Temperature Field Reconstruction Research Articles

Perception and reconstruction of temperature field in forgings based on physical model and CNN model

Accurately perceiving and reconstructing the temperature field of hot-deformed large forgings is essential for intelligent forging and enhancing forging quality. In this study, the surface temperature of large forgings is measured using an infrared thermal imager. The relationship between the grayscale of infrared images and the surface temperature of large forgings is analyzed. A physical model, incorporating a convolutional neural network (CNN) model, is established to ascertain the surface temperature of large forgings. The CNN model utilizes standard convolution and multilayer perceptron modules for the backbone network and implements an adaptive learning rate. The correlation between three channel information and the surface temperature besides the grayscale is considered in the CNN model. Lastly, this study proposes a method for reconstructing the temperature field of large forgings based on the developed perception model, enabling accurate surface temperature measurement even when the large forgings are covered by oxide skin.

High-Resolution Reconstruction of Temperature Fields Based on Improved ResNet18.

High-precision measurement of temperature value distributions in production scenarios is of great significance for industrial production, but traditional temperature field reconstruction algorithms rely on the design of manual feature extraction methods with high computational complexity and poor generalization ability. In this paper, we propose a high-precision temperature field reconstruction algorithm based on deep learning, using an efficient adaptive feature extraction method for temperature field reconstruction. We design an improved temperature field reconstruction algorithm based on the ResNet18 neural network; introduce the CBAM attention mechanism in the model; and design a feature pyramid, using M-FPN, a multi-scale feature aggregation network fusing PAN and FPN, to make the extracted feature information propagate multi-dimensionally among different layers to improve the feature characterization ability. Finally, the mean square error is used to guide the model to optimize the training so that the model pays more attention to the data and reduces the large error to ensure that the gap between the predicted value and the real value is small. The experimental results show that the reconstruction accuracy of the improved algorithm presented in this paper is significantly better than that of the original algorithm in the case of typical peaked temperature field distributions.

Multispectral two-dimensional temperature field reconstruction using particle swarm optimization and the Broyden–Fletcher–Goldfarb–Shanno algorithm with multiple extreme value optimization

The flame temperature is a crucial parameter in combustion, indicating heat generation and transfer. However, determining the correct temperature becomes challenging when flame emissivity is unknown. This article presents a multispectral 2D temperature field reconstruction method using particle swarm optimization and the Broyden–Fletcher–Goldfarb–Shanno algorithm, combined with multivariate extreme-value optimization. This method establishes an objective function based on the correlation between the true temperature and spectral data. The objective function was minimized by adjusting the emissivity, enabling the reconstruction of 2D temperature and emissivity images without assuming an emissivity model. Calibration with a multispectral camera yielded the spectral response coefficients, and three-time spline interpolation was used to verify the calibration data. The reconstructed 2D temperature field showed a mean absolute error (MAE) of 4.61 and structural similarity (SSIM) of 0.995. The reconstructed emissivity image had an MAE range of 0.04 to 0.053 and an SSIM range of 0.897 to 0.915. The system, tested on a butane flame, showed a temperature range of 1001 K to 1356 K, with an average temperature of 1194 K and emissivity values ranging from 0.3166 to 0.8621. The results align with the expected combustion process and radiation characteristics distribution of the butane flame.

C-type two-thermocouple sensor design between 1000 and 1700 °C.

At the present stage of transient ultra-high-temperature energy release of boron-containing warm-pressure explosives, single thermocouples are often used for multi-point measurements in the process of their temperature field changes, and the results of their temperature field reconstruction are not satisfactory due to the limited consistency of the thermocouples. Aiming at the above-mentioned problems, a C-type two-thermocouple suitable for transient temperature measurement in high-temperature environments is designed; the system characterization of the two-thermocouple is carried out by using the blind system identification method of the inter-relationships; the identification process is evaluated by a new cost function; and the optimal solution on the new cost function is realized by using the gradient descent method. The temperature reconstruction of the two-thermocouple output excited by the simulated heat source is carried out by using the auto-regressive with extra inputs model, and the feasibility of the reconstruction results is verified. In the experimental part, a thermocouple dynamic characteristic calibration system based on a high-temperature furnace is constructed and experimental validation is carried out in the high-temperature furnace to compare the effects of different exposure lengths and different wire diameters on the output of the two-thermocouple, and as a result, the outputs are corrected and analyzed. The results show that two-thermocouple methods with different combinations of wire diameters are better for temperature measurement, with a reconstructed root mean squared error of 0.0162 and a goodness of fit of 89.13%.

