语言竞争与演化动力学本质上是语言在动态对话网络中的传承与传播，语言通过对话交流转化为语言使用者表示语言类型和掌握程度的内部结构改变，内部结构又通过该个体与其他个体交流实现语言的再次传播。运用数据和agent行为驱动的动态社会网络建模方法，构建朋友婚姻家庭动态社会网络和agent日常活动行为驱动的在家庭和公共场景交替发生的对话交流网络，刻画人员在不同场景下语言使用的社会异质结构。通过模糊逻辑和模糊认知图方法对个体语言使用学习建模，将语言使用频次、机会和语言声誉感知转化为语言使用者的语言内部结构、语言态度和地位认同度。建立考虑双语者等语言因素、社会因素、人口因素和环境因素及其相互作用机制的语言竞争动态agent网络社会仿真系统，结合云南少数民族语言使用现状和历史演变，研究语言竞争、濒危和政策调控演化动力学机制，形成对语言濒危与共存的成因、演化规律深刻认识，为制定语言保护政策提供预警和决策支持。

Language competition and evolution is essentially language inheritance and transformation on conversation dynamic network. Language conversation and communication can be changed into language internal structure that stands for language type and language usages capability, and language internal structure of one speaker can be transformed again by communicating with other speaker. The data and individual activities driven approach for modeling the dynamic social network will be used to developed friends and household dynamic social network and alternative conversation network between household and public environment driven by people daily activities. Language conversation inhomogeneous environment can be characterized with the developed dynamic networks. Individual words learning model is developed with a Fussy-Cognitive-Based representation model, and language usage frequency and perception of language prestige can be transferred into language internal structure and language status, attitude change and social identity. Language competition dynamic agent network model is developed by considering the interactions among bilingual factor, society structure factors, demographic factors and environment factors. Language competition, extinction and language maintenance mechanism will be investigated binding with ethnic minorities' language usage present situation and historic evolution in Yunnan. Profound understanding for mechanisms of language extinction, coexistence and evolution law will be obtained. Study results will provide decision making support for language maintenance strategies and early warning.