameter attribution vector for each pipe segment region :

[0094] in, , , and These represent the proportions of parameter layer residuals attributed to the furnace-side thermal boundary, in-tube heat transfer, flow resistance, and material properties, respectively.

[0095] The parameter attribution vector satisfies:

[0096] When the residual can be clearly attributed, at least one attribution proportion is greater than zero. When the residuals manifest as isolated measurement point anomalies or cross-physical field relationships not being valid, the proportions of each attribution are reduced.

[0097] The parameter attribution vector Instead of being multiplied directly with the parameter layer residual vector, it is used to construct the residual-parameter attribution mapping matrix. ,Depend on The temperature residual, flow residual, and structural response residual are mapped to the furnace-side thermal boundary correction factor, tube-side heat transfer correction factor, flow resistance correction factor, and material property correction factor.

[0098] Combination Figure 3 The residual-parameter attribution mapping matrix It is a 4x3 matrix, with each row corresponding to... , , and The correction directions are respectively corresponding to the temperature residual, flow residual, and structural response residual.

[0099] The residual-parameter attribution mapping matrix is ​​represented as follows:

[0100] in, This represents the correction factor for the temperature residual on the furnace-side thermal boundary. Mapping weights, and These represent the correction factors for heat transfer inside the tube due to temperature residual and flow residual, respectively. Mapping weights, This represents the correction factor for flow residuals with respect to flow resistance. Mapping weights, This represents the correction factor for material properties based on the structural response residuals. Mapping weights.

[0101] The remaining mapping weights are set to zero or a limit value based on the residual attribution results.

[0102] In this embodiment, the residual-parameter attribution mapping matrix Based on parameter attribution vector The residual attribution criterion was determined.

[0103] When the thermal boundary offset criterion is met, set and will , and Setting the mapping weight to zero or below the limiting threshold with a weak mapping weight ensures that the temperature residual is primarily written into the furnace-side thermal boundary correction factor. .

[0104] When the flow heat transfer offset criterion is met, set , , and reduce and This allows the temperature residual and flow residual to be primarily incorporated into the in-tube heat transfer correction factor. and flow resistance correction factor .

[0105] When the structural parameter offset criterion is met, set Furthermore, the mapping weights of temperature and flow residuals to other parameters are reduced, so that the structural response residuals are primarily written into the material property correction coefficients. .

[0106] When any criterion is not met, the mapping weight corresponding to that criterion is reset to zero or reduced to the limit value.

[0107] By using the matrix construction method described above, the three types of residuals can be selectively mapped to the four types of slowly varying model parameters, thus avoiding the indiscriminate writing of residuals from different sources into all model parameters.

[0108] The parameter layer assimilation unit is based on the residual-parameter attribution mapping matrix. and parameter layer residuals Calculate the candidate correction amount for the parameter:

[0109] in, Indicates the first Candidate parameter correction values ​​for each pipe section area.

[0110] The fourth-order diagonal parameter assimilation gain matrix is:

[0111] in, , , and These are the parameter assimilation gains corresponding to the furnace-side thermal boundary, in-tube heat transfer, flow resistance, and material properties, respectively.

[0112] thus, It is a 4x4 matrix. It is a 4x3 matrix. Given a three-dimensional column vector, the calculated... It is a four-dimensional column vector, and a parameter vector. The dimensions are consistent.

[0113] The parameter assimilation gain is determined based on the historical assimilation error, the allowable range of parameter changes, and the stability of the simulation model. Its effect intensity is less than the state update intensity corresponding to the state assimilation mapping matrix, so that the parameters of the slowly varying model have a lower update rate relative to the current state variables.

[0114] Slow update coefficient of parameters Used to control the speed at which candidate parameter corrections are written into the parameters of the slowly varying model, and .

[0115] The candidate parameter vector after parameter layer assimilation is:

[0116] When a residual occurs only within a single assimilation period, even a large residual magnitude will not directly lead to... Significant changes; Parametric layer assimilation only gradually changes the corresponding parameters when the residual persists across multiple cycles and is consistent with the physical correlation between temperature, flow, and structural response.

[0117] The physical consistency constraint unit is used to verify the candidate results after state layer assimilation and parameter layer assimilation.

[0118] The verification content should include at least the consistency of heat transfer, the continuity of flow inside the pipe, and the consistency of structural response.

