Spiking neural P systems with rules on synapses (RSSNP systems) are a class of computation models which are inspired by the information processing and communication manner of neurons. In this work, we consider labeled RSSNP systems (lRSSNP systems), where each rule is assigned either with a label chosen from an alphabet Σ or with the empty label λ. A string over an alphabet is accepted by an lRSSNP system if the string is processed from left to right by the system in the sense that in a step only rules labeled with the processed symbol or with λ are used (not both), and the system reaches a final configuration in the moment when the whole string is processed. The set of all accepted strings is categorized as the restricted (computations are done with the application of the rules labeled with symbols from the given alphabet) control language and the unrestricted (rules labeled with λ are allowed in the computations) control language of the system. We study the language accepting power of lRSSNP systems using only standard spiking rules by comparing the family of control languages of lRSSNP systems with the families of languages in the Chomsky hierarchy. It is proved that restricted lRSSNP systems can accept no more than context-sensitive languages, and unrestricted lRSSNP systems can accept recursively enumerable languages.

突触规则脉冲神经P系统（RSSNP系统）是一类受神经元信息处理和通信方式启发的计算模型。在这项工作中，我们考虑标记的 RSSNP 系统（lRSSNP 系统），其中每个规则都分配有从字母表 Σ 中选择的标签或空标签λ。如果一个字符串被系统从左到右处理，那么在一个步骤中只有用处理过的符号或λ标记的规则使用（不是同时使用），并且系统在处理整个字符串的那一刻达到最终配置。所有接受字符串的集合被归类为系统的受限（计算是通过应用给定字母表中的符号标记的规则来完成）控制语言和不受限制的（计算中允许使用λ标记的规则）控制语言. 我们通过比较 lRSSNP 系统的控制语言家族与乔姆斯基层次结构中的语言家族，研究了仅使用标准尖峰规则的 lRSSNP 系统的语言接受能力。证明受限制的lRSSNP系统只能接受上下文敏感的语言，不受限制的lRSSNP系统可以接受递归可枚举的语言。