A physics-driven sensor placement optimization methodology for temperature field reconstruction

Perceiving the global field from sparse sensors has been a grand challenge in the monitoring, analysis, and design of physical systems. In this context, sensor placement optimization is a crucial issue. Most existing works require large and sufficient data to construct data-based criteria, which are intractable in data-free scenarios without numerical and experimental data. To this end, we propose a novel physics-driven sensor placement optimization (PSPO) method for temperature field reconstruction using a physics-based criterion to optimize sensor locations. In our methodological framework, we firstly derive the theoretical upper and lower bounds of the reconstruction error under noise scenarios by analyzing the optimal solution, proving that error bounds correlate with the condition number determined by sensor locations. Furthermore, the condition number, as the physics-based criterion, is used to optimize sensor locations by the genetic algorithm. Finally, the best sensors are validated by reconstruction models, including non-invasive end-to-end models, non-invasive reduced-order models, and physics-informed models. Experimental results, both on a numerical and an application case, demonstrate that the PSPO method significantly outperforms random and uniform selection methods, improving the reconstruction accuracy by nearly an order of magnitude. Moreover, the PSPO method can achieve comparable reconstruction accuracy to the existing data-driven placement optimization methods.

An Extremely Sparse Tomography Reconstruction of a Multispectral Temperature Field without Any a Priori Knowledge.

When undertaking optical sparse projection reconstruction, the reconstruction of the tested field often requires the utilization of a priori knowledge to compensate for the lack of information due to the sparse projection angle. In order to reconstruct the radiation field of unknown materials or in situations where a priori knowledge cannot be obtained, this paper proposes an extremely sparse tomography multispectral temperature field reconstruction algorithm that analyzes the similarity (the similarity here compares and calculates the Euclidean distance of the spectral emissivity values at various wavelengths between different spectral curves) of radiation characteristics of materials under the same pressure and concentration but different temperature, describes the similarity between the radiation information of the tested field using the dynamic time warping (DTW) algorithm, and uses the similarity sum of the radiation information among the subregions of the temperature field as the optimization objective. This is combined with the equation-constrained optimization algorithm and multispectral thermometry to establish the statistical law between the missing information and finally realize the reconstruction of the temperature field. Simulation experiments show that, without any a priori knowledge, the method in this paper can realize reconstruction of the temperature field with an accuracy of 1.53-12.05% under two projection angles and has fewer projection angles and stronger robustness than other methods.

Temperature field reconstruction of hot blast furnace shell based on DE-RBF hybrid algorithm

Temperature field reconstruction of hot blast furnace shell based on DE-RBF hybrid algorithm

3D Reconstruction of flame temperature field based on lightweight residual network with spatial attention mechanism

3D Reconstruction of flame temperature field based on lightweight residual network with spatial attention mechanism

2D temperature field reconstruction using optimized Gaussian radial basis function networks

Monitoring temperature fields by sensors and mapping the reconstructed temperature field holds great importance in numerous contexts. Concurrently, as sensing technologies continue to advance, there is a parallel development of temperature reconstruction algorithms. In this study, we present an enhanced Quadratic Optimization Gaussian Radial Basis Function approximation (QO-GRBF). We demonstrate the effectiveness of this algorithm by reconstructing a 10 cm x 10 cm temperature field using distributed optical fiber sensing system. The absolute error across four different temperature stages from 35 °C to 65 °C ranged from 0.21 °C to 0.45 °C, indicating a strong reconstruction ability. Furthermore, we compared the performance of QO-GRBF with GRBF in different point densities. The results reveal an over 30 % improvement in scenarios with denser data points, highlighting the efficacy of proposed algorithm. The data size chosen was quantified as well. There was a 10.22 % improvement observed when transitioning from selecting 4 points to selecting 6 points.