[0119] Heat transfer consistency is used to determine whether there is an unreasonable energy difference between the corrected furnace-side heating, tube wall heat conduction and tube internal heat exchange; The continuity of flow within the pipe is used to determine whether the corrected flow distribution and pressure distribution satisfy the mass conservation and pressure drop relationship of the same loop. Structural response consistency is used to determine whether the corrected temperature gradient corresponds to the location of thermal stress concentration.

[0120] Physical consistency constraint unit calculation Overall consistency deviation of each pipe section area :

[0121] in, This indicates a deviation in the consistency of heat transfer. Indicates a deviation in flow continuity. Indicates structural response consistency deviation. , and These are the corresponding weights.

[0122] when When the error exceeds the consistency threshold, it indicates that although the current assimilation correction may have reduced the local measurement point residuals, it has disrupted the thermal-fluid-structure interaction. The system then determines the consistency limiting coefficient based on the consistency deviation. ,and .

[0123] When the consistency deviation is larger The smaller the value, the better, to prevent unreasonable model drift caused by local abnormal measurement points or incorrect residual attribution.

[0124] The physical consistency constraint unit determines the final state correction and the final parameter correction based on the consistency limiting coefficient:

[0125]

[0126] in, This represents the final state correction amount after consistency limiting. This represents the final parameter correction amount after the combined effect of the slow update coefficient and the consistency limiting coefficient.

[0127] thus, This represents the candidate correction amount for parameters that have not been multiplied by the slow update coefficient. This indicates the parameter correction amount after a slow update. This indicates the final parameter correction amount after consistency limiting.

[0128] After physical consistency constraints, the system obtains the first... The final coupling state and final model parameters for each assimilation cycle:

[0129]

[0130] And use this as the initial condition for the next assimilation period:

[0131]

[0132] in, This represents the coupled state after state assimilation and physical consistency verification. This represents the model parameters after parameter assimilation and physical consistency verification.

[0133] Therefore, short-term operational disturbances are mainly absorbed through the state layer, while long-term model offsets are mainly absorbed through the parameter layer, and the two are not confused with each other.

[0134] Combination Figure 4 During runtime, the following process is executed for each assimilation cycle: First, read the measured data of the water-cooled wall and complete the correspondence between the measuring points and the simulation nodes; Then, the thermo-fluid-solid multiphysics coupling simulation model is run to obtain the predicted temperature, flow state, and structural response for the current cycle; Next, the original coupling residuals between the measured data and the simulation prediction results are calculated, and the residuals of different dimensions are standardized. Subsequently, based on the persistence, temperature-structure consistency, temperature-flow consistency, and spatial neighborhood consistency of the standardized residuals, the parameter layer allocation coefficients are determined, and the residuals are decomposed into state layer residuals and parameter layer residuals. Then, the attribution direction of the parameter layer residual is determined by the thermal boundary offset criterion, the flow heat transfer offset criterion, and the structural parameter offset criterion, and a residual-parameter attribution mapping matrix is ​​constructed based on the parameter attribution vector. Then, the current coupling state is corrected by using the state layer residuals through the state assimilation mapping matrix, and the parameter layer residuals are used to form parameter candidate correction quantities through the residual-parameter attribution mapping matrix and the parameter assimilation gain matrix; Finally, the candidate correction results are verified for thermal, flow, and structural consistency, and the final coupling state and final model parameters are determined based on the consistency limiting coefficient.

[0135] Example 2 This embodiment also provides a system for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall, including: The measured data alignment unit is used to collect on-site measured temperature data of the water-cooled wall, flow data inside the tube, and structural response data, and to map the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model. The coupled simulation prediction unit is used to run the thermo-fluid-solid multiphysics coupled simulation model and output temperature prediction, flow state prediction, and structural response prediction values ​​that match the measured data. The residual decomposition unit is used to calculate the coupled residuals between the measured data and the corresponding predicted values, including temperature residuals, flow residuals, and structural response residuals, and to standardize the coupled residuals. Combining the standardized residual persistence index and the cross-physics correlation index, the coupled residuals are decomposed into state-level residuals and parameter-level residuals. The residual attribution unit is used to determine the offset type of the parameter layer residual as thermal boundary offset, flow heat transfer offset or structural parameter offset based on the combined distribution characteristics of temperature residual, flow residual and structural response residual, and to construct the residual-parameter attribution mapping matrix. The dual-layer assimilation unit is used to correct the model's temperature field, flow field, and structural field within the current assimilation cycle using the state layer residuals; and to simultaneously correct the furnace-side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters based on the parameter layer residuals combined with the residual-parameter attribution mapping matrix. The physical consistency constraint unit is used to limit the obtained field correction results and model parameter correction results based on the rules of heat transfer consistency, flow continuity in the pipe, and structural response consistency; the coupled field state after the limit and the updated model parameters are used as the initial conditions for the next assimilation period simulation.

