A TDLAS two-dimensional temperature field reconstruction method, system, equipment and medium

Abstract

This invention relates to the field of spectral detection technology and discloses a TDLAS two-dimensional temperature field reconstruction method, system, device, and medium. The method includes: acquiring initial integrated absorbance data of a target region using a TDLAS measurement system; using a first CNN model to output a first temperature field distribution of the target region based on the initial integrated absorbance data; based on the first temperature field distribution and the absorption spectral physical model of the TDLAS measurement system, inverting to obtain the integrated absorbance data corresponding to the missing optical path of the TDLAS measurement system when acquiring the initial integrated absorbance data, so as to obtain the target integrated absorbance data of the target region by combining the initial integrated absorbance data; using a second CNN model to output a second temperature field distribution of the target region based on the target integrated absorbance data; and iteratively reconstructing the target temperature field distribution of the target region using a polynomial fitting method, starting from the second temperature field distribution, achieving higher accuracy.

Description

Technical Field

[0001] This invention relates to the field of spectral detection technology, and in particular to a TDLAS two-dimensional temperature field reconstruction method, system, device and medium. Background Technology

[0002] Tunable Diode Laser Absorption Spectroscopy (TDLAS) technology, with its outstanding advantages such as non-contact operation, high sensitivity, fast response, and strong anti-interference capabilities, is widely used in temperature field measurement and monitoring in high-temperature and complex environments such as industrial combustion fields, aero-engines, and boilers. In high-temperature and complex environments, uneven temperature distribution and numerous interference factors pose challenges to the accurate reconstruction of two-dimensional temperature fields using TDLAS technology. Furthermore, in practical engineering, factors such as laser obstruction, optical path attenuation, and sensor malfunctions can easily lead to optical path defects, resulting in incomplete effective absorbance data and ill-posed reconstruction problems, further exacerbating the difficulty of accurate two-dimensional temperature field reconstruction.

[0003] Reconstruction methods based on deep learning, such as convolutional neural networks (CNNs), rely on massive amounts of labeled data for model training. They can quickly extract feature information from TDLAS spectral data to achieve rapid reconstruction of the temperature field. However, under conditions of data loss due to missing optical paths or abnormal distribution, the stability and accuracy of model inference drop sharply, making it impossible to complete effective temperature field reconstruction. Summary of the Invention

[0004] The purpose of this invention is to provide a TDLAS two-dimensional temperature field reconstruction method, system, device and medium, which can solve the problem that CNN models cannot effectively complete temperature field reconstruction under the condition of missing optical path.

[0005] To address the aforementioned technical problems, embodiments of the present invention provide a TDLAS two-dimensional temperature field reconstruction method, comprising the following steps: Initial integrated absorbance data of the target area were obtained using the TDLAS measurement system; The first CNN model is used to output the first temperature field distribution of the target area based on the initial integrated absorbance data; Based on the initial temperature field coarse distribution and the absorption spectrum physical model of the TDLAS measurement system, the integrated absorbance data corresponding to the missing optical path of the TDLAS measurement system when acquiring the initial integrated absorbance data is obtained by inversion, so as to combine the initial integrated absorbance data to obtain the target integrated absorbance data of the target area. The second CNN model is used to output the second temperature field distribution of the target region based on the target integrated absorbance data; Based on the polynomial fitting method, the target temperature field distribution of the target region is iteratively reconstructed starting from the second temperature field distribution.

[0006] Furthermore, after acquiring the initial integrated absorbance data of the target area using the TDLAS measurement system, the method further includes: Based on the initial integrated absorbance data, the missing optical path of the TDLAS measurement system with respect to the target region was determined when acquiring the initial integrated absorbance data.

[0007] Further, the integral absorbance data corresponding to the missing optical path of the TDLAS measurement system for the target region is obtained by inverting the data based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, including: Based on the first temperature field distribution, Beer-Lambert's law of the TDLAS measurement system, and the HITRAN spectral database, the integrated absorbance data of the missing optical path was determined.

