Intelligent Process Monitoring of Laser-Induced Graphene Production With Deep Transfer Learning

Abstract

Three-dimensional (3D) graphene has been increasingly used in many applications due to its superior properties. The laser-induced graphene (LIG) technique is an effective way to produce 3D graphene by combining graphene preparation and patterning into a single step using direct laser writing. However, the variation in process parameters and environment could largely affect the formation and crystallization quality of 3D graphene. This paper develops a vision and deep transfer learning-based processing monitoring system for LIG production. To solve the problem of limited labeled data, novel convolutional de-noising auto-encoder (CDAE)-based unsupervised learning is developed to utilize the available unlabeled images. The learned weights from CDAE are then transferred to a gaussian convolutional deep belief network (GCDBN) model for further fine-tuning with a very small amount of labeled images. The experimental results show the proposed method can achieve the state-of-art performance of precise and robust monitoring for the quality of the LIG formation.

Portable glucose sensing analysis based on laser-induced graphene composite electrode

“…Laser-induced 3D porous patterning of graphene on PI was achieved at room temperature through continuous semiconductor lasers, which enabled highprecision preparation of LIG electrode without using harmful chemical reagents and is favorable for developing green, lowcost and high conductivity sensor devices. 36,37 Next, 3,4ethylene dioxythiophene (EDOT) monomers were electrochemically polymerized on the surface of LIG electrode to form a PEDOT lm. As a polymer material with excellent conductivity, deformability and biocompatibility, PEDOT can effectively protect the mesoporous structure of graphene 3D space, and enhances the conductivity and exibility of electrodes.…”

Design and Optimization of an Intelligent Monitoring System for Overhead Lines Based on Common Information Model

“…Each CIM package contains multiple classes, and UML's classes illustrate all the classes in the package and their relationships. When a class has a relationship with classes in VOLUME 12, 2024 other packages, those classes will also be displayed with an annotation to show which package they belong to [18] and [19]. Classes have properties that describe the characteristics of objects.…”

Reliable and Intelligent Fault Diagnosis With Evidential VGG Neural Networks

“…The present study considers both sensor noise and sensor failure, in evaluating the uncertainty estimation capability of the proposed EVGG method. First, background noise during the operation of the equipment and communication noise through the transmission channels can be presented by additive white Gaussian noise (AWGN) [22]. Second, sensor failures may have various patterns.…”