The academic community has made outstanding achievements in researching the March algorithm. However, the current fault modeling method, which centers on fault primitives, cannot be directly applied to analyzing the March algorithm. This paper proposes a new test primitive. The test primitives, which decouple the cell states from sensitization and detection operations, describe the common features that must be possessed for the March algorithm to detect corresponding faults, forming a highly flexible and scalable March algorithm analysis unit. The theoretical analysis proves that the test primitives demonstrate completeness, uniqueness, and conciseness. On this foundation, the utilization of test primitives within the March analysis procedure is elucidated.

翻译：学术界在March算法的研究上取得了显著成就。然而，当前以故障基元为中心的故障建模方法无法直接应用于March算法的分析。本文提出了一种新的测试基元。测试基元通过将存储单元状态与敏化和检测操作解耦，描述了March算法检测相应故障所必须具备的共同特征，形成了一种高度灵活且可扩展的March算法分析单元。理论分析证明，测试基元具有完备性、唯一性和简洁性。在此基础上，阐明了测试基元在March分析流程中的应用。