Fuel cells are electrochemical devices that convert the chemical energy of a fuel and an oxidizing agent such as oxygen into electrical energy through a chemical reaction. They are considered a promising and environmentally friendly technology for generating electricity, particularly to power vehicles, homes and portable electronics.
However, microdefects on fuel cell surfaces can have various implications depending on their size, nature and location. These defects can include imperfections, irregularities or anomalies in the materials that make up the fuel cell components, such as the electrodes, electrolyte and catalyst layers. Microdefects disrupt the smooth flow of ions and electrons within the fuel cell. As a consequence, the cell resistance increases and the overall efficiency and power output of the cell are reduced.
The traditional method to detect these defects is by Scanning Electron Microscopy (SEM). This is information about the morphology and topography of the surface to identify defects. Researchers at the Korea Research Institute of Standards and Science have developed a technology based on deep learning techniques that enables real-time 3D measurements using a single-point pattern projection method.
Its single-shot deflectometer method uses a high carrier frequency pattern. However, the visibility of the fringe pattern captured using these methods is not feasible when this pattern is projected onto a metal surface with low polish quality, such as battery fuel. Due to low reflectivity, the quality of the captured image could be better and the phase cannot be recovered correctly. Many surfaces with highly deformed levels generate complex reflected fringe patterns that include closed-loop and open-loop features, demonstrating a low-frequency composite pattern from which phase recovery is difficult.
To overcome this limitation, the team created an ai algorithm for the pattern projection method inspired by the DL technique in optical meteorology. They used DYnet++, trained with measurement data on thousands of surface shapes. This allows DYnet++ to perform real-time 3D morphology measurements of surfaces with low reflectivity or complex shapes. They added more convolution layers to the Ynet model based on the Unet++ architecture to generate a nested DYnet++ or Y-net model. Basically, the concept they propose is a standard encoder and decoder block to help the network learn better from marginal patterns.
Getting a good training data set is essential in every DL task to ensure the best result. Deflectometry training data can be generated by simulation and experimentally. However, the simulation data will only partially reflect the actual physical imaging process. This will lead to a problem with very good results with the simulation data but not good experimental results. They designed a deformable mirror (DM) to quickly obtain experimental training data. It is a specialized optical device used in adaptive optics systems to correct distortions and aberrations in incoming light.
In conclusion, the strong and novel point of their proposed method is that even when the surface has low reflectivity and a very complex topology that could generate marginal closed and open loop patterns together, their DL network can still measure them in seconds. The model could predict results quickly and automatically without human intervention. This is extremely useful in speeding up the manufacturing process of these surfaces in modern industry.
Review the Paper. All credit for this research goes to the researchers of this project. Also, don’t forget to join. our SubReddit of more than 30,000 ml, Facebook community of more than 40,000 people, Discord channel, and Electronic newsletterwhere we share the latest news on ai research, interesting ai projects and more.
If you like our work, you’ll love our newsletter.
Arshad is an intern at MarktechPost. He is currently pursuing his international career. Master’s degree in Physics from the Indian Institute of technology Kharagpur. Understanding things down to the fundamental level leads to new discoveries that lead to the advancement of technology. He is passionate about understanding nature fundamentally with the help of tools such as mathematical models, machine learning models, and artificial intelligence.
<!– ai CONTENT END 2 –>