A Sparse Autoencoder and Softmax Regression based Diagnosis Method for the Attachment on the Blades of Marine Current Turbine

Abstract

The development and application of marine current energy are attracting more and more attention in the world. Due to the hardness of its working environment, it is important and difficult to study the fault diagnosis of marine current generation system. In this paper, underwater image is chosen as the fault diagnosing signal after different sensors are compared. This paper proposes a diagnosis method based on the sparse autoencoder (SA) and softmax regression (SR). The SA is used to extract the features and SR is used to classify them. Images are used to monitor whether the blade is attached by benthos and to determine its corresponding degree of attachment. Compared with the other techniques, experiment results show that the proposed method can diagnose the blade attachment with higher accuracy.

Applications of machine learning to machine fault diagnosis: A review and roadmap

“…Some publications [310,[317][318][319][320][321][322][323][324][325][326][327][328][329][330] have introduced AE and its common varieties into machine fault diagnosis. Among them, Jia et al [310] used the stacked AE to automatically learn features from the frequency-domain data and subsequently complete the diagnosis tasks of rolling element bearings and gears, which was one of the earliest studies in applications of stacked AE.…”

A Mini Review of Research on Blade Fault Detection of Marine Current Turbine

“…ACET.000596. 4(5).2021 as valuable fault detection information [10]. The reference [11] proposed a diagnosis method based on a deep separable convolutional neural network for the biofouling on the blades of MCTs.…”

Multi‐Channel Domain Adaptation Deep Transfer Learning for Bridge Structure Damage Diagnosis

“…Otherwise, visual images of some damages are not as easy to obtain as vibration data, which results in very expensive. Compared with extracting features from the raw data, the features extracted from the conversion map may ignore some useful conditional information and increase the feature distribution discrepancy between different domains [22,23]. Therefore, the transfer damage diagnosis between bridges is particularly important.…”