As an important part of non-destructive testing, infrared thermography testing is widely used in various fields of industrial development for monitoring the quality of metal parts. Considering the problem of low detection rate of surface defects on steel parts, we explored the application of neural architecture search (NAS) in infrared thermography area for the first time. On the one hand, we compared different time-series temperature features of defect locations in infrared images and validate the performance of three different features such as heating, cooling and full process by machine learning methods. On the other hand, we searched for multilayer perceptron through NAS technology to classify defects with different depths. Experiments have proved that the time-series temperature feature is very effective when used in the depth classification of defects, and the accuracy rate can reach 93% under the verification of traditional machine learning methods. The NAS technique used in this paper can search 100 multilayer perceptrons in a minimum of 121s and achieve 100% defect classification accuracy.

In this paper, we focus on the inverse filtering method for restoring transient heat conduction processes. First, we address the defects of contact temperature sensor used in transient high temperature test, further we expand the distortion of transient calibration result caused by the lack of dynamic characteristics of bare thermocouple, then the inverse filtering idea is proposed. Second, we propose an improved deep belief network (DBN) to extract the characteristic of thermocouple dynamic nonlinearity on dynamic response signal. The continuous restricted Boltzmann machine (CRBM) is adopted to improve the input layer structure which the input samples can be continuous, and the back propagation (BP) neural network is adopted to realize input-output approximation at output layer. Additionally, dynamic response signal of the thermocouple is obtained through the constructed laser narrow pulse calibration system, and dynamic inverse filtering of thermocouple is realized by using the standard source as the approximation target. The performance of the inverse filtering is analyzed by comparing with the traditional linear least square(LS) method, and it is verified that this method can more accurately approximate the transient high temperature physical process in the calibration environment.

Since filament thermocouple is limited by its junction structure and dynamic characteristics, the actual heat conduction process cannot be reproduced during the transient thermal shock. In order to solve this problem, we established a thermocouple dynamic calibration system with laser pulse as excitation source to transform the problem of the restoring excitation source acting on the surface temperature of thermocouple junction into the problem of solving the one-dimensional (1D) inverse heat conduction process, proposed a two-layer domain filtering kernel regularization method for double conductors of thermocouple, analyzed the factors causing unstable two-layer domain solution, and solved the regular solution of two-layer domain by the filtering kernel regularization strategy. By laser narrow pulse calibration experiment, we obtained experimental samples of filament thermocouples with two kinds of junction structures, butt-welded and ball-welded; established error estimation criterion; and obtained the optimal filtering kernel parameters by the proposed regularization strategy, respectively. The regular solutions solved for different thermocouples were very close to the exact solution under the optimal strategy, indicating that the proposed regularization method can effectively approach the actual surface temperature of the thermocouple junction.

In high-temperature testing scenarios that rely on contact, fine-wire thermocouples demonstrate commendable dynamic performance. Nonetheless, their thermal inertia leads to notable dynamic nonlinear inaccuracies, including response delays and amplitude reduction. To mitigate these challenges, a novel dynamic error correction approach is introduced, which combines a Continuous Restricted Boltzmann Machine, Deep Belief Network, and Physics-Informed Neural Network (CDBN-PINN). The unique heat transfer properties of the thermocouple’s bimetallic structure are represented through an Inverse Heat Conduction Equation (IHCP). An analysis is conducted to explore the connection between the analytical solution’s ill-posed nature and the thermocouple’s dynamic errors. The transient temperature response’s nonlinear characteristics are captured using CRBM-DBN. To maintain physical validity and minimize noise amplification, filtered kernel regularization is applied as a constraint within the PINN framework. This approach was tested and confirmed through laser pulse calibration on thermocouples with butt-welded and ball-welded configurations of 0.25 mm and 0.38 mm. Findings reveal that the proposed method achieved a peak relative error of merely 0.83%, superior to Tikhonov regularization by −2.2%, Wiener deconvolution by 20.40%, FBPINNs by 1.40%, and the ablation technique by 2.05%. In detonation tests, the corrected temperature peak reached 1045.7 °C, with the relative error decreasing from 77.7% to 5.1%. Additionally, this method improves response times, with the rise time in laser calibration enhanced by up to 31 ms and in explosion testing by 26 ms. By merging physical constraints with data-driven methodologies, this technique successfully corrected dynamic errors even with limited sample sizes.

In high-temperature testing scenarios that rely on contact, fine-wire thermocouples demonstrate commendable dynamic performance. Nonetheless, their thermal inertia leads to notable dynamic nonlinear inaccuracies, including response delays and amplitude reduction. To mitigate these challenges, a novel dynamic error correction approach is introduced, which combines a Continuous Restricted Boltzmann Machine, Deep Belief Network, and Physics-Informed Neural Network (CDBN-PINN). The unique heat transfer properties of the thermocouple's bimetallic structure are represented through an Inverse Heat Conduction Equation (IHCP). An analysis is conducted to explore the connection between the analytical solution's ill-posed nature and the thermocouple's dynamic errors. The transient temperature response's nonlinear characteristics are captured using CRBM-DBN. To maintain physical validity and minimize noise amplification, filtered kernel regularization is applied as a constraint within the PINN framework. This approach was tested and confirmed through laser pulse calibration on thermocouples with butt-welded and ball-welded configurations of 0.25mm and 0.38mm. Findings reveal that the proposed method achieved a peak relative error of merely 0.83%, surpassing the 2.05% of the ablation technique and the 2.2% of Tikhonov regularization. In detonation tests, the corrected temperature peak reached 1045.7°C, with the relative error decreasing from 77.6% to 4.9%, and the rise time enhanced by 26 milliseconds. By merging physical constraints with data-driven methodologies, this technique successfully corrected dynamic errors even with limited sample sizes.