Surface temperature field real-time reconstruction of hot forging die based on 1DCNN

Accurate reconstruction of the die surface temperature field during the hot forging process is crucial for ensuring optimal control of die temperature and high-quality metal forgings. In this study, we propose a novel method that combines hot forging process parameters with a deep neural network to reconstruct the temperature field on the die surface in real-time. Firstly, a high-fidelity finite element model of the hot forging process is created and validated with a forging experiment. Secondly, temperature observation points are initially selected based on the simulated temperature distribution of the forging die, and the correlation coefficient method is used to optimize the layout of observation points to improve reconstruction accuracy and efficiency. Subsequently, multiple high-quality simulations with varying process parameters are conducted, generating a comprehensive dataset. The 1DCNN algorithm is then presented to produce a mapping from the input to the surface temperature field. Finally, the effectiveness of the surface temperature field reconstruction method is verified. The results indicate that the overall performance of the proposed 1DCNN algorithm is better than the traditional regression algorithms. In terms of reconstruction accuracy, it achieves a prediction accuracy within an absolute error of 3 °C for the outer surface points and within a relative error of 5 % for the cavity region. Regarding reconstruction efficiency, it requires an approximate execution time of 0.064 s for predicting a new sample. This paper presents a method for real-time reconstruction of the die surface temperature field, leveraging hot forging process parameters and temperature data from observation points. This approach enables reliable online monitoring of transient temperature variations on the hot forging die surface, offering valuable support for operational control and maintenance.

Modified reconstruction of boiler numerical temperature field based on flame image feature extraction

ABSTRACT Computational fluid dynamics (CFD) method is a common method to obtain the temperature field distribution within the furnace. However, in order to reduce computational complexity, the numerical simulation involve simplifications in boundary conditions and model parameters. As a result, the temperature field distribution deviates from the actual operating conditions. This paper proposes a numerical temperature field modification method based on flame images. Flame images corresponding to the operational conditions are collected using the Industrial Flame Monitoring System (IFMS). The flame images are preprocessed, and the contour of the flame’s core region is extracted using the Mask Region-based Convolutional Neural Network (Mask R-CNN) method. The geometric features of the flame are extracted, and matrix calculations are applied to modify the numerical temperature field. The modified temperature field exhibits a deviation in the combustion center, aligning with the actual operation of the boiler. A comparison between the modified temperature field and the temperature measurements from flue gas shows consistent temperature trends. The absolute error (AE) under different operating conditions is under 8.8 K, and the relative error (RE) remains below 1.3%. The analysis results demonstrate that it can enhance the accuracy of temperature field calculations by the modification from flame image features.

A Deep Transfer Operator Learning Method for Temperature Field Reconstruction in a Lithium-Ion Battery Pack

A Deep Transfer Operator Learning Method for Temperature Field Reconstruction in a Lithium-Ion Battery Pack

Real-time reconstruction of 3D transient non-uniform temperature field for thermal protection system based on machine learning

Reconstructed temperature field will provide critical input for monitoring the aerodynamic heating on the thermal protection system (TPS). In this paper, a novel data-driven method is proposed that allows real-time reconstruction of the 3D transient non-uniform temperature field in a TPS ceramic tile. The measured temperature is obtained by embedding sparse sensors at different layers of the tile. The spatial feature bases and time projection coefficients of the non-uniform temperature fields are extracted using the multi-parameter proper orthogonal decomposition (POD) method. The Fully-connected Long Short-Term Memory (FC-LSTM) networks are used to establish the prediction relationship between the measurements and the full-field temperatures. The problem to estimate the projection coefficients of the spatial feature bases is solved by optimizing the loss function using Adam algorithm. The numerical examples show an approximation error less than 5% and the reconstruction time-consuming less than 1 s in 3D temperature field reconstruction (TFR). The proposed reconstruction method is robust at different measurement noise levels and tested point arrangements.

