Digital Twin-Based Modeling of Complex Systems for Smart Aging

Abstract

In this paper, we use digital twin technology to conduct in-depth analysis and research on the modeling of complex systems for smart aging. The solution of the digital twin model based on a big data platform is proposed, and the problem of asynchronous and incomplete digital twin real-time monitoring data is solved, and the algorithm is applied to the digital twin model based on the big data platform for data preprocessing to achieve better results. To improve the real-time data transmission, the OPCUA information modeling method is optimized by using node merging and adaptive compression, which statutes the system data and achieves the effect of information fusion. The Web protocol is also used to unify the digital twin information interaction form. After experimental testing, the effectiveness of the information modeling scheme designed in this paper is verified. The fall detection results based on the digital twin selected thresholds are significantly better than the experimental results of the manually set threshold method; the thresholds set by the digital twin can more accurately identify daily behaviors, especially the more violent daily behaviors such as lying down, and the accurate alarm rate of the falling behavior reaches 92.5%. In contrast, the artificially set thresholds have a lower overall recognition rate for human behaviors and are prone to misclassification, with a misclassification rate of 3.8%. Therefore, it can be determined that using the digital twin method to set feature thresholds is better than the manual setting threshold method in terms of detection accuracy, and the digital twin method is chosen to select the thresholds for each stage of the fall process in this project. The validation results demonstrate the system's excellence in information interaction, optimization of numerical analysis, and display of results for smart aging.