Multi-source joint inversion method and system for adaptive activation of combustion field function partition and physical closure correction

Abstract

The present application belongs to the technical field of combustion diagnosis and physical field inversion, and provides a multi-source joint inversion method and system for adaptive activation of combustion field function partition and correction of physical closure, which obtains multi-source observation data to generate an initial inversion result, and adaptively activates a target function partition corresponding to the current iteration step from a pre-constructed function partition library according to the combustion field state characteristics, wherein different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data for different target function partitions; observation response residuals and physical closure residuals are constructed for each partition to identify the dominant mismatch type; accordingly, constraint switching and closed-loop correction are performed locally to update the inversion field; a double convergence criterion is judged, and if it is satisfied, the result and diagnostic information are output, and if it is not satisfied, iteration is continued. The present application realizes partition-level accurate correction and improves the inversion accuracy and robustness.

Description

Technical Field

[0001] This invention relates to the field of combustion diagnostics and physical field inversion technology, specifically to a multi-source joint inversion method and system for adaptive activation and physical closure correction of combustion field functional zoning. Background Technology

[0002] In aero-engine combustors and other high-temperature reactive flow combustion devices, the combustion process is typically accompanied by the strongly coupled evolution of multiple physical quantities, including temperature, pressure, composition, heat release, and flow structure. The spatial distribution and interrelationships of these physical quantities not only affect combustion efficiency, emission characteristics, and heat load levels, but also directly relate to combustion stability, component life assessment, and combustion organization optimization. Therefore, accurately acquiring combustion field state information within complex and confined spaces has always been a key technical challenge in combustion mechanism research, combustion system optimization design, and operational status diagnosis.

[0003] Because combustion chambers are typically characterized by high temperature, high pressure, transients, and strong turbulence, a single observation method often struggles to simultaneously achieve adequate spatial coverage, measurement accuracy, and response speed. In practical research and engineering applications, it is usually necessary to comprehensively utilize multiple observation methods to acquire combustion field information. For example, path integral optical observations can obtain absorption or radiation response information along the optical path; wall temperature measurements can reflect local thermal boundary responses; pressure measurements can reflect combustion-induced flow and thermoacoustic characteristics; and outlet temperature measurements can characterize the overall thermal distribution. Historical operating data or low-order flow prior information can also be combined to provide auxiliary constraints on the state to be reconstructed. The observation data from these different sources usually exhibit significant differences in measurement principles, spatial resolution, temporal resolution, sensitive regions, and noise characteristics. Therefore, combustion field state reconstruction often requires a multi-source joint inversion approach.

[0004] In existing technologies, multi-source combustion field reconstruction methods can generally be categorized as follows: one type is based on inversion using a single observation model, followed by empirical correction of the results; another type incorporates multiple source observations into a single objective function, employing weighted least squares, regularization, or statistical estimation methods for joint solution; and yet another type introduces prior fields, smoothing terms, or general physical constraints to improve the stability of the reconstruction process. While these methods can mitigate the information deficiency caused by single observations under certain conditions, they still exhibit the following significant limitations in complex combustion scenarios:

[0005] First, existing joint inversion methods mostly employ uniform observation fusion rules and constraints across the entire field, typically assuming that different regions can apply the same residual construction method, the same regularization mechanism, or the same weight adjustment strategy. However, actual combustion fields often exhibit significant mechanistic inhomogeneities in space. Different locations may be dominated by mixing and heating processes, strong reaction and heat release processes, recirculation stabilization processes, near-wall heat transfer processes, or subsequent decay processes. The sensitivity of different regions to different observation information is not consistent, and the physical relationships they should satisfy also differ. In this case, if uniform inversion rules are still used across the entire field, it is easy to encounter problems such as local regions fitting observations but violating physical laws, or local conflicts being diffused into global biases.

[0006] Second, existing multi-source inversion methods typically focus more on the matching degree between observation data, while insufficiently considering the differences in physical closure relationships across different regions. In a combustion field, even if a reconstruction result is relatively close to some observation data, there may still be inconsistencies in local thermal states and reaction intensities, inconsistencies between near-wall thermal responses and wall boundary conditions, and incompatibility between the hotspot residence location in the recirculation zone and the flame stabilization structure. If minimizing observation errors is taken as the primary optimization objective, it is often difficult to ensure that the obtained results simultaneously meet the local physical constraints that a combustion field should have, thus affecting the reliability and interpretability of the inversion results.

[0007] Third, existing methods, when faced with local mismatches, often employ global weighting, global smoothing enhancement, or global re-solution, lacking local diagnosis and correction mechanisms specific to the mismatch region and type. In reality, error sources in the combustion field exhibit clear regional and typological characteristics. For example, the main problem in some areas may stem from insufficient observation coverage, while others may primarily manifest as thermochemical closure mismatches, and still others may exhibit wall thermal boundary mismatches or flame stabilization structure mismatches. If the main sources of mismatch cannot be identified and targeted constraint switching and closed-loop corrections implemented locally, the reconstruction process is prone to slow convergence, persistent local distortions, or the propagation of local errors to other areas.

[0008] Therefore, for multi-source joint inversion tasks of aero-engine combustion chambers and other high-temperature reactive flow combustion devices, there is an urgent need to propose a new combustion field reconstruction method. Summary of the Invention

[0009] To address the technical problems of existing multi-source joint inversion methods for combustion fields under complex high-temperature reaction flow conditions, such as insufficient adaptability of the whole-field unified constraint, difficulty in identifying local physical mismatches, easy propagation of local errors to the global, and insufficient physical reliability of reconstruction results, this invention discloses a multi-source joint inversion method and system for adaptive activation and physical closure correction of combustion field functional zones.

[0010] This method is designed for aero-engine combustors and other combustion devices with confined spaces, high-temperature reaction flows, and multi-source heterogeneous observation conditions. Based on the correspondence between multi-source observation information and the local dominant mechanism of the combustion field, it no longer adopts a uniform observation fusion rule and a uniform constraint form for the entire field. Instead, it adaptively determines functional zones according to the combustion process characteristics of different regions, and constructs observation constraints and physical closure constraints for each functional zone. Specifically, for different target functional zones, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. Based on the identification of local dominant mismatch types, local constraint switching and closed-loop correction are implemented to achieve stable joint inversion of the combustion field.

[0011] Specifically, the technical solution for achieving the objective of this invention is as follows: In a first aspect, this invention discloses a multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning, the method comprising the following steps: S1. Generate initial inversion results based on multi-source observation data of the combustion field to be measured; S2. Based on the combustion field state characteristics of the initial inversion results, adaptively activate the target functional partition corresponding to the current iteration step from the pre-constructed functional partition library; S3. For each activated target functional partition, construct observation response residuals and physical closure residuals respectively. For different target functional partitions, configure different observation response residual construction methods and / or constraint weights for different observation sources in the multi-source observation data. S4. Based on the observed response residual and the physical closure residual, identify the dominant mismatch type of each target functional partition; S5. Based on the dominant mismatch type, perform constraint switching and closed-loop correction within the local range of the corresponding target functional partition, and update the inversion field; S6. Determine whether the updated inversion field satisfies the preset dual convergence criterion. If it does, output the final inversion result and diagnostic information. If it does not, use the updated inversion field as the new initial inversion result for iteration.