Direct reconstruction of tissue temperature fields during thermal therapy by dynamic response matching and steady-state correction

In the field of oncology thermotherapy technology, online and real-time reconstruction of biological tissue temperature field is a fundamental work with important practical needs. In this paper, we proposed a new method for direct reconstruction of biological tissue temperature field by dynamic response feature matching and steady-state correction. Based on the bioheat transfer equation, the step response model of tissue temperature field during thermal therapy is established, and the dynamic response eigenvectors of biological tissue temperature and the steady-state gain of temperature response are determined offline. According to the optimal matching principle between the temperature dynamic response of the observation point and the internal space point of tissue, the optimal matching matrix of the dynamic response features between the observation point temperature and the internal temperature field of tissue is determined with the use of the dynamic response eigenvectors. Further, through the weighted synthesis of observation point temperature response by using the optimal matching matrix of dynamic response features, the tissue transient temperature field pre-estimation model is established. Finally, by correcting the tissue temperature field pre-estimation model with steady-state gain correction factor, the direct reconstruction scheme of temperature field based on dynamic response feature matching and steady-state gain correction is established. In this work, the non-destructive reconstruction of the internal temperature field of biological tissues during laser-induced thermal therapy is achieved based on the above direct reconstruction scheme. Numerical simulation experiments are performed to investigated the effects of pre-estimation time domain, the number of measurement points, laser irradiation form and measurement error on the temperature field reconstruction results.

Temperature and Acoustic Field Reconstruction in Fiber Optic DAS System Based on Variational Mode Decomposition

Temperature and Acoustic Field Reconstruction in Fiber Optic DAS System Based on Variational Mode Decomposition

Temperature field reconstruction based on femtosecond laser prepared of multiwavelength fibre bragg grating array

Temperature field measurements are of great significance in industrial production, scientific research, and other fields. Contact temperature measurements usually achieve high spatial resolution by increasing the layout density of the sensors. However, in practical applications, achieving high-density sensor placement is often difficult. Therefore, when the number of sensors is limited, it is necessary to perform function fitting or interpolation of the temperature of the sampling points to reconstruct the temperature field distribution. This study proposed a temperature field reconstruction method based on femtosecond laser prepared multiwavelength fibre Bragg grating (FBG) array. A multiwavelength FBG array was prepared using the femtosecond laser phase mask method combined with stress stretching, and applied for measuring the temperature field distribution in a tubular high-temperature furnace. Cubic spline interpolation and backpropagation (BP) neural networks were used to construct two-dimensional temperature field models for the temperature field distribution data measured by the FBGs, and the prediction accuracies of the two models were compared. The test results show that the root mean square error of the temperature field distribution constructed using the BP neural network is 0.7333 °C, which is approximately 23.18% of the predicted results of the cubic spline interpolation model, indicating that it is a high-precision temperature field reconstruction method.

A real-time solution method for three-dimensional steady temperature field of transformer windings based on mechanism-embedded cascade network

To enhance the computation efficiency and accuracy of three-dimensional steady temperature field of transformer windings, we propose a new non-invasive Reduced Order Model (ROM) based on a mechanism-embedded cascade network. Initially, a snapshot matrix is formed from the Full Order Model (FOM) and then combined with Proper Orthogonal Decomposition (POD) to extract key modal features that characterize the temperature field. Subsequently, a cascade network architecture, integrating Multilayer Perceptron (MLP) and Radial Basis Function Neural Network (RBFNN), is devised to swiftly map working condition parameters to modal coefficients. Additionally, the cascade network is embedded with condition sensitivity and modal contribution mechanisms to further enhance prediction accuracy. Finally, by linearly weighting the modes with predicted modal coefficients, a rapid reconstruction of the steady temperature field in transformer windings is achieved. Validation against Fluent software simulations and experimental measurements demonstrate a close agreement, with computational errors of less than 4K and an impressive single solution time of only 0.0087 s, which is 48760 times faster compared to Fluent software.

