经典细胞自动机是一类显式的离散时空格点模型，其复杂的大尺度现象源于简单的确定性规则。为揭示不同物理上相异的遍历行为类别，我们对三态细胞自动机（含一个稳定的‘真空’态及带±电荷的‘粒子’）开展了系统性研究。该分类借助自动机在离散对称性（电荷共轭、空间宇称和时间反演）下的不同变换性质予以辅助。特别地，我们提出一种简洁分类方案，依据如下可观测量来区分此类系统中遍历行为的类型与层级：平均返回时间、守恒量数目以及关联函数的标度行为。在每一类物理上相异的类别中，我们给出具体示例并讨论其部分现象学特征，包括混沌或遍历动力学、相空间碎片化、Ruelle-Pollicott共振、拟局域守恒量的存在性，以及具有多种动力学指数的反常输运。

Classical cellular automata represent a class of explicit discrete spacetime lattice models in which complex large-scale phenomena emerge from simple deterministic rules. With the goal to uncover different physically distinct classes of ergodic behavior, we perform a systematic study of three-state cellular automata (with a stable `vacuum' state and `particles' with $\pm$ charges). The classification is aided by the automata's different transformation properties under discrete symmetries: charge conjugation, spatial parity and time reversal. In particular, we propose a simple classification that distinguishes between types and levels of ergodic behavior in such system as quantified by the following observables: the mean return time, the number of conserved quantities, and the scaling of correlation functions. In each of the physically distinct classes, we present examples and discuss some of their phenomenology. This includes chaotic or ergodic dynamics, phase-space fragmentation, Ruelle-Pollicott resonances, existence of quasilocal charges, and anomalous transport with a variety of dynamical exponents.