[0012] In one embodiment of step S1, the multi-source observation data includes at least two of the following: path integral optical observation data, wall temperature measurement data, pressure measurement data, outlet temperature distribution data, and at least two of the following: historical operating condition data, low-order flow prior information, and numerical simulation prior information.

[0013] In one embodiment of step S2, based on the combustion field state characteristics of the initial inversion result, the target functional partition corresponding to the current iteration step is adaptively activated from the pre-built functional partition library, including: S21. Extract the temperature gradient distribution, reflux prior index, and wall distance parameter from the initial inversion results as combustion field state features; S22. The temperature gradient distribution, the reflux prior index, and the wall distance parameter are compared with the activation threshold of each candidate partition in the functional partition library. S23. The candidate partitions that meet the activation threshold are determined as the target functional partitions of the current iteration step. The target functional partitions include one or more of the following: low reaction intensity mixing and heating zone, main reaction zone, reflux stabilization zone, and near-wall heat exchange zone.

[0014] In one embodiment of step S3, for each activated target functional partition, the observed response residual and the physical closure residual are constructed, including at least one of the following: For the main reaction zone in the target functional area, a thermochemical consistency residual is constructed to constrain the deviation between the normalized temperature rise and the equivalent reaction intensity index, and an energy closure residual is constructed to constrain the conservation relationship between the convection transport term, the heat conduction term and the equivalent heat source term. For the reflux stabilization zone in the target functional partition, a flame stabilization consistency residual is constructed to constrain the Euclidean distance between the hot spot residence centroid coordinates and the reflux core location. For the near-wall heat transfer zone in the target functional area, a near-wall heat flux closure residual is constructed to constrain the Fourier thermal conductivity relationship between the wall equivalent heat flux and the near-wall normal temperature gradient. For the low-reaction-intensity mixing and heating zone in the target functional area, a physical closed residual of the low-reaction-intensity mixing and heating zone is constructed to constrain the weighted sum of the upper limit of reaction intensity and the upper limit of temperature.

[0015] Furthermore, the formula for calculating the thermochemical consistency residual of the main reaction zone is: ; in, Thermochemical consistency residuals in the main reaction zone Main reaction zone Number of units inside, For the first Normalized temperature rise per unit, For the first Local reaction intensity index of each unit; The formula for calculating the energy closure residual of the main reaction zone is: ; in, The energy closure residual of the main reaction zone, For the first Fluid density within each unit; For the first The local velocity vector of each element or its prior value; For the first Specific enthalpy of each unit; Indicates the first Thermal conductivity of each unit; This is the equivalent heat source term.

[0016] Furthermore, the formula for calculating the flame stabilization consistency residual in the reflow stabilization region is as follows: ; in, For the flame stabilization consistency residual in the reflux stabilization region, The coordinates of the centroid of the high-temperature, high-reaction zone within the reflux stability region. The location of the reflux core is determined by flow priors or low-velocity reflux. The formula for calculating the near-wall heat flux closure residual in the near-wall heat transfer zone is: ; in, For the near-wall heat flux closure residual in the near-wall heat transfer region, Near-wall heat exchange zone The number of units, For the first Thermal conductivity of the near-wall unit For the first Temperature gradient of near-wall elements along the wall normal This represents the equivalent heat flux of the wall.

[0017] Furthermore, the formula for calculating the physical closure residual of the low-reaction-intensity mixing heating zone is as follows: ; in, The physical closure residual of the low-reaction-intensity mixing and heating zone. For partitioning Number of internal units; This represents the upper limit of the permissible reaction intensity within the low-reaction-intensity mixing and heating zone; This represents the upper limit of the allowable temperature for the low-reaction-intensity mixing heating zone; This represents the weighting coefficient for the temperature term.

[0018] In one embodiment of step S4, based on the observed response residuals and the physical closure residuals, the dominant mismatch type of each target functional partition is identified, including: S41, regarding the first Within the target functional area Class residual is The normalized exceedance index is calculated according to the following formula: ； in, For the corresponding criterion threshold, To prevent tiny positive numbers with a denominator of zero; S42. Based on the normalized out-of-limit index, the dominant mismatch type index is determined using the following formula: ; In the formula, Indicates the first The dominant mismatch type index for each target functional partition.

[0019] In one embodiment of step S6, the preset dual convergence criterion includes: The global convergence criterion is that the relative change in the inversion field between two adjacent iterations is less than the preset global convergence threshold. The partition convergence criterion is that the physical closure residuals of all activated target function partitions are lower than the preset partition convergence threshold. The system is deemed to satisfy the dual convergence criterion only when both the global convergence criterion and the partition convergence criterion are satisfied.

[0020] Secondly, the present invention discloses a multi-source joint inversion system for adaptive activation and physical closure correction of combustion field functional zoning. The system includes one or more processors and a memory storing instructions. When the instructions are executed by the one or more processors, the system performs the above-mentioned method. The memory includes a data acquisition module, an initial inversion module, a functional zoning activation module, a zoning residual construction module, a dominant mismatch identification module, a local constraint switching and closed-loop correction module, and a convergence judgment module.

[0021] The data acquisition module is used to acquire multi-source observation data of the combustion field to be tested. The initial inversion module is used to generate initial inversion results based on the multi-source observation data; The functional partition activation module is used to adaptively activate the target functional partition corresponding to the current iteration step from the pre-built functional partition library based on the combustion field state characteristics of the initial inversion result. The partition residual construction module is used to construct observation response residuals and physical closure residuals for each activated target functional partition. For different target functional partitions, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. The dominant mismatch identification module is used to identify the dominant mismatch type of each target functional partition based on the observed response residual and the physical closure residual; The local constraint switching and closed-loop correction module is used to perform constraint switching and closed-loop correction within the local range of the corresponding target functional partition according to the dominant mismatch type, and update the inversion field; The convergence judgment module is used to determine whether the updated inversion field meets the preset dual convergence criteria. If it does, the final inversion result and diagnostic information are output. If it does not, the updated inversion field is used as the new initial inversion result for iteration.