[0008] Further, obtaining the target integrated absorbance data of the target region by combining the initial integrated absorbance data includes: Based on the initial integrated absorbance data, the missing optical path, and the corresponding integrated absorbance data, the target integrated absorbance data of the TDLAS measurement system for the target region is determined.

[0009] Furthermore, the step of iteratively reconstructing the target temperature field distribution of the target region based on the polynomial fitting method, starting from the second temperature field distribution, includes: An initial polynomial temperature model for the target region is constructed based on the polynomial fitting method. The minimum squares method is adopted, with the goal of minimizing the residual between the integrated absorbance data corresponding to the temperature calculated by the initial polynomial temperature model and the measured integrated absorbance data. Starting from the second temperature field distribution, the polynomial coefficients of the two-dimensional polynomial temperature model are iteratively optimized to obtain the target polynomial temperature model for characterizing the target temperature field distribution.

[0010] Furthermore, the order of the initial polynomial temperature model is determined based on the temperature field complexity of the target region.

[0011] Furthermore, the target area is the combustion chamber of an aircraft engine.

[0012] Embodiments of the present invention also provide a TDLAS two-dimensional temperature field reconstruction, including: The integrated absorbance acquisition module is used to acquire the initial integrated absorbance data of the target area through the TDLAS measurement system; The first temperature field construction module is used to output the first temperature field distribution of the target area based on the initial integrated absorbance data using the first CNN model; The integral absorbance completion module is used to invert the integral absorbance data corresponding to the missing optical path of the TDLAS measurement system when the initial integral absorbance data is acquired, based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, so as to obtain the target integral absorbance data of the target area by combining the initial integral absorbance data. The second temperature field construction module is used to output the second temperature field distribution of the target region based on the target integral absorbance data using the second CNN model; The target temperature field construction module is used to iteratively reconstruct the target temperature field distribution of the target region based on the polynomial fitting method, starting from the second temperature field distribution.

[0013] Embodiments of the present invention also provide a computer device, including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the above-described TDLAS two-dimensional temperature field reconstruction method.

[0014] Embodiments of the present invention also provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described TDLAS two-dimensional temperature field reconstruction method.

[0015] The TDLAS two-dimensional temperature field reconstruction method provided by this invention has at least the following beneficial effects: After acquiring the initial integrated absorbance data of the target area using the TDLAS measurement system, this data contains missing data corresponding to optical paths. Therefore, the first CNN model outputs an initial coarse temperature field distribution based on this missing optical path data, providing a priori temperature field basis for completing the missing optical path. Combined with the absorption spectrum physical model of the TDLAS measurement system—that is, based on the physical correlation between the temperature field and spectral absorption—the data corresponding to the missing optical path is retrieved and the initial data is completed, ensuring the rationality and accuracy of the completed data. Then, a second CNN model is used to output a high-precision temperature field distribution. This collaborative design of two-stage CNNs results in a temperature field accuracy far exceeding that of a single CNN model, providing high-quality initial values ​​for subsequent polynomial fitting. The iterative process of polynomial fitting ensures the physical consistency and smoothness of the temperature field, compensating for the shortcomings of the CNN model in terms of physical rationality, thereby improving the reconstruction accuracy of the final temperature field distribution. Attached Figure Description

