联盟形成是多Agent系统中一种重要的合作方式。人们设计了一系列联盟形成框架，较好地解决了联盟值最大化、任务分配、组合拍卖等问题。已有关于联盟形成的研究，较多地从效用、任务等角度来考虑问题。在一些情况下，仅从这些角度考虑联盟形成是不够的，于是我们从约束的角度来研究联盟形成。首先深刻分析了联盟形成时的约束问题，采用命题逻辑来描述对Agent的约束，给出了联盟偏好语言及其语义描述；接着给出了它的一些性质；最后将动态约束下的联盟形成机制与常见的一些联盟形成机制作了对比，体现了动态约束下联盟形成机制的特点。

Coalition formation is an important form of cooperation in multiagent systems. A series of frameworks for coalition formation have been designed,to some extent, which solves many important questions, such as the maximization of the sum of values of all coalitions, task allocation, combinatorial auctions, and so on.In their systems, the agents which form coalition are driven by interests or accomplishing tasks, which is helpless in supervising the participators in the system. Therefore, investigating the process of coalition formation from the aspect of restriction is necessary. We analyse the restriction problem in coalition formation, and then use proposition logic to describe the restriction of agents. Firstly, we design the coalition preference language(CPL), and describe its semantics. Secondly, we present some properties of CPL. Finally, we compare coalition formation under dynamic restrictions with some familiar frameworks of coalition formation, and educe the advantages of ours.