多领域动力学系统状态序列仿真数据集

多领域动力学系统对象状态序列仿真数据包含来自物理学、流体、生物、气象等多个领域的动力学的观测对象状态，根据各系统的动力学微分方程，设置不同的参数为具有交互关系的观测对象生成同步演化的状态序列。对于每个动力学，根据其微分方程设置不同的系统特定参数组合、交互观测对象数量、序列观测长度，来设置不同的参数组合作为不同系统设定。对于每个系统，随机初始化每个对象的初始状态，并根据微分方程生成各对象的观测序列，该共同演化的观测序列与各对象间的交互关系图组成一个样本。通过随机设置多种初始状态，为每个系统生成多组样本。

This multi-domain dynamical system object state sequence simulation dataset contains observed states of dynamical system objects from diverse domains including physics, fluid dynamics, biology, meteorology, and others. By leveraging the dynamical differential equations of individual systems, diverse parameters are configured to generate synchronously evolving state sequences for interacting observed objects. For each type of dynamics, distinct system settings are defined by varying three aspects: system-specific parameter combinations, the number of interacting observed objects, and the length of the observation sequence, based on its corresponding differential equation. For each system setting, the initial state of each observed object is randomly initialized, and the observation sequence for each object is generated according to the differential equation. Each sample consists of this co-evolving observation sequence and the interaction graph between the observed objects. Multiple samples are generated for each system setting by randomly setting various initial states.