[0016] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0017] Figure 1 This is a schematic flowchart of a TDLAS two-dimensional temperature field reconstruction method provided by the present invention. Figure 2 A schematic diagram of an optical path provided by the present invention; Figure 3 This invention provides a centrally symmetric Gaussian distribution real temperature field cloud map; Figure 4 A reconstructed temperature field cloud map provided by the present invention; Figure 5 This invention provides a single CNN method for predicting temperature field cloud maps. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0019] Currently, traditional reconstruction methods (such as ART and SA) suffer from a significant reduction in effective projected data due to data gaps, resulting in insufficient effective constraints on equations, a sharp increase in matrix rank deficiency and ill-conditioned phenomena, and an inability to achieve stable and reliable reconstruction, making them unsuitable for the precise measurement requirements of complex high-temperature environments. Reconstruction methods based on deep learning, such as convolutional neural networks (CNNs), rely on massive amounts of labeled data for model training, enabling rapid extraction of feature information from TDLAS spectral data and rapid reconstruction of the temperature field. However, these methods require even larger training datasets for high-resolution temperature field reconstruction, leading to a surge in model complexity and computational costs, thus increasing the cost of practical engineering applications. Furthermore, the reconstruction results from deep learning models are prone to insufficient physical plausibility, and the model's generalization ability is limited. Under conditions of data gaps due to missing optical paths or abnormal distribution, the model's inference stability and accuracy drop sharply, making it impossible to reconstruct an effective temperature field. Simultaneously, pure deep learning methods lack the high-resolution refinement and continuous distribution characteristics of polynomial analytical models, making it difficult to balance accuracy and detailed representation.

[0020] Therefore, there is a need for a TDLAS two-dimensional temperature field reconstruction method that can adapt to the working conditions of missing optical paths, balance reconstruction accuracy and cost control, and solve the defects of existing methods, so as to achieve stable and high-precision reconstruction under missing optical paths and reduce the cost of high-resolution reconstruction.

[0021] To overcome the problems of reconstruction failure and sharp drop in accuracy of existing TDLAS two-dimensional temperature field reconstruction technology under optical path missing conditions, this invention provides a TDLAS two-dimensional temperature field reconstruction method for optical path missing conditions. This method can solve engineering problems such as incomplete data, ill-conditioned solutions, and easy divergence in iteration caused by optical path missing, while also addressing issues such as high training cost, insufficient physical plausibility, and poor generalization of high-resolution CNNs. It provides stable and reliable real-time high-resolution temperature field monitoring technology support for harsh environments such as industrial combustion fields and aero-engine combustion chambers, which are characterized by high temperatures, complexity, and susceptibility to optical path missing.

[0022] The technical solutions provided by the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0023] One embodiment of the present invention relates to a TDLAS two-dimensional temperature field reconstruction method. The specific process of the TDLAS two-dimensional temperature field reconstruction method in this embodiment can be as follows: Figure 1 As shown, it includes: Step 101: Obtain the initial integrated absorbance data of the target area using the TDLAS measurement system; Step 102: Using the first CNN model, output the first temperature field distribution of the target area based on the initial integrated absorbance data; Step 103: Based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, the integrated absorbance data corresponding to the missing optical path of the TDLAS measurement system when acquiring the initial integrated absorbance data is obtained by inversion, so as to combine the initial integrated absorbance data to obtain the target integrated absorbance data of the target area. Step 104: Using the second CNN model, output the second temperature field distribution of the target region based on the target integrated absorbance data; Step 105: Based on the polynomial fitting method, starting from the second temperature field distribution, iteratively reconstruct the target temperature field distribution of the target region.

[0024] The following is a detailed description of the implementation details of the TDLAS two-dimensional temperature field reconstruction method in this embodiment. The following content is only for the convenience of understanding the implementation details and is not necessary for implementing this solution.

[0025] In step 101, absorbance measurement data under conditions of missing optical paths is acquired. Multi-path integrated absorbance data of the target detection area is obtained using the TDLAS measurement system as initial integrated absorbance data. Then, based on this initial integrated absorbance data, the missing optical paths of the TDLAS measurement system regarding the target area are determined when the initial integrated absorbance data was acquired. This is because some optical paths in the obtained integrated absorbance data suffer signal loss due to factors such as obstruction, sensor failure, or vibration interference. Therefore, the locations of the missing optical paths and the corresponding missing data are marked.

[0026] In step 102, a coarse CNN prediction is performed. The absorbance data with missing optical paths obtained in step 101 is input into the pre-trained first CNN model. Here, the CNN model is trained with "incomplete absorbance data - coarse temperature field distribution" as the training target, which has strong anti-missing ability and outputs the initial coarse temperature field distribution of the target detection area. This initial coarse temperature field distribution is used to provide a priori temperature field basis for completing the missing optical path.