[0022] Compared with existing multi-source observation fusion and joint inversion methods for combustion fields, the multi-source joint inversion method and system for combustion field functional zoning adaptive activation and physical closure correction proposed in this invention has the following advantages: 1. This invention implements zoned inversion control to address differences in the dominant mechanisms of the combustion field, enhancing the relevance of complex combustion field reconstruction. Existing methods typically employ uniform observation fusion rules, residual construction methods, or regularization processing across the entire field, which struggles to adapt to the mechanistic differences in mixing and heating, strong reactions, flame stabilization, near-wall heat transfer, and subsequent attenuation in different regions. This invention constructs a functional zone library and adaptively activates corresponding zones based on the current combustion object and state characteristics. This allows each region to adopt constraint forms and correction strategies that match its dominant processes, avoiding the masking of local mechanistic differences by uniform processing methods and significantly enhancing the adaptability to combustion field non-uniformity.

[0023] 2. Balancing observational fitting capability with local physical reliability enhances the physical consistency of joint inversion results. Existing schemes often focus on minimizing the error between reconstructed results and observational data, neglecting the physical relationships that should be satisfied between local thermal states, reaction intensities, flame stabilization structures, and boundary thermal responses. This can easily lead to situations where observational fitting is good but physical closure is unreasonable. This invention constructs observational response residuals and physical closure residuals separately within each functional zone. This ensures that the reconstructed results not only satisfy multi-source observation constraints but also better conform to the dominant mechanistic characteristics of the corresponding region, thereby improving the physical interpretation capability and engineering reliability of the results.

[0024] 3. Identifying dominant local mismatch types enables a shift from "overall error processing" to "mismatch mechanism diagnosis." Existing technologies, when reconstruction biases occur, typically only perform overall weighting, smoothing, or recalculation based on the magnitude of the overall error, making it difficult to distinguish the specific sources of local biases (such as underfitting of observations, thermochemical closure mismatch, flame-stabilized structure mismatch, or boundary thermal response mismatch). This invention, by comparing the degree to which different residuals in each functional zone exceed corresponding thresholds, can identify the dominant mismatch type in the current region and take targeted local correction measures accordingly, giving the inversion process stronger diagnostic and control capabilities.

[0025] 4. Implementing constraint switching and closed-loop correction within a local scope suppresses the propagation of local errors to the global scope, improving inversion stability. Existing methods often employ a unified global update approach. When there are significant observation conflicts or obvious physical closure disruptions in a certain region, local errors can easily spread to other regions through the global solution process. This invention implements local constraint switching and local closed-loop updates within the corresponding functional zones and their adjacent buffer areas. This allows local conflicts to be identified and corrected within a local scope, avoiding unnecessary disturbances to other regions, thereby improving the numerical stability and convergence reliability of the joint inversion process for complex combustion fields.

[0026] 5. Improve the collaborative utilization of multi-source heterogeneous observation information and enhance adaptability to actual engineering observation conditions. Observational data in actual combustion devices come from diverse sources, and various observations differ significantly in spatial coverage, response time scale, and sensitive mechanisms. Existing methods struggle to effectively handle the differences in the roles of different observations in different regions, easily leading to inefficient use of some observational information or the simple averaging of local observation conflicts. This invention, through functional partitioning adaptive activation and regionalized constraint construction, enables different types of observations to play a role in more representative and constrained regions, thereby improving the overall utilization efficiency of multi-source observation information and making it more suitable for complex engineering objects such as aero-engine combustors.

[0027] 6. It possesses excellent scalability and engineering application value. The method does not rely on a single observation device or a single combustion configuration. It allows for flexible adjustment of functional zone types, zone activation conditions, residual construction methods, and local correction strategies based on specific combustion objects and observation conditions. It is applicable to combustion devices with significant backflow flame stabilization structures, as well as combustion objects without significant backflow but exhibiting near-wall thermal coupling or subsequent attenuation characteristics. Therefore, this invention provides a multi-source joint inversion technology route with significant potential for widespread application for aero-engine combustors and other high-temperature reactive flow combustion devices, and provides reliable field information support for combustion state diagnosis, flame stabilization analysis, heat load assessment, and operational optimization.

[0028] 7. This invention provides a novel technical framework for multi-source inversion of complex combustion fields. Compared with existing methods that primarily rely on unified weight fusion or unified objective function minimization, this invention organically combines functional partition adaptive activation, regionalized residual construction, dominant mismatch identification, and local closed-loop correction to form a new joint inversion framework oriented towards local mechanism differences in combustion fields. This framework can more effectively coordinate the relationship between observation consistency requirements and physical closure requirements under multi-source observation conditions, thus providing a new approach for the development of complex combustion field reconstruction technology. Attached Figure Description

[0029] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a flowchart of the multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning in this invention; Figure 2 This is the execution flow of the multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning in this invention; Figure 3 This is an architecture diagram of the multi-source joint inversion system for adaptive activation and physical closure correction of combustion field functional zoning in this invention; The modules are as follows: 301. Data acquisition module; 302. Initial inversion module; 303. Functional partition activation module; 304. Partition residual construction module; 305. Dominant mismatch identification module; 306. Local constraint switching and closed-loop correction module; 307. Convergence judgment module; 308. Result output module. Detailed Implementation

[0031] The embodiments of this application will now be described in detail with reference to the accompanying drawings.

[0032] The following specific examples illustrate the implementation of this application. Those skilled in the art can easily understand other advantages and effects of this application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. This application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this application. It should be noted that, in the absence of conflict, the following embodiments and features of the embodiments can be combined with each other. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0033] like Figure 1 and Figure 2As shown, this invention discloses a multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning. This method includes steps such as multi-source observation data acquisition and preprocessing, initial state field generation, functional zoning library construction and zoning adaptive activation, zoning observation response residual construction, zoning physical closure residual construction, dominant mismatch type identification, local constraint switching, local subdomain closed-loop correction, convergence determination, and output of joint inversion results and diagnostic information. The convergence determination implements process branch control through a judgment node.

[0034] Specifically, the method includes the following steps: S1. Generate initial inversion results based on multi-source observation data of the combustion field to be measured; S2. Based on the combustion field state characteristics of the initial inversion results, adaptively activate the target functional partition corresponding to the current iteration step from the pre-constructed functional partition library; S3. For each activated target functional partition, construct observation response residuals and physical closure residuals respectively. For different target functional partitions, configure different observation response residual construction methods and / or constraint weights for different observation sources in the multi-source observation data. S4. Based on the observed response residual and the physical closure residual, identify the dominant mismatch type of each target functional partition; S5. Based on the dominant mismatch type, perform constraint switching and closed-loop correction within the local range of the corresponding target functional partition, and update the inversion field; S6. Determine whether the updated inversion field satisfies the preset dual convergence criterion. If it does, output the final inversion result and diagnostic information. If it does not, use the updated inversion field as the new initial inversion result for iteration.