Multi-View Synthesis of Sparse Projection of Absorption Spectra Based on Joint GRU and U-Net

Tunable diode laser absorption spectroscopy (TDLAS) technology, combined with chromatographic imaging algorithms, is commonly used for two-dimensional temperature and concentration measurements in combustion fields. However, obtaining critical temperature information from limited detection data is a challenging task in practical engineering applications due to the difficulty of deploying sufficient detection equipment and the lack of sufficient data to invert temperature and other distributions in the combustion field. Therefore, we propose a sparse projection multi-view synthesis model based on U-Net that incorporates the sequence learning properties of gated recurrent unit (GRU) and the generalization ability of residual networks, called GMResUNet. The datasets used for training all contain projection data with different degrees of sparsity. This study shows that the synthesized full projection data had an average relative error of 0.35%, a PSNR of 40.726, and a SSIM of 0.997 at a projection angle of 4. At projection angles of 2, 8, and 16, the average relative errors of the synthesized full projection data were 0.96%, 0.19%, and 0.18%, respectively. The temperature field reconstruction was performed separately for sparse and synthetic projections, showing that the application of the model can significantly improve the reconstruction accuracy of the temperature field of high-energy combustion.

Joint identification method for thermal–mechanical coupling loads

The crucial role in evaluating structural fatigue life, implementing health monitoring, and optimizing lightweight and reliable design is played by load identification technology. With the development of the aviation industry towards high-speed aircraft, the challenges under high temperature and high load conditions have been intensified. Unfortunately, in environments characterized by the coexistence of force and heat, difficulties are faced by existing load identification methods in achieving key decoupling and accurate identification of force and temperature fields. To address this gap, the Joint Identification Method (JIM), a new method that integrates force and thermal strain patterns, is introduced in this article. This method has been developed through in-depth research on “joint physical fields” and “joint basis functions.” Multiple physical fields for load identification are systematically decomposed by our method, each field is theoretically deduced, and they are seamlessly reassembled using the same structural spatial field. Through the method, mechanical strain and thermal strain are separated in high-precision temperature field reconstruction, and a basis function method for identifying the full field load distribution under a few measurement points is obtained. To verify the accuracy of our proposed Joint Identification Method (JIM), five experiments were designed, and the results consistently demonstrated the superiority of our method in accuracy (with an error always below 5%), efficiency (avoiding a large number of redundant calculations in traditional Duhamel methods), and reliability in static and dynamic load static heat identification. In the end, accurate, efficient, and reliable identification of the structural strain field, structural load field, and temperature field were achieved by our method.

Flame soot absorption coefficient and temperature reconstruction through line-of-sight attenuation and light field imaging

The accuracy and efficiency of 3-D flame temperature field reconstruction using the light field imaging technique heavily depend on soot radiation characteristics. In this study, we employ the line-of-sight attenuation method to reconstruct the soot absorption coefficient distribution in a pure absorbing flame. Utilizing these soot absorption coefficients, the radiative transfer equation is effectively transformed within the framework of the light field imaging technique into a linear inverse problem and also outlines the flame boundary. This proposed strategy reduces the unwanted detection rays significantly, thus eliminating the extensive computational processing. Consequently, the proposed approach substantially enhances the accuracy and efficiency of flame temperature reconstruction. Numerical simulations were carried out on a bimodal asymmetric flame to validate the noise tolerance capabilities, investigate the effects of varying voxel numbers on flame division and carry out a comparative study. Experimental studies were also conducted to reconstruct flame temperature and soot absorption coefficient distributions under different combustion operating conditions. Thermocouple measurements were performed and compared with the reconstructed temperatures. Outcomes achieved from both numerical simulations and experimental studies demonstrate the feasibility, accuracy and robustness of the proposed method.