Hongye Zhang | Searcharxiv

arXiv2024

Optimisation and Loss Analyses of Pulsed Field Magnetisation in a Superconducting Motor with Cryocooled Iron Cores

A 2D electromagnetic-thermal coupled numerical model has been developed using the finite element method and validated against experimental data to investigate a superconducting machine featuring high-temperature superconducting (HTS) tape stacks and cryocooled iron cores. The HTS stacks are transformed into trapped field stacks (TFSs) through pulsed field magnetisation (PFM), generating rotor fields. After PFM, the superconducting motor operates on the same principle as permanent magnet synchronous motors. This study explores the behaviour of HTS stacks by altering the stack's layer number from one to nine and adjusting the pulsed current amplitude from 250 A to 1000 A. The primary objective of this paper is to identify the optimal combination of pulsed current amplitudes and TFS layer numbers for achieving maximum magnetisation fields. The secondary objective is to evaluate the overall losses in both superconducting and non-superconducting parts of the machine during magnetisation, including heat generated in various layers of the TFS, and losses in the motor's active materials (copper windings and iron cores). Two motor configurations were proposed, and two calculation methods using linear interpolation of iron losses and steel grades were introduced to estimate the iron losses for the studied iron material, M270-35A. This pioneering study is expected to serve as a valuable reference for loss analyses and structural design considerations in developing superconducting machines.

cond-mat.supr-con↗

arXiv2020

VELC: A New Variational AutoEncoder Based Model for Time Series Anomaly Detection

Anomaly detection is a classical but worthwhile problem, and many deep learning-based anomaly detection algorithms have been proposed, which can usually achieve better detection results than traditional methods. In view of reconstruct ability of the model and the calculation of anomaly score, this paper proposes a time series anomaly detection method based on Variational AutoEncoder model(VAE) with re-Encoder and Latent Constraint network(VELC). In order to modify reconstruct ability of the model to prevent it from reconstructing abnormal samples well, we add a constraint network in the latent space of the VAE to force it generate new latent variables that are similar with that of training samples. To be able to calculate anomaly score in two feature spaces, we train a re-encoder to transform the generated data to a new latent space. For better handling the time series, we use the LSTM as the encoder and decoder part of the VAE framework. Experimental results of several benchmarks show that our method outperforms state-of-the-art anomaly detection methods.

cs.LG↗