[0027] In step 103, missing optical path completion is performed. Based on the initial coarse temperature field distribution (i.e., the first temperature field distribution) obtained in step 102, and combined with the TDLAS absorption spectral physical model (Beer-Lambert's law and the HITRAN spectral database), the integrated absorbance data corresponding to the missing optical path is calculated by inversion. Based on the initial integrated absorbance data, the missing optical path, and the corresponding integrated absorbance data, the target integrated absorbance data for the TDLAS measurement system regarding the target region is determined, i.e., the missing optical path data is completed, resulting in complete absorbance data. This completion process is not simple data interpolation, but rather based on the physical relationship between the temperature field and spectral absorption, ensuring the rationality and accuracy of the completed data.

[0028] In step 104, CNN-based fine prediction is performed. The complete integrated absorbance data, supplemented in step 103, is input into the pre-trained second CNN model. Here, the second CNN model uses "complete absorbance data - high-precision temperature field distribution" as its training objective and outputs a high-precision temperature field. The second CNN model does not need to pursue excessively high resolution (resolution can be flexibly improved later through polynomial fitting), and its training cost is far lower than that of high-resolution CNN models.

[0029] In step 105, polynomial fitting iterative optimization is performed. Taking the high-precision temperature field obtained in step 104 as the starting point of the iteration, the polynomial fitting method is used for iterative reconstruction. Among them, the order of the polynomial and the reconstruction resolution are adjusted according to the actual detection requirements (the resolution can be set arbitrarily without adding extra training costs). After the iteration converges, the final two-dimensional temperature field is obtained.

[0030] Specifically, based on the polynomial fitting method, an initial polynomial temperature model for the target region is constructed. Using the minimum square method, with the goal of minimizing the residual between the integral absorbance data corresponding to the temperature calculated by the initial polynomial temperature model and the measured integral absorbance data, the polynomial coefficients of the two-dimensional polynomial temperature model are iteratively optimized, starting from the second temperature field distribution, to obtain a target polynomial temperature model for characterizing the target temperature field distribution.

[0031] The following example uses the TDLAS two-dimensional temperature field reconstruction of an aero-engine combustion chamber. The target detection area is a rectangular area of ​​5cm*5cm, the preset reconstruction resolution is 150*150, that is, 0.33mm*0.33mm, and the maximum allowable optical path loss ratio is 30%.

[0032] First, absorbance measurement data was acquired under conditions of missing optical paths. Using the TDLAS measurement system, 40 detection optical paths (4 angles, 10 optical paths per angle) were arranged in the target combustion area, such as... Figure 2 As shown, absorbance data of each optical path was collected; simulating industrial site shading conditions, signal loss was artificially set in 8 optical paths (20%), and the positions of the missing optical paths (optical paths 2, 3, 5, 12, 27, 28, 34, and 35) were marked to obtain integrated absorbance data with missing optical paths.

[0033] Then, the first stage of CNN coarse prediction is performed. A training set is constructed: generating the true temperature field distribution (resolution 50*50, i.e., 1mm*1mm). Figure 3 As shown, the integral absorbance data of the corresponding randomly missing maximum 30% optical path is obtained by combining numerical simulation; a first CNN model is constructed, the training set is input into the first CNN model, the mean squared error (MSE) is used as the loss function, and the Adam optimizer is used to train the model; the integral absorbance data with missing optical path is input into the trained first CNN model, and the initial coarse temperature distribution of the target area (resolution of 50*50, i.e. 1mm*1mm) is output as the prior basis for optical path completion.

[0034] Next, missing optical paths are filled in. Based on the initial coarse temperature field distribution, combined with Beer-Lambert's law and the HITRAN spectral database, the theoretical integrated absorbance of each missing optical path is calculated to physically fill in the missing data. This ensures that the filled absorbance data conforms to the laws of spectral physics and avoids physical distortion problems caused by pure data interpolation. The calculated integrated absorbance values ​​are then added to the missing positions to obtain complete integrated absorbance data (all 40 optical paths are complete).

