Visual Inspection System for Uncoupling Status of Control-Rod Drive Rod in Nuclear Reactor.

Under some unexpected conditions, drive rods and control-rod assemblies may not be disconnected. If this situation is not detected, the control rod will be lifted out of the reactor core together with the upper reactor internals. This situation will seriously affect the follow-up work and reduce the economy and safety protection of the nuclear power plant. To ensure safety, the tripping status must be checked after tripping. Follow-up work can be carried out after checking and confirming that all drive rods are in the tripping status. There are many problems for traditional inspection methods, such as misjudgment, low accuracy, and labor consumption. This paper proposes a visual inspection system for the uncoupling state of the control-rod drive rod of the nuclear reactor. The proposed method is based on the fitting model of the ellipse parameter of the drive-rod head and the height of the drive rod. The ellipse of the drive-rod head is firstly accurately detected. Then, a mathematical model between the ellipse parameter and the height of the drive rod is established. The measurement error caused by the swing of the head of the drive rod is eliminated. The accurate measurement of the height difference before and after the tripping of the drive rod is computed. Finally, the status of the uncoupling of the drive rod is judged according to the difference. Many experiments are carried out with our developed system. The experimental results show that the proposed system realizes remote operation, ensures the quality of trip-status inspection, improves work efficiency, and reduces the workload of staff.

MAMA Net: Multi-Scale Attention Memory Autoencoder Network for Anomaly Detection.

Anomaly detection refers to the identification of cases that do not conform to the expected pattern, which takes a key role in diverse research areas and application domains. Most of existing methods can be summarized as anomaly object detection-based and reconstruction error-based techniques. However, due to the bottleneck of defining encompasses of real-world high-diversity outliers and inaccessible inference process, individually, most of them have not derived groundbreaking progress. To deal with those imperfectness, and motivated by memory-based decision-making and visual attention mechanism as a filter to select environmental information in human vision perceptual system, in this paper, we propose a Multi-scale Attention Memory with hash addressing Autoencoder network (MAMA Net) for anomaly detection. First, to overcome a battery of problems result from the restricted stationary receptive field of convolution operator, we coin the multi-scale global spatial attention block which can be straightforwardly plugged into any networks as sampling, upsampling and downsampling function. On account of its efficient features representation ability, networks can achieve competitive results with only several level blocks. Second, it's observed that traditional autoencoder can only learn an ambiguous model that also reconstructs anomalies "well" due to lack of constraints in training and inference process. To mitigate this challenge, we design a hash addressing memory module that proves abnormalities to produce higher reconstruction error for classification. In addition, we couple the mean square error (MSE) with Wasserstein loss to improve the encoding data distribution. Experiments on various datasets, including two different COVID-19 datasets and one brain MRI (RIDER) dataset prove the robustness and excellent generalization of the proposed MAMA Net.