A machine learning based calibration method for differential scanning calorimetry

Aohua Xiao, Zhengyang Zhang, Zhongxin Qian, Ao Xiao, Dongfang Hu, Jinxin Xu, Xudong Lu, Mohammad Mahdi Heyhat, Jiong Ding,
A machine learning based calibration method for differential scanning calorimetry,
Thermochimica Acta,
Volume 754,
2025,
180158,
ISSN 0040-6031,
https://doi.org/10.1016/j.tca.2025.180158.
(https://www.sciencedirect.com/science/article/pii/S0040603125002333)
Abstract: Due to temperature differences, heating non-uniformity, and sensor drift, the measurements of differential scanning calorimetry (DSC) usually require calibration; however, traditional calibration methods are difficult to apply under different experimental conditions. To solve the problem, a neural network-based calibration method is proposed in this article. First, a second-order heat transfer model of DSC is established, and its temperature difference is calculated using the sapphire. Second, a neural network is developed to model the relationship between the experimental and model-based temperature differences. Finally, the neural network is trained on the experimental data. The resulting model is then validated using experimental data obtained under different conditions. The results demonstrate that the machine-learning-based calibration method significantly improves the efficiency and accuracy of DSC calibration.
Keywords: Differential scanning calorimetry; Neural network calibration; Heat transfer modeling; Correction