[0035] This invention takes an aero-engine combustor with a swirling flame stabilization structure as an example. The object to be reconstructed is a two-dimensional or three-dimensional discrete computational region inside the combustor. This region is divided into... The spatial units to be reconstructed are represented as follows: ; In the formula, This is the combustion field state vector; Indicates the first Temperature within each spatial unit; Indicates the first The equivalent reaction intensity index within each spatial unit can be normalized to... An interval is used to characterize the degree of local reaction activity. Number of spatial discrete units; superscript This indicates transposition. The reason for using an equivalent reaction intensity index instead of a specific single chemical quantity is that complete component information may not be directly obtained in actual engineering observations. Instead, a comprehensive index reflecting local reactivity can be constructed by combining optical response, temperature rise level, historical operating conditions, or prior information.

[0036] In one embodiment of step S1, the multi-source observation data includes at least two of the following: path integral optical observation data, wall temperature measurement data, pressure measurement data, outlet temperature distribution data, and at least two of the following: historical operating condition data, low-order flow prior information, and numerical simulation prior information.

[0037] In addition, since different observation sources differ in measurement principles, spatial coverage, time response characteristics and noise levels, after obtaining the raw data, it is preferable to perform time synchronization, spatial registration, dimensional unification and validity screening on various types of observation data to form a standardized input dataset for subsequent joint inversion.

[0038] After obtaining multi-source observation data, a state characterization model of the combustion field to be reconstructed and corresponding observation mapping relationships are established. The state characterization model describes the spatial distribution of the physical quantities to be reconstructed within the combustion field. These physical quantities can be temperature fields, reaction intensity correlation fields, component distribution correlation fields, or other field variables characterizing the combustion state. For different observation sources, observation response relationships corresponding to the state characterization model are established to describe the mapping relationship between the state to be reconstructed and various types of observation data. Based on this, initial inversion results of the combustion field are obtained using multi-source observation data. These initial inversion results can be obtained through a unified solution or by combining prior information and single-source inversion results. Their purpose is to provide an initial state basis for subsequent functional zoning identification and local loop closure correction.

[0039] In specific implementation, for the first Class observations, their observation models are uniformly represented as: ; In the formula, Indicates the first The measurement vector corresponding to the observation class; Represents the current state vector Through the first Predicted observations obtained from the class-observation mapping relationship; Indicates observation noise; This indicates the total number of observation types. For discrete measurements such as wall temperature and outlet temperature, This can be represented as a sampling mapping from the state field to the response at the measurement point; for path integral optical observations, This can be represented as a forward modeling operator that integrates or calculates the equivalent response of a local field along the optical path; for pressure measurement, It can represent the equivalent mapping relationship between the local thermal state and the distribution of reaction intensity, which leads to the pressure response characteristics. In practical applications, each observation model can be specifically set according to the sensor type and device structure.

[0040] After obtaining multi-source observation data, it is preferable to first generate an initial state field for subsequent functional zone identification and local closed-loop correction. This initial state field is not required to reach the final optimum in the global scope, but mainly serves to provide the initial temperature distribution, reaction intensity distribution, and main structural features of the combustion field. To avoid the initial state generation process relying too much on a single unified objective function, this embodiment preferably uses a multi-source observation residual recursive correction method to construct the initial inversion results.

[0041] Let the initial prior state be Then the first The initial state after round recursion correction can be expressed as: ; In the formula, Indicates the first The state vector obtained by round recursion; Indicates the first Measurement vectors for observations; The first term predicted from the current state is... Observational response; This represents the corresponding observation residual; Indicates the first A mapping operator from observation residuals to state corrections is used to convert observation biases into correction contributions to the temperature field, reaction intensity field, or other states to be reconstructed. Indicates the first Recursive correction coefficients for similar observations; This indicates the total number of observation types.

[0042] Preferably, take As the starting point of the recursion, after several rounds of recursion, when the state corrections of two adjacent rounds satisfy: At this point, the initial correction is stopped, and the initial inversion result is obtained. In the formula, This is the stopping threshold for the initial state generation phase.

[0043] The purpose of the aforementioned initial state generation method is to enable multiple types of observation information to progressively correct the initial state, thereby forming an initial field that reflects the main thermal state distribution and structural characteristics. The resulting initial inversion results are mainly used for subsequent functional partition adaptive activation, partition residual construction, and local closed-loop correction, and are not the final output of this invention.

[0044] In step S2, this specific implementation further constructs a functional partition library and adaptively activates the functional partitions of the region to be reconstructed based on the flow organization characteristics, heat release level, boundary thermal coupling characteristics, and observed response characteristics of the combustion object. The functional partition library includes at least a low-reaction-intensity mixing and heating zone and a main reaction zone, and may further include one or more of the following: a reflux stabilization zone, a near-wall heat exchange zone, and a subsequent decay zone, depending on the structure and combustion organization of the specific combustion object. Specifically, the low-reaction-intensity mixing and heating zone corresponds to a region with low reaction intensity but where mixing and temperature rise have already begun; the main reaction zone corresponds to a region with significant heat release and rapid changes in temperature and state quantities; the reflux stabilization zone corresponds to a region with low-speed reflux, hot spot retention, or significant pressure-thermal response coupling; the near-wall heat exchange zone corresponds to a region significantly affected by the wall thermal boundary; and the subsequent decay zone corresponds to a region where heat release weakens after the main reaction and state quantities tend to evolve more gradually. The activation of the functional partitions can be achieved based on one or more combinations of reaction progress indicators, reaction intensity indicators, temperature gradients, pressure pulsation characteristics, wall distances, prior flow structure information, and other characteristic quantities that can characterize local dominant mechanisms.

[0045] Specifically, in one embodiment of step S2, based on the combustion field state characteristics of the initial inversion result, the target functional partition corresponding to the current iteration step is adaptively activated from the pre-built functional partition library, including: S21. Extract the temperature gradient distribution, reflux prior index, and wall distance parameter from the initial inversion results as combustion field state features; S22. The temperature gradient distribution, the reflux prior index, and the wall distance parameter are compared with the activation threshold of each candidate partition in the functional partition library. S23. The candidate partitions that meet the activation threshold are determined as the target functional partitions of the current iteration step. The target functional partitions include one or more of the following: low reaction intensity mixing and heating zone, main reaction zone, reflux stabilization zone, and near-wall heat exchange zone.

[0046] In practice, after obtaining the initial inversion results, a functional partition library can be further constructed, and the functional partitions of the region to be reconstructed can be adaptively activated. In this embodiment, four types of target functional partitions can be used, namely, low-reaction-intensity mixed-heating regions. Main reaction zone Reflux Stable Region and near-wall heat exchange zone Among them, the low reaction intensity mixing and heating zone is used to characterize the region where the local reaction intensity is low but mixing and temperature rise have begun; the main reaction zone is used to characterize the region where heat release is significant and temperature and reaction state change rapidly; the reflux stabilization zone is used to characterize the region that is significantly affected by swirling induced reflux, hot spot residence and flame stabilization; and the near-wall heat transfer zone is used to characterize the region where the wall thermal boundary has a significant impact.