[0136] The system is used to introduce on-site data such as wall temperature of the water-cooled wall, distributed fiber temperature, tube pressure, inlet flow rate, outlet steam and water parameters, and local strain into the thermo-fluid-solid multiphysics coupling simulation model during boiler operation, so that the simulation model can perform layered corrections on the current coupling state and slowly varying model parameters based on the measured data.

[0137] The current coupling state represents the temperature field, the flow state inside the tube, and the structural response state within the current assimilation cycle. The slowly varying model parameters represent the long-term shift of the furnace-side thermal boundary, the heat transfer capacity inside the tube, the flow resistance, and the tube wall material response relative to the initial simulation model.

[0138] Combination Figure 2 The measured data alignment unit is used to establish the correspondence between the measured data and the water-cooled wall simulation model.

[0139] In summary, this invention does not simply use measured data of water-cooled walls to correct simulation results. Instead, based on the persistence and cross-physics correlation of standardized residuals, it decomposes the coupling residuals between measured data and simulation predictions into state-level residuals and parameter-level residuals. This ensures that short-term load fluctuations, combustion bias, or instantaneous flow disturbances primarily affect the state correction of the current temperature, flow, and structural fields. Meanwhile, long-term residuals that conform to the thermal-fluid-structure interaction are further attributed to slowly varying model parameters such as furnace-side thermal boundaries, tube heat transfer, flow resistance, or material properties. This avoids inadvertently incorporating short-term operational disturbances into model parameters, reducing unreasonable model drift. It also prevents long-term boundary offsets or parameter degradation from being retained as state errors, improving the water-cooled wall thermal-fluid-structure interaction simulation model's ability to track on-site operating conditions and its long-term predictive stability. By using the residual-parameter attribution mapping matrix and physical consistency constraints, the system can also maintain physical consistency between heat transfer, flow continuity within the tube, and structural response during the parameter correction process, thereby improving the reliability of the prediction results for the whole-field temperature distribution of the water-cooled wall, the flow state within the tube, and the thermal stress state.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for assimilating measured data from multi-field coupled thermal-fluid-solid-structure interaction simulations of water-cooled walls, characterized in that, include: Collect on-site measured temperature data of the water-cooled wall, flow data inside the tube, and structural response data, and map the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model; Run the thermo-fluid-solid multiphysics coupled simulation model and output the temperature prediction, flow state prediction, and structural response prediction values ​​that match the measured data. Calculate the coupled residuals between the measured data and the corresponding predicted values, including temperature residuals, flow residuals, and structural response residuals, and standardize the coupled residuals. Combining standardized residual persistence index and cross-physics correlation index, the coupled residual is decomposed into state layer residual and parameter layer residual. Based on the combined distribution characteristics of temperature residual, flow residual, and structural response residual, the offset type corresponding to the parameter layer residual is determined to be thermal boundary offset, flow heat transfer offset, or structural parameter offset, and a residual-parameter attribution mapping matrix is ​​constructed. The temperature field, flow field, and structure field of the model within the current assimilation cycle are corrected using the state layer residuals; Based on the parameter layer residual combined with the residual-parameter attribution mapping matrix, the furnace side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material property parameters are simultaneously corrected. Based on the rules of heat transfer consistency, flow continuity inside the pipe, and structural response consistency, the obtained field correction results and model parameter correction results are subject to amplitude constraints. The coupled field state after amplitude limiting and the updated model parameters are used as the initial conditions for the next assimilation period simulation.

2. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The specific process of mapping the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model is as follows: divide the independent pipe segment regions according to the orientation of the water-cooled furnace, the pipe row number, and the elevation, and associate each pipe segment region with the thermal field, flow field, and structural field calculation nodes respectively; temperature data is matched to the thermal field calculation node, the flow data inside the pipe is matched to the flow field calculation node, and the structural response data is matched to the structural field calculation node.

3. The method for assimilating measured data from a multi-field coupled simulation of a water-cooled wall's thermal-fluid-solid dynamics as described in claim 1, characterized in that... The coupling residual standardization process is as follows: temperature residual, flow residual, and structural response residual are integrated into a coupling residual vector, and the coupling residual vector is standardized according to the normalization scale of each of the three types of residuals.

4. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The acquisition and stratification rules for the standardized residual persistence index include extracting the mean of standardized residuals from multiple consecutive assimilation periods and calculating the residual persistence index from the changes in standardized residuals in adjacent periods. Based on the residual persistence index, the proportion of coupled residuals allocated to the parameter layer residuals is determined.

5. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The cross-physics correlation includes temperature-structure consistency, temperature-flow consistency, and spatial neighborhood consistency. Temperature-structure consistency characterizes the matching relationship between temperature residuals and structural response residuals, temperature-flow consistency characterizes the matching relationship between temperature residuals and flow anomalies within the pipe, and spatial neighborhood consistency characterizes the continuity of residual distribution between adjacent pipe segments.

6. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 5, characterized in that, Before performing consistency determination, an effective residual threshold is set; if the standardized residual does not reach the effective residual threshold, the weight of the corresponding consistency index on the parameter layer residual allocation is reduced.

7. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The method of determining the offset type of parameter layer residual as thermal boundary offset, flow heat transfer offset, or structural parameter offset based on the combined distribution characteristics of temperature residual, flow residual, and structural response residual specifically includes: When the temperature residual is continuously distributed in the same direction along the same side of the furnace or adjacent pipe sections, and the flow residual does not reach the flow anomaly threshold, it is determined that there is a thermal boundary offset. When the temperature residual is consistently high, and the flow rate in the pipe is low or the pressure drop in the pipe is high, it is determined that there is a flow heat transfer deviation. When temperature and flow residuals cannot explain the recurring structural response residuals, it is determined that there is a structural parameter offset.

8. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The constructed residual-parameter attribution mapping matrix is ​​a four-row, three-column matrix; the four rows of the matrix correspond to the correction directions of furnace side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters, respectively, and the three columns of the matrix correspond to temperature residuals, flow residuals, and structural response residuals, respectively.

9. The method for assimilating measured data from a multi-field coupled thermal-fluid-solid simulation of a water-cooled wall according to claim 1, characterized in that, The method of simultaneously correcting furnace-side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material property parameters based on parameter layer residuals combined with residual-parameter attribution mapping matrices includes: When a thermal boundary offset is determined to exist, the mapping weight of the thermal boundary parameter pointing to the furnace side of the temperature residual is increased; When a flow heat transfer offset is determined to exist, the mapping weights of temperature, flow residual pointing to in-tube heat transfer parameters, and flow resistance parameters are increased. When a structural parameter offset is determined to exist, the mapping weight of the structural response residual to the material performance parameters is increased.

10. A system for assimilating measured data from a multi-field coupled thermal-fluid-solid-structure interaction simulation of a water-cooled wall, characterized in that, include: The measured data alignment unit is used to collect on-site measured temperature data of the water-cooled wall, flow data inside the tube, and structural response data, and to map the measured data to the corresponding calculation nodes inside the thermal-fluid-solid multiphysics coupling simulation model. The coupled simulation prediction unit is used to run the thermo-fluid-solid multiphysics coupled simulation model and output temperature prediction, flow state prediction, and structural response prediction values ​​that match the measured data. The residual decomposition unit is used to calculate the coupled residual between the measured data and the corresponding predicted value, including temperature residual, flow residual and structural response residual, and to standardize the coupled residual. Combining standardized residual persistence index and cross-physics correlation index, the coupled residual is decomposed into state layer residual and parameter layer residual. The residual attribution unit is used to determine the offset type of the parameter layer residual as thermal boundary offset, flow heat transfer offset or structural parameter offset based on the combined distribution characteristics of temperature residual, flow residual and structural response residual, and to construct the residual-parameter attribution mapping matrix. A two-layer assimilation unit is used to correct the model's temperature field, flow field, and structural field within the current assimilation cycle using state layer residuals. Based on the parameter layer residual combined with the residual-parameter attribution mapping matrix, the furnace side thermal boundary parameters, tube heat transfer parameters, flow resistance parameters, and material performance parameters are simultaneously corrected. The physical consistency constraint unit is used to limit the amplitude of the obtained field correction results and model parameter correction results based on the rules of heat transfer consistency, flow continuity in the pipe, and structural response consistency. The coupled field state after amplitude limiting and the updated model parameters are used as the initial conditions for the next assimilation period simulation.

Images