The Research On Building And Simulating To The Knowledge Learning Matrix Of Section Team For Large-scale Construction Project

Along with the increasingly complex construction project management, and the higher requirements to knowledge learning and learning ability of team, knowledge has become the basic means to improve the efficiency of the team. In construction of engineering project, section team knowledge learning faces some problems, such as the staff awareness, the learning knowledge mining strength and learning efficiency are not high enough. In this context, it is necessary to make a deeper research on knowledge learning of section team for construction project.Firstly, the meaning of section team of Large-scale construction project was defined, which was on the basis of the broad access to domestic and foreign related literature and research results. Learning principles of section team were expounded from four angles, namely tacit knowledge into principle, knowledge chain cohesion knowledge map, knowledge of fusion principle, principle of deviation. Meanwhile, the knowledge learning of section team was analyzed deeply. Secondly, through the analysis of the connotation, the basic process of Boston matrix, combining with the section team knowledge learning characteristics, the thesis constructed three different kinds of Learning Matrix, namely the Explicit/implicit-Complexity, Forgetting-Learning, Aging-Cycle; Moreover, this thesis elaborated the multi-agent reinforcement learning process, and expounds the related theory of markov games on the basis of thorough analysis to multi-agent learning theory, in order to build multi-agent learning algorithm of markov games in section team, make algorithm simulation and prove the convergence of the algorithm. In the meantime, according to the results of the experiment simulation analysis, the thesis put forward the suggestions and measures to knowledge reinforcement learning of Large-scale construction project combining with the knowledge learning of section team.This thesis introduced the multi-agent learning mechanism and markov games model, and made it applied to section team knowledge learning study of Large-scale construction project. The author built the knowledge learning matrix, and it was a new attempt in the study of construction project team learning, aimed to provide references for the section team knowledge learning study of Large-scale construction project.