[0047] In this embodiment, the local temperature gradient index can be defined as: ; In the formula, Indicates the first Normalized temperature gradient index for each unit; Indicates the first Temperature gradient of each unit; To prevent small positive numbers with a denominator of zero, the reflux prior index is defined as follows: Its value range is This can be given by a low-order flow prior field, a historical average flow field, or a candidate region for low-velocity recirculation. Define the first... The distance from each unit to the nearest wall is .

[0048] Based on the aforementioned local temperature gradient indicators, the activation of the target functional area can be performed according to the following rules: (1) The optimal definition of the near-wall heat exchange zone is: In the formula, This is the distance threshold for determining proximity to the wall.

[0049] (2) The optimal definition of the reflux stability region is: In the formula, This is the reflow activation threshold.

[0050] (3) The preferred definition of the main reaction zone is: In the formula, The reaction intensity threshold of the main reaction zone; The temperature gradient threshold of the main reaction zone.

[0051] (4) The preferred definition of the low reaction intensity mixing and heating zone is: ; In the formula, This is the lower threshold of the heating zone for low-reaction-intensity mixing; The inlet reference temperature; This is the temperature rise threshold used to determine when localized heating has begun. If a zone does not meet the activation conditions under the current operating conditions, that zone may not be activated. The purpose of this approach is to allow the zoning strategy to adapt to different combustion objects, rather than imposing fixed zoning on all operating conditions.

[0052] In this embodiment, for different target functional zones, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. Specifically, for the main reaction zone, the constraint weight of path integral optical observation is strengthened; for the near-wall heat transfer zone, the constraint weight of wall temperature measurement data is strengthened; and for the reflux stabilization zone, flame stabilization consistency auxiliary constraints constructed based on pressure measurement data are introduced.

[0053] In one embodiment of step S3, for each activated target functional partition, the observed response residual and the physical closure residual are constructed, including at least one of the following: (1) For the main reaction zone in the target functional area, construct thermochemical consistency residuals to constrain the deviation between the normalized temperature rise and the equivalent reaction intensity index, and construct energy closure residuals to constrain the conservation relationship between the convective transport term, the heat conduction term and the equivalent heat source term. (2) For the reflux stabilization zone in the target functional partition, construct the flame stabilization consistency residual to constrain the Euclidean distance between the hot spot residence centroid coordinates and the reflux core position; (3) For the near-wall heat transfer zone in the target functional area, construct the near-wall heat flux closed residual to constrain the Fourier thermal conductivity relationship between the wall equivalent heat flux and the near-wall normal temperature gradient; (4) For the low reaction intensity mixing and heating zone in the target functional area, construct the physical closed residual of the low reaction intensity mixing and heating zone to constrain the weighted sum of the upper limit of reaction intensity and the upper limit of temperature.

[0054] After activating the functional partitions, the observed response residuals and physical closure residuals are constructed for each target functional partition. In an optional embodiment, for the first... Functional partitions With the The optimal definition for the response residuals of class-based and regional observations is: ； In the formula, For the first The target functional partition corresponds to the first Regional observation response residuals of similar observations; This is a region association operator used to extract or weight the representation of the first region. The observations are mainly affected by the partitioning The affected portion; its specific form can be determined by the location of the measurement point, the optical path crossing relationship, the wall adjacency relationship, or the exit section mapping relationship. The meaning of the above definition is that it no longer only focuses on the overall observation error, but also identifies the local contribution of different observation errors in different zones.

[0055] In one embodiment, for the main reaction zone In this embodiment, the coordination between the thermal state and the reaction state, as well as the local energy closure relationship, are preferably considered simultaneously. Therefore, the normalized temperature rise is first defined as follows: ; In the formula, Indicates the first Normalized temperature rise per unit; This represents the reference high temperature level under the current operating conditions, which can be given by the reference operating conditions, prior calculations, or empirical upper limits. Based on this, the formula for calculating the thermochemical consistency residual of the main reaction zone is: ; in, Thermochemical consistency residuals in the main reaction zone Main reaction zone Number of units inside, For the first Normalized temperature rise per unit, For the first The local reaction intensity index of each unit. This formula represents the local reaction intensity index within the main reaction zone. With normalized temperature rise There should be a good correspondence between the two. If the deviation between the two is large, it indicates that there is a problem of incoordination between temperature rise and reaction state. In other words, the residual quantifies the degree of inconsistency between the reaction progress (normalized temperature rise) derived from the temperature field and the reaction progress directly characterized by the reaction intensity index in the main reaction zone. The smaller the value, the more coordinated the thermo-chemical coupling.

[0056] The formula for calculating the energy closure residual of the main reaction zone is: ; in, The energy closure residual of the main reaction zone, For the first Fluid density within each unit; For the first The local velocity vector of each element or its prior value; For the first Specific enthalpy of each unit; Indicates the first Thermal conductivity of each unit; The equivalent heat source term is used to characterize the heat contribution of the local reaction. The above residual is used to measure the degree of local energy closure between convection, conduction and heat source terms in the main reaction zone. In other words, the residual reflects the degree of violation of local energy conservation. The convection, conduction and heat source terms should satisfy a natural balance relationship. The smaller the value, the higher the degree of energy closure.

[0057] In one embodiment, for the reflux stabilization region The main consideration is the consistency between the hotspot's location and the reflux core's position. The centroid coordinates of the high-temperature, high-reaction zone within the reflux stability region are defined as follows: ; In the formula, For the first The spatial coordinates of each unit; As the weight, the preferred value is . This is used to highlight regions with high temperatures and strong reactions. The location of the reflux nucleus, determined by flow priors or low-velocity reflux, is defined as... The formula for calculating the flame stabilization consistency residual in the reflux stabilization zone is: ; in, For the flame stabilization consistency residual in the reflux stabilization region, The coordinates of the centroid of the high-temperature, high-reaction zone within the reflux stability region. The residual represents the location of the reflux core, determined by flow priors or low-velocity reflux. In other words, this residual reflects the degree of spatial offset between the centroid of the high-temperature, high-reaction zone and the reflux core; the larger the value, the less consistent the current reconstruction result is with the reflux-stabilized flame structure.

[0058] In one embodiment, for the low-reaction-intensity mixing heating zone This embodiment primarily constrains the occurrence of excessively strong local reaction peaks or abnormally high temperatures exceeding reasonable limits. The formula for calculating the physical closure residual of the low-reaction-intensity mixing and heating zone is as follows: ; in, The physical closure residual of the low-reaction-intensity mixing and heating zone. For partitioning Number of internal units; This represents the upper limit of the permissible reaction intensity within the low-reaction-intensity mixing and heating zone; This represents the upper limit of the allowable temperature for the low-reaction-intensity mixing heating zone; This is the weighting coefficient for the temperature term. In other words, this residual is used to suppress high reaction intensities and abnormally high temperatures that do not conform to their functional properties in the low-reaction-intensity mixing and heating zone, and is achieved through a weighted sum of the upper limit constraints on reaction intensity and temperature, respectively.