[0035] The second stage involves precise CNN prediction. A training set is constructed: a true temperature field distribution (50*50 resolution, i.e., 1mm*1mm) is generated, and the integrated absorbance data of the complete optical path is obtained through numerical simulation. The training samples are input into the second CNN model, using MSE as the loss function and the Adam optimizer to train the model. The completed integrated absorbance data is then input into the trained second CNN model, outputting a high-precision temperature field (50*50 resolution, i.e., 1mm*1mm). Compared to the coarsely reconstructed temperature field, this temperature field significantly reduces error, provides clearer details of the temperature distribution, and accurately reflects the temperature gradient and peak distribution of the measurement area, providing high-quality initial values ​​for subsequent polynomial iterative optimization.

[0036] Finally, polynomial fitting and iterative optimization were performed. Using a high-precision temperature field as the initial value for iteration, a two-dimensional 6th-order polynomial temperature model is constructed (the order is adjusted according to the complexity of the temperature field in the target region). An objective function is constructed based on the least squares method, with the goal of minimizing the residual between the measured integrated absorbance and the model-calculated absorbance. The polynomial coefficients are iteratively solved. During the iteration process, the polynomial coefficients are continuously corrected to gradually reduce the temperature field reconstruction deviation until the residual meets the preset accuracy requirements, at which point the iteration stops. Since the resolution of the polynomial fitting method can be arbitrarily adjusted, no additional model training is required to output the final high-resolution two-dimensional temperature field (150*150, i.e., 0.33mm*0.33mm).

[0037] This invention addresses the common and challenging industrial environment of missing optical paths by constructing a collaborative reconstruction system of "optical path completion - two-stage CNN initial value optimization - polynomial fitting iteration." It is not a simple aggregation of existing technologies, but rather combines the advantages of polynomial fitting and CNN to solve the problem of high cost in high-resolution reconstruction. Specifically:

[0038] (1) Targeted solution to the reconstruction problem of optical path missing conditions: Specifically designed for the optical path missing problem caused by optical path occlusion and sensor failure in industrial sites, an optical path completion strategy of "CNN coarse prediction-physical completion" was designed. The absorbance data of the missing optical path is back-inferred through the temperature field prior, which solves the problem that traditional methods cannot reconstruct or have extremely low accuracy under optical path missing conditions, and has clear technical targeting; (2) Two-stage CNN collaborative optimization of temperature field, balancing accuracy and cost: Two functionally differentiated CNN models are used. The first-stage lightweight CNN is used to coarsely predict the temperature field to assist in optical path completion. The second-stage CNN does not require high-resolution design and only outputs a high-precision temperature field, avoiding the need for a large amount of training data for high-resolution CNN models and significantly reducing training costs. At the same time, the collaborative design of the two-stage CNN makes the temperature field accuracy much higher than that of a single CNN model, providing high-quality initial values ​​for subsequent polynomial fitting. (3) Integrating the resolution advantage of polynomial fitting to reduce the cost of high-resolution reconstruction: The core advantage of polynomial fitting is that the reconstruction resolution can be arbitrarily adjusted. Without relying on high-resolution CNN models, the temperature field reconstruction resolution can be flexibly adjusted according to actual needs, which solves the pain point of extremely high training cost of high-resolution CNN. At the same time, the iterative process of polynomial fitting can ensure the physical consistency and smoothness of the temperature field, making up for the shortcomings of CNN models in physical rationality. (4) Synergistic progression of technical processes: The entire reconstruction process (missing optical path data → coarse CNN prediction → optical path completion → fine CNN prediction → polynomial fitting) is interconnected, with each step providing the necessary basis for the next step, forming a complete technical link of "data completion - initial value optimization - physical reconstruction".

