Thermal degradation studies and hybrid neural network modelling of eutectic phase change material composites

Abstract

Summary Using a thermogravimetric analyzer, the thermal stability of pure eutectic phase change material (PEPCM) (LiNO3 + NaCl) and composite eutectic PCM (CEPCM) mixture (ie, PCM containing 9% expanded graphite [EG]) was examined. PEPCM and CEPCM degradation kinetics were studied using model free kinetics methods. The activation energy of both PCM samples was evaluated using the Kissinger‐Akahira‐Sunose (KAS), Flynn‐Wall‐Ozawa (FWO), Starink, Friedman and Vyazovkin kinetic models. The calculated activation energies for Vyazovkin, Frideman, Ozawa, KAS and Starink techniques for PEPCM were 80.62‐149.2, 108.1‐180.18, 83.74‐136.17, 73.55‐127.02 and 74.64‐149.9 kJ/mol, respectively. Likewise, the activation energy of CEPCM vary between 59.4‐161.41, 83.97‐188.69, 57.1‐147.32, 54.19‐137.43 and 55‐160.65 kJ/mol. Hybrid neural networks such as ANN‐PSO and ANFIS were used to model the degradation of PCM samples. The type of PCM, the heating rate, and the temperature were applied as input parameters, while the sample's mass loss was utilized as an output parameter. The created hybrid models are capable of effectively predicting experimental TGA data.

Research on active disturbance rejection control technique for underwater welding robot based on model compensation

“…Meanwhile, to design the tracking controller of a robot by using adaptive control technology, which need not know the dynamic model of the robot, and without considering joint frictions and external disturbances (Baek, 2016;Zhang and Cheng, 2019;Baek, 2018;Tang, 2021). To control trajectory tracking, a neural network was used in some different conditions (Rahmani, 2016;Balasubramanian, 2022;Izadbakhsh and Nikdel, 2022;Uyulan and Pek, 2021;Antonio, 2019). A designed controller for the 6-DOF robot, which was using a disturbance observer to combine with sliding mode control technology (Cui, 2016;Wang et al 2020aWang et al , 2020bWu, 2020;Jiang, 2021).…”