[0059] In one embodiment, for the near-wall heat exchange zone In this embodiment, constraint is preferably achieved by combining wall temperature measurement and near-wall heat flux closure relationship. If the first The wall temperature measurement value corresponding to each near-wall unit is The residual of the wall temperature response can then be expressed as: ; In the formula, This represents the corresponding wall temperature calculated based on the current reconstructed field and wall heat transfer relationship. The near-wall heat flux closure residual in the near-wall heat transfer zone can be expressed as: ;in, For the near-wall heat flux closure residual in the near-wall heat transfer region, Near-wall heat exchange zone The number of units, For the first Thermal conductivity of the near-wall unit For the first Temperature gradient of near-wall elements along the wall normal This represents the wall-equivalent heat flux. In other words, this residual quantifies the degree of inconsistency between the near-wall vapor temperature field and the wall thermal boundary response, where the Fourier conduction term and the wall-equivalent heat flux should satisfy the heat flow continuity condition.

[0060] In step S4, after obtaining the residual information of each functional partition, the present invention further identifies the mismatch state within each functional partition to determine the dominant mismatch type of the current region. The dominant mismatch type refers to the type of bias that primarily affects the current reconstruction result within the corresponding functional partition, such as underfitting of observations, local thermochemical closure mismatch, local flame-stabilized structure mismatch, near-wall thermal boundary mismatch, or other dominant mismatches. Preferably, the residual category with the largest bias contribution in the current functional partition can be determined by comparing the degree of exceedance of various residuals relative to their corresponding criterion thresholds, and the mismatch mechanism corresponding to this category can be identified as the dominant mismatch type of the functional partition. Through this process, the general "large error" in traditional methods can be further distinguished into "which type of mismatch mechanism is dominant," thus providing a basis for subsequent local corrections.

[0061] In one embodiment of step S4, based on the observed response residuals and the physical closure residuals, the dominant mismatch type of each target functional partition is identified, including: S41, regarding the first Within the target functional area Class residual is The normalized exceedance index is calculated according to the following formula: ； in, For the corresponding criterion threshold, To prevent tiny positive numbers with a denominator of zero; S42. Based on the normalized out-of-limit index, the dominant mismatch type index is determined using the following formula: ; In the formula, Indicates the first The dominant mismatch type index for each target functional partition, as defined above, means that it is not a simple comparison of the absolute values ​​of different residuals, but a comparison of their degree of exceeding the limit relative to the corresponding allowable level, so as to determine which type of mismatch dominates the current region.

[0062] In one embodiment of step S5, after identifying the dominant mismatch type, this embodiment further implements constraint switching and closed-loop correction within a local scope.

[0063] Specifically, local constraint switching refers to adaptively adjusting the mode of action and relative strength of various observation constraints, physical constraints, and auxiliary regularization constraints within a functional partition, based on the dominant mismatch type of that partition. For example: (1) When the dominant mismatch in a functional zone is manifested as insufficient observation fitting, the direct dominant role of conflicting observations in that zone can be reduced, and the constraint effect of other observation information and physical closure relationships in that zone can be enhanced. (2) When the dominant mismatch manifests as thermochemical closure mismatch in the main reaction zone, it can enhance the physical closure constraint that reflects the coordination relationship between temperature rise, reaction intensity and component change, and suppress unreasonable local peaks caused by forced traction through overall smoothing or single observation. (3) When the dominant mismatch manifests as a mismatch in the flame stabilization structure of the recirculation stabilization region, the consistency constraint between the hot spot residence location, pressure response and recirculation candidate region can be enhanced. (4) When the dominant mismatch manifests as the boundary thermal response mismatch of the near-wall heat exchange zone, it can enhance the closed constraint between the wall temperature and the near-wall field and limit the diffusion of local wall errors to the mainstream region.

[0064] Specifically, the so-called local closed-loop correction means that the above constraint switching is not performed uniformly across the entire field, but only locally updated within the corresponding functional partition and its adjacent buffer area to avoid local conflicts interfering with irrelevant areas. After each local update, the system returns the updated state field to the residual calculation and mismatch identification stage of the next iteration (i.e., steps S2 to S4), thus forming a closed-loop correction mechanism across iteration steps: "identification (S4) → switching (S5) → update (S5) → re-identification (next round S4)".

[0065] In specific implementation, for the first For each functional partition, a local objective function is preferably established: ; In the formula, Indicates the first Within the first partition Weighting coefficients for observation residuals; Indicates the first Within the first partition Weighting coefficients for physical closure residuals; Indicates the first The number of physical closure residuals corresponding to each partition; This indicates the local update step size constraint coefficient; Indicates the first Projection operators for each partition and its adjacent buffer region; Indicates the first The current state obtained from round iteration. By introducing This ensures that the current update mainly occurs in local areas related to the current mismatch, rather than spreading indiscriminately throughout the entire field.

[0066] In this embodiment, when the first When the dominant mismatch type of a partition corresponds to a certain type of observation mismatch, it is preferable to reduce the weight of the conflicting observation residuals and appropriately increase the weight of the physically closed residuals, i.e.: ; ; In the formula, To observe the weighting coefficient of the residuals; These are the weighting coefficients for the physical closure residuals. Conversely, when the dominant mismatch type corresponds to a certain type of physical closure mismatch, it is preferable to increase the weight of the corresponding physical closure residual and appropriately reduce the local update step size, i.e. ; ; In the formula, Indicates the currently dominant physical closure residual category; These are the weighting factors for the physical residuals; This represents the local step size contraction coefficient. Using this method, differentiated local corrections can be implemented for different dominant mismatch types, rather than uniform weighting.

[0067] After the local constraint switching is completed, the solution is obtained based on the updated weights and the local objective function. Local correction amount for each functional zone And update the overall status using the following formula: ; Subsequently, each functional zone, various residuals, and their normalized out-of-limit indicators are recalculated, and the closed-loop process of "functional zone status determination - residual calculation - dominant mismatch identification - local constraint switching - local correction update" is repeated.

[0068] In step S6, after completing the local closed-loop correction, the present invention re-solves the updated combustion field state using a multi-source joint solution, and again performs functional zone state determination, residual calculation, and dominant mismatch identification. When the observed response residuals and physical closure residuals of each functional zone meet the preset convergence conditions, or when the residual improvement between two adjacent updates is less than a preset threshold, the iteration stops, and the final combustion field joint inversion result is output. The final result includes not only the spatial distribution of the field variables to be reconstructed, but also mismatch type information, local correction state information, and reconstruction reliability evaluation information for each functional zone, thus providing a basis for subsequent combustion state diagnosis and combustion organization analysis.

[0069] In one embodiment of step S6, the preset dual convergence criterion includes: The global convergence criterion is that the relative change in the inversion field between two adjacent iterations is less than the preset global convergence threshold. The partition convergence criterion is that the physical closure residuals of all activated target function partitions are lower than the preset partition convergence threshold. The system is deemed to satisfy the dual convergence criterion only when both the global convergence criterion and the partition convergence criterion are satisfied.