[0039] This invention reconstructs the temperature field in scenarios with missing optical paths through the above embodiments. The accuracy of the reconstruction results is evaluated below, and compared with a single CNN reconstruction method to verify the superiority of the method presented in this invention. The temperature field reconstruction cloud maps and error comparison results for each method in scenarios with missing optical paths are as follows: Figure 4 The temperature field contour map reconstructed by the method of the present invention. Figure 5 This is a temperature field cloud map predicted by a single CNN method. Table 1 compares the two core metrics, Mean Absolute Relative Error (MARE) and Maximum Absolute Relative Error (MAXRE), between the two methods.

[0040] Table 1. Comparison of the two core indicators, mean absolute relative error and maximum absolute relative error, between the two methods. The reconstruction accuracy of the method of this invention is as follows: Under the condition of 20% optical path loss, the average relative error between the final reconstructed temperature field and the real temperature field is 0.78%, and the maximum relative error is 5.80%, which fully meets the accuracy requirements of industrial temperature monitoring. Comparative verification results show that, under the condition of 20% optical path loss, the average absolute relative error of the single CNN reconstruction method is as high as 15.9%, and the maximum absolute relative error is as high as 26.2%, which is significantly lower than the method of this invention and cannot meet the needs of practical applications. Therefore, this invention, through a "three-level collaborative" fusion strategy, effectively improves the reconstruction accuracy in scenarios with missing optical paths and significantly enhances robustness.

[0041] The above demonstrates that the method of the present invention has significant advantages in the case of optical path loss, combining high resolution, high precision and fast convergence, and is suitable for real-time monitoring needs of high-temperature and complex combustion fields.

[0042] The steps of the various methods described above are only for clarity. In practice, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the protection scope of this invention. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, without changing the core design of the algorithm and process, are also within the protection scope of this invention.

[0043] Another embodiment of the present invention relates to a TDLAS two-dimensional temperature field reconstruction system. The implementation details of this embodiment's TDLAS two-dimensional temperature field reconstruction system are described below. The following details are provided for ease of understanding and are not essential for implementing this solution. This embodiment's TDLAS two-dimensional temperature field reconstruction system includes: The integrated absorbance acquisition module is used to acquire the initial integrated absorbance data of the target area through the TDLAS measurement system; The first temperature field construction module is used to output the first temperature field distribution of the target area based on the initial integrated absorbance data using the first CNN model; The integral absorbance completion module is used to invert the integral absorbance data corresponding to the missing optical path of the TDLAS measurement system when the initial integral absorbance data is acquired, based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, so as to obtain the target integral absorbance data of the target area by combining the initial integral absorbance data. The second temperature field construction module is used to output the second temperature field distribution of the target region based on the target integral absorbance data using the second CNN model; The target temperature field construction module is used to iteratively reconstruct the target temperature field distribution of the target region based on the polynomial fitting method, starting from the second temperature field distribution.

[0044] It is not difficult to see that this embodiment is a system embodiment corresponding to the above method embodiments, and this embodiment can be implemented in conjunction with the above method embodiments. The relevant technical details and technical effects mentioned in the above embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above embodiments.

[0045] It is worth mentioning that all modules involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this invention, this embodiment does not introduce units that are not closely related to solving the technical problem proposed by this invention; however, this does not mean that other units are absent from this embodiment.

[0046] Another embodiment of the present invention relates to a computer device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the TDLAS two-dimensional temperature field reconstruction method in the above embodiments.

[0047] The memory and processor are connected via a bus, which can include any number of interconnecting buses and bridges, connecting various circuits of one or more processors and memories. The bus can also connect various other circuits, such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single element or multiple elements, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0048] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0049] Another embodiment of the present invention relates to a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the method embodiments described above.

[0050] That is, those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a device (which may be a microcontroller, chip, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0051] Those skilled in the art will understand that the above embodiments are specific embodiments for implementing the present invention, and in practical applications, various changes can be made to them in form and detail without departing from the spirit and scope of the present invention.

