The construction of digital twin for reliability of liquid level volatile products

Abstract

Abstract Due to the lack of air tightness and environmental stress, the liquid level volatile products will gradually volatilize, which will affect the performance and reliability of products. Due to the small volume, complex working environment, strong acidity of electrolyte and various electric environments, it is difficult to monitor the liquid level in real time, so it is impossible to predict the remaining life of such products. In view of this, this paper proposes a scheme to estimate the liquid level information and predict the remaining life of volatile liquid level products through reliability digital twin. First, a set of test device is designed to control the environmental factors of liquid level volatile products in real time and collect the information of liquid level height. Then, according to the product working environment and liquid level data collected by the test device in real time, a general mathematical model of liquid level volatile of products is fitted. According to the model, the twin mapping of this kind of products in virtual space can be built, which can make them have liquid level volatile products surplus life prediction ability. Finally, the digital twin real-time collects the actual use data of the product, and continuously evolves the remaining life prediction model of the product through learning, so as to improve the prediction accuracy and better guarantee the reliable use of the product.