[0070] Specifically, this invention employs dual convergence criteria. First, a global convergence criterion: the relative change in the inversion field between two adjacent iterations is less than a preset global convergence threshold. Second, a partitioned convergence criterion: all normalized out-of-limit indices for all activated functional partitions are not greater than 1, i.e.: when When the residuals of each region are considered to have met the preset requirements, convergence is determined only when both criteria are met simultaneously, and the iteration can be stopped.

[0071] Furthermore, to prevent infinite iteration, when the residual improvement between two adjacent iterations is less than a given threshold... When the iteration stops, that is: ; In the formula, Indicates the first During each iteration, the vector consists of normalized out-of-limit indices corresponding to all activated functional partitions and all residual types. Indicates the first During each iteration, the vector consists of normalized out-of-limit indices corresponding to all activated functional partitions and all residual types. This is the convergence threshold. The final joint inversion result is output after any criterion is met.

[0072] Specifically, the output of the final inversion results and diagnostic information includes the reconstructed temperature field. Equivalent reaction intensity field The results also include the activation status, dominant mismatch type, and local correction status information of each functional zone. For engineering applications, these results not only reflect the main thermal state and reaction intensity distribution of the combustion field, but also help identify whether local abnormal areas are mainly caused by observation conflicts, or by thermochemical closure mismatch, backflow flame stabilization mismatch, or near-wall thermal boundary mismatch. This provides a basis for combustion state diagnosis, flame stabilization analysis, heat load assessment, and subsequent control.

[0073] It should be noted that the types of functional zones, activation conditions, residual construction methods, local constraint switching strategies, and convergence criteria described in this invention can all be adjusted according to specific combustion objects, specific observation configurations, and specific engineering requirements. However, its core lies in establishing an adaptive activation mechanism for functional zones based on differences in the local dominant mechanisms of the combustion field, and on this basis, achieving stable joint inversion under multi-source observation conditions through regionalized residual diagnosis and local closed-loop correction. This invention is not limited to a single observation form, a fixed combustion configuration, or a fixed solver implementation.

[0074] The state variables, partition activation conditions, residual expressions, and weight update methods used in the above embodiments are all implementations of this invention. In other embodiments, other forms of equivalent reaction intensity indices, other forms of functional partition criteria, other forms of physical closure residuals, and other forms of local update strategies can also be adopted according to different combustion objects, different observation configurations, and different solution requirements. Any technical solution that implements adaptive activation of functional partitions based on differences in the local dominant mechanisms of the combustion field, and performs dominant mismatch identification and local closed-loop correction based on regional observation residuals and physical closure residuals, should be considered to fall within the protection scope of this invention.

[0075] Based on the same inventive concept, this invention also provides a multi-source joint inversion system for adaptive activation and physical closure correction of combustion field functional zones, as described in the following embodiments. Since the principle of the multi-source joint inversion system in solving the problem is similar to the multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zones disclosed in the above embodiments, the implementation of the multi-source joint inversion system can refer to the implementation of the multi-source joint inversion method, and repeated details will not be elaborated further. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0076] Figure 3 This is a structural block diagram of a multi-source joint inversion system for adaptive activation and physical closure correction of combustion field functional zoning disclosed in an embodiment of the present invention, as shown below. Figure 3 As shown, the multi-source joint inversion system includes a data acquisition module 301, an initial inversion module 302, a functional partition activation module 303, a partition residual construction module 304, a dominant mismatch identification module 305, a local constraint switching and closed-loop correction module 306, and a convergence judgment module 307. The structure is described below.

[0077] Among them, the data acquisition module 301 is used to acquire multi-source observation data of the combustion field to be tested; The initial inversion module 302 is used to generate initial inversion results based on the multi-source observation data; The functional partition activation module 303 is used to adaptively activate the target functional partition corresponding to the current iteration step from the pre-built functional partition library according to the combustion field state characteristics of the initial inversion result; The partition residual construction module 304 is used to construct observation response residuals and physical closure residuals for each activated target functional partition. For different target functional partitions, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. The dominant mismatch identification module 305 is used to identify the dominant mismatch type of each target functional partition based on the observed response residual and the physical closure residual; The local constraint switching and closed-loop correction module 306 is used to perform constraint switching and closed-loop correction in the local range of the corresponding target functional partition according to the dominant mismatch type, and update the inversion field; The convergence judgment module 307 is used to determine whether the updated inversion field meets the preset dual convergence criteria. If it does, the final inversion result and diagnostic information are output. If it does not, the updated inversion field is used as the new initial inversion result for iteration.

[0078] In addition, the system also includes a result output module 308, which is used to output the final joint inversion results and related diagnostic information.

[0079] The modules mentioned above interact with each other through data interfaces and control logic to complete the entire process of multi-source joint inversion of the combustion field.

[0080] The embodiments of the present invention achieve the following technical effects: 1. Implementing zoned adaptive inversion to address local mechanism differences in the combustion field overcomes the masking of non-uniformity by uniform processing across the entire field, thereby improving the reconstructive accuracy.

[0081] 2. By balancing observational fitting and physical closure, the partitioned construction of double residuals ensures that the results conform to the dominant mechanism, thereby enhancing the physical explanatory power and engineering credibility.

[0082] 3. Identify the dominant mismatch type in each partition, realize the transformation from overall error processing to mismatch mechanism diagnosis, and improve the diagnostic and control capabilities of the inversion process.

[0083] 4. Local constraint switching and closed-loop correction suppress global error propagation, avoid inter-regional disturbances, and improve numerical stability and convergence reliability.

[0084] 5. Enhance the collaborative utilization of multi-source heterogeneous observations, enabling various types of observations to play a role in suitable areas and adapt to complex engineering observation conditions.

[0085] 6. This invention is highly scalable and can be flexibly adapted to different combustion configurations and observation conditions, providing reliable field information support for devices such as aero engines.

[0086] 7. By constructing a new framework that organically combines partition activation, residual construction, mismatch identification, and local correction, the requirements for observation consistency and physical closure are coordinated.

[0087] In this embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-mentioned multi-source joint inversion method for adaptive activation and physical closure correction of any combustion field functional zone.

[0088] Specifically, the computer device can be a computer terminal, a server, or a similar computing device.

[0089] In this embodiment, a computer-readable storage medium is provided, which stores a computer program that performs the above-described multi-source joint inversion method for adaptive activation and physical closure correction of any combustion field functional zoning.

[0090] Specifically, computer-readable storage media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer-readable storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable storage media does not include transient media, such as modulated data signals and carrier waves.

[0091] Obviously, those skilled in the art should understand that the modules or steps of the above-described embodiments of the present invention can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those presented here, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the embodiments of the present invention are not limited to any particular hardware and software combination.

