Advanced experimental-based data-driven model for the electromechanical behavior of twisted YBCO tapes considering thermomagnetic constraints

Abstract

Data-driven models can predict, estimate, and monitor any highly nonlinear and multi-variable behavior of high temperature superconducting (HTS) materials, and superconducting devices to analyze their characteristics with very high accuracy in almost real time, which is a significant figure of merit as compared with traditional numerical approaches. The electromechanical behavior of twisted HTS tapes under different strains, magnetic fields, and temperatures is a complicated problem to be solved using conventional approaches, including finite element-based methods, otherwise, experimental testing is needed to characterize it. This paper aims to offer a data-driven model based on artificial intelligence techniques to predict the electromechanical behavior of HTS tapes operating under various thermomagnetic conditions. By using the proposed model, normalized critical current value and stress of twisted tapes can be predicted under different temperatures and magnetic flux densities. For this purpose, experimental data were used as inputs to design an Adaptive Neuro-Fuzzy Inference System (ANFIS). To gain the best performance of the designed prediction system, multiple clustering methods have been used, such as the grid partitioning method, fuzzy c-means clustering method, and sub-clustering method. Sensitivity analyses were conducted to find the best architecture of ANFIS to predict and model electromechanical behavior of twisted tapes with high accuracy.

DC Electro-Magneto-Mechanical Characterization of 2G HTS Tapes for Superconducting Cable in Magnet System Using Artificial Neural Networks

“…As shown in Figure 10(a), the Jc estimation requires 5 neurons in each layer while this value for stress estimation is 3. [24] To show the capability of the proposed structure of ANN model, Table 7 is presented. In this table, firstly the 7 inputs are used to form an ANN model.…”

“…Another ANN model is also used for estimating with only 5 inputs. At last, the results are compared with the estimations of ANFIS method [24] to show the improvements that the proposed ANN with 7 inputs is offering. For Jc, the proposed 7 inputs ANN has about 10% higher R 2 value in comparison to 5 inputs ANN and 15% higher than the proposed method of [24].…”

“…At last, the results are compared with the estimations of ANFIS method [24] to show the improvements that the proposed ANN with 7 inputs is offering. For Jc, the proposed 7 inputs ANN has about 10% higher R 2 value in comparison to 5 inputs ANN and 15% higher than the proposed method of [24]. The RMSE of the 7 inputs ANN is also 10% and 26% lower than two other methods, respectively.…”

“…Last recently, AI models were used to design [21], monitor [22], and estimate [23] different characteristics and performances of superconducting devices. In [24], a novel AI approach was introduced to estimate the stress and Jc in HTS tapes. However, it did not consider the thickness of stabilisers and substrates in the model.…”

Investigation on the Electrothermal Performance of a High-Temperature Superconducting Cable in an Offshore Wind Farm Integrated Power System: Fault and Islanding Conditions

“…Cryo-electrified transportation applications [4], [5], such as HTS transformers [6], [7], superconducting fault current limiters [8]- [10], superconducting rotational machines [11], [12], and HTS cables [13]- [15]. Although the integration of HTS cables into power systems with wind farms was investigated in the literature, most of the related papers have focused on investigating the steady-state characteristic of HTS cables, affected by the presence of wind farms [16]- [18].…”