A New Convolutional Neural Network With Random Forest Method for Hydrogen Sensor Fault Diagnosis

Abstract

Hydrogen is considered to be a hazardous substance. Hydrogen sensors can be used to detect the concentration of hydrogen and provide an ideal monitoring means for the safe use of hydrogen energy. Hydrogen sensors need to be highly reliable, so fault identification and diagnosis for gas sensors are of vital practical significance. However, traditional machine learning methods for fault diagnosis are based on features extracted by experts, prior knowledge requirements and the sensitivity of system changes. In this study, a new convolutional neural network (CNN) using the random forest (RF) classifier is proposed for hydrogen sensor fault diagnosis. First, the 1-D time-domain data of fault signals are converted into 2-D gray matrix images; this process does not require noise suppression and no signal information is lost. Secondly, the features of the gray matrix images are automatically extracted by using a CNN, which does not rely on expert experience. Dropout and zero-padding are used to optimize the structure of the CNN and reduce overfitting. Random forest, which is robust and has strong generalization ability, is introduced for the classification of gas sensor signal modes, in order to obtain the final diagnostic results. Finally, we design and implement a prototype hydrogen sensor array for experimental verification. The accuracy of fault diagnosis in hydrogen sensors is 100% under noisy environment with the proposed method, which is superior of CNN without RF and other methods. The results show that the proposed CNN with RF method provides a good solution for hydrogen sensor fault diagnosis. INDEX TERMS Fault diagnosis, hydrogen sensor, convolutional neural network, random forest, feature extraction.

A New Hydrogen Sensor Fault Diagnosis Method Based on Transfer Learning With LeNet-5

“…The typical LeNet-5 structure consists of two alternating convolutional layers, two pooling layers, and the two-layer FC artificial neural network. Compared with Alenet, GoogLenet, VGG16, and other CNN algorithms, LeNet-5 method has simple structure and high accuracy (Wen et al, 2018;Lu et al, 2019), and has achieved good results in hydrogen sensor fault diagnosis (Sun et al, 2020). Therefore, this study adpots LeNet-5 as classifier.…”

“…There are many CNN models; for example, GoogLeNet (Szegedy et al, 2015), AlexNet (Krizhevsky et al, 2017), andLeNet-5 (LeCun, 2015). As a classic CNN, LeNet-5 is widely used for handwritten digital character recognition (Tivive and Bouzerdoum, 2005) and fault diagnosis (Wen et al, 2018;Sun et al, 2020). LeNet-5 is a CNN with a gradient-based learning structure, and its input layer is an image with a size of 32× 32 pixels.…”

“…Song et al proposed a fault diagnosis and reconfiguration strategy for hydrogen sensor array in 2019 (Song et al, 2019). Sun et al proposed a new convolutional neural network method for hydrogen sensor fault diagnosis in 2020 (Sun et al, 2020).…”

“…As a branch of ML, deep learning (DL) has powerful functionality and flexibility. DL does not need to rely on expert experience and manual feature extraction (Zhang W. et al, 2018), so many scholars have gradually introduced DL methods, such as the deep belief network (Shao et al, 2018;Wang et al, 2020), sparse autoencoders, and convolution neural networks (CNNs) (Wen et al, 2018;Wu and Zhao, 2018;Gou et al, 2020;Sun et al, 2020) into fault diagnosis processes. These methods can improve the accuracy of fault diagnosis, but there are some limitations.…”

Hydrogen Leakage Diagnosis for Proton Exchange Membrane Fuel Cell Systems: Methods and Suggestions on Its Application in Fuel Cell Vehicles

“…Chauhan et al [23] reviewed different sensing technologies used in existing hydrogen concentration sensors and various materials used in hydrogen sensing. In addition, Sun et al [122] proposed a data-driven diagnosis method based on the convolutional neural network and the random forest for different failures of hydrogen concentration sensors (such as impact fault, stuck fault, etc. )…”