[0092] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, various modifications and variations can be made to the embodiments of 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 multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning, characterized in that, include: Initial inversion results are generated based on multi-source observation data of the combustion field to be measured. Based on the combustion field state characteristics of the initial inversion results, the target functional partition corresponding to the current iteration step is adaptively activated from the pre-constructed functional partition library; For each activated target functional partition, observation response residuals and physical closure residuals are constructed respectively. For different target functional partitions, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. Based on the observed response residuals and the physical closure residuals, the dominant mismatch type of each target functional partition is identified; Based on the dominant mismatch type, constraint switching and closed-loop correction are performed within the local range of the corresponding target functional partition to update the inversion field; Determine whether the updated inversion field satisfies the preset dual convergence criterion. If it does, output the final inversion result and diagnostic information. If it does not, use the updated inversion field as the new initial inversion result for iteration.

2. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 1, characterized in that, The multi-source observation data includes at least two of the following: path integral optical observation data, wall temperature measurement data, pressure measurement data, outlet temperature distribution data, and at least two of the following: historical operating condition data, low-order flow prior information, and numerical simulation prior information.

3. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 1, characterized in that, Based on the combustion field state characteristics of the initial inversion results, the target functional partition corresponding to the current iteration step is adaptively activated from the pre-built functional partition library, including: The temperature gradient distribution, reflux prior index, and wall distance parameter in the initial inversion results are extracted as combustion field state features. The temperature gradient distribution, the reflux prior index, and the wall distance parameter are compared with the activation threshold of each candidate partition in the functional partition library. Candidate partitions that meet the activation threshold are determined as target functional partitions for the current iteration step. The target functional partitions include one or more of the following: low reaction intensity mixing and heating zone, main reaction zone, reflux stabilization zone, and near-wall heat exchange zone.

4. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 1 or 3, characterized in that, For each activated target functional partition, construct the observation response residual and the physical closure residual, including at least one of the following: For the main reaction zone in the target functional area, a thermochemical consistency residual is constructed to constrain the deviation between the normalized temperature rise and the equivalent reaction intensity index, and an energy closure residual is constructed to constrain the conservation relationship between the convection transport term, the heat conduction term and the equivalent heat source term. For the reflux stabilization zone in the target functional partition, a flame stabilization consistency residual is constructed to constrain the Euclidean distance between the hot spot residence centroid coordinates and the reflux core location. For the near-wall heat transfer zone in the target functional area, a near-wall heat flux closure residual is constructed to constrain the Fourier thermal conductivity relationship between the wall equivalent heat flux and the near-wall normal temperature gradient. For the low-reaction-intensity mixing and heating zone in the target functional area, a physical closed residual of the low-reaction-intensity mixing and heating zone is constructed to constrain the weighted sum of the upper limit of reaction intensity and the upper limit of temperature.

5. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 4, characterized in that, The formula for calculating the thermochemical consistency residual of the main reaction zone is: ； in, Thermochemical consistency residuals in the main reaction zone Main reaction zone Number of units inside, For the first Normalized temperature rise per unit, For the first Local reaction intensity index of each unit; The formula for calculating the energy closure residual of the main reaction zone is: ； in, The energy closure residual of the main reaction zone, For the first Fluid density within each unit; For the first The local velocity vector of each element or its prior value; For the first Specific enthalpy of each unit; Indicates the first Thermal conductivity of each unit; This is the equivalent heat source term.

6. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 4, characterized in that, The formula for calculating the flame stabilization consistency residual in the reflow stabilization region is: ； in, For the flame stabilization consistency residual in the reflux stabilization region, The coordinates of the centroid of the high-temperature, high-reaction zone within the reflux stability region. The location of the reflux core is determined by flow priors or low-velocity reflux. The formula for calculating the near-wall heat flux closure residual in the near-wall heat transfer zone is: ； in, For the near-wall heat flux closure residual in the near-wall heat transfer region, Near-wall heat exchange zone The number of units, For the first Thermal conductivity of the near-wall unit For the first Temperature gradient of near-wall elements along the wall normal This represents the equivalent heat flux of the wall.

7. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 4, characterized in that, The formula for calculating the physical closure residual of the low-reaction-intensity mixing and heating zone is as follows: ； in, The physical closure residual of the low-reaction-intensity mixing and heating zone. For partitioning Number of internal units; This represents the upper limit of the permissible reaction intensity within the low-reaction-intensity mixing and heating zone; This represents the upper limit of the allowable temperature for the low-reaction-intensity mixing heating zone; This represents the weighting coefficient for the temperature term.

8. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 1, characterized in that, Based on the observed response residuals and the physical closure residuals, the dominant mismatch types for each target functional partition are identified, including: For the Within the target functional partition, the first Class residual is The normalized exceedance index is calculated according to the following formula: ； in, For the corresponding criterion threshold, To prevent tiny positive numbers with a denominator of zero; Based on the normalized out-of-limit index, the dominant mismatch type index is determined using the following formula: ； In the formula, Indicates the first The dominant mismatch type index for each target functional partition.

9. The multi-source joint inversion method for adaptive activation and physical closure correction of combustion field functional zoning according to claim 1, characterized in that, The preset dual convergence criteria include: The global convergence criterion is that the relative change in the inversion field between two adjacent iterations is less than the preset global convergence threshold. The partition convergence criterion is that the physical closure residuals of all activated target function partitions are lower than the preset partition convergence threshold. The system is deemed to satisfy the dual convergence criterion only when both the global convergence criterion and the partition convergence criterion are satisfied.

10. A multi-source joint inversion system for adaptive activation and physical closure correction of combustion field functional zoning, characterized in that, Includes one or more processors; as well as A memory storing instructions that, when executed by the one or more processors, cause the system to perform the method as described in any one of claims 1 to 9, wherein the memory includes: The data acquisition module is used to acquire multi-source observation data of the combustion field under test; The initial inversion module is used to generate initial inversion results based on the multi-source observation data; The functional partition activation module is used to adaptively activate the target functional partition corresponding to the current iteration step from the pre-built functional partition library based on the combustion field state characteristics of the initial inversion result. The partition residual construction module is used to construct observation response residuals and physical closure residuals for each activated target functional partition. For different target functional partitions, different observation response residual construction methods and / or constraint weights are configured for different observation sources in the multi-source observation data. The dominant mismatch identification module is used to identify the dominant mismatch type of each target functional partition based on the observed response residual and the physical closure residual; The local constraint switching and closed-loop correction module is used to perform constraint switching and closed-loop correction within a local range of the corresponding target functional partition according to the dominant mismatch type, and update the inversion field. The convergence judgment module is used to determine whether the updated inversion field meets the preset dual convergence criteria. If it does, the final inversion result and diagnostic information are output. If it does not, the updated inversion field is used as the new initial inversion result for iteration.

Images