A Survey on Uncertainty Quantification in Dynamical Systems
本文综述了动态系统中不确定性量化的方法，分为前向传播和后向传播两类，帮助工程、物理、技术和医疗领域的研究者应对不确定性。
There is uncertainty in various domains, requiring a robust understanding and quantification in order to make informed decisions. This article presents a comprehensive discussion of uncertainty quantification (UQ) in dynamical systems. Beginning with an overview, we examine the multifaceted nature of uncertainty and the challenges in its modeling and analysis. Identifying key sources that shape uncertainty is essential for accurate quantification. Whether models are deterministic or stochastic, statistical inference remains crucial for robust modeling and decision-making in UQ. This article introduces a state-of-the-art classification of UQ methods, comprising two fundamental categories: 1) forward propagation techniques and 2) backward propagation techniques, each designed to address specific challenges in UQ. By equipping readers with essential tools for managing uncertainty, this article aims to broaden the application of UQ across engineering, physics, technology, and healthcare.