Claims

1. A TDLAS two-dimensional temperature field reconstruction method, characterized in that, The method includes: Initial integrated absorbance data of the target area were obtained using the TDLAS measurement system; The first CNN model is used to output the first temperature field distribution of the target area based on the initial integrated absorbance data; Based on the first temperature field distribution and the physical model of the absorption spectrum of the TDLAS measurement system, the integrated absorbance data corresponding to the missing optical path of the TDLAS measurement system when acquiring the initial integrated absorbance data is obtained by inversion, so as to obtain the target integrated absorbance data of the target area by combining the initial integrated absorbance data. The second CNN model is used to output the second temperature field distribution of the target region based on the target integrated absorbance data; Based on the polynomial fitting method, the target temperature field distribution of the target region is iteratively reconstructed starting from the second temperature field distribution.

2. The TDLAS two-dimensional temperature field reconstruction method according to claim 1, characterized in that, After acquiring the initial integrated absorbance data of the target area using the TDLAS measurement system, the method further includes: Based on the initial integrated absorbance data, the missing optical path of the TDLAS measurement system with respect to the target region was determined when acquiring the initial integrated absorbance data.

3. The TDLAS two-dimensional temperature field reconstruction method according to claim 2, characterized in that, The integral absorbance data corresponding to the missing optical path of the TDLAS measurement system for the target region is obtained by inversion based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, including: Based on the first temperature field distribution, Beer-Lambert's law of the TDLAS measurement system, and the HITRAN spectral database, the integrated absorbance data of the missing optical path was determined.

4. The TDLAS two-dimensional temperature field reconstruction method according to claim 3, characterized in that, The step of obtaining the target integrated absorbance data of the target region by combining the initial integrated absorbance data includes: Based on the initial integrated absorbance data, the missing optical path, and the corresponding integrated absorbance data, the target integrated absorbance data of the TDLAS measurement system for the target region is determined.

5. The TDLAS two-dimensional temperature field reconstruction method according to claim 1, characterized in that, The method based on polynomial fitting, starting with the second temperature field distribution, iteratively reconstructs the target temperature field distribution of the target region, including: An initial polynomial temperature model for the target region is constructed based on the polynomial fitting method. The minimum squares method is adopted, with the goal of minimizing the residual between the integrated absorbance data corresponding to the temperature calculated by the initial polynomial temperature model and the measured integrated absorbance data. Starting from the second temperature field distribution, the polynomial coefficients of the two-dimensional polynomial temperature model are iteratively optimized to obtain the target polynomial temperature model for characterizing the target temperature field distribution.

6. The TDLAS two-dimensional temperature field reconstruction method according to claim 5, characterized in that, The order of the initial polynomial temperature model is determined based on the complexity of the temperature field in the target region.

7. The TDLAS two-dimensional temperature field reconstruction method according to claim 1, characterized in that, The target area is the combustion chamber of an aircraft engine.

8. A TDLAS two-dimensional temperature field reconstruction system, characterized in that, The system includes: The integrated absorbance acquisition module is used to acquire the initial integrated absorbance data of the target area through the TDLAS measurement system; The first temperature field construction module is used to output the first temperature field distribution of the target area based on the initial integrated absorbance data using the first CNN model; The integral absorbance completion module is used to invert the integral absorbance data corresponding to the missing optical path of the TDLAS measurement system when the initial integral absorbance data is acquired, based on the first temperature field distribution and the absorption spectrum physical model of the TDLAS measurement system, so as to obtain the target integral absorbance data of the target area by combining the initial integral absorbance data. The second temperature field construction module is used to output the second temperature field distribution of the target region based on the target integral absorbance data using the second CNN model; The target temperature field construction module is used to iteratively reconstruct the target temperature field distribution of the target region based on the polynomial fitting method, starting from the second temperature field distribution.

9. A computer device, characterized in that, include: At least one processor; And a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the TDLAS two-dimensional temperature field reconstruction method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the TDLAS two-dimensional temperature field reconstruction method as described in any one of claims 1 to 7.

Images