ABSTRACT This paper formulates a new control problem of achieving the global objective of a discrete event system with networked control agents that pursue only their own objectives and compete with each other. In this framework, each agent makes a decision to meet its own private objective with a given mask during state observation of the system and such a decision can be changed by the influence of other agents connected through a directed network. Agents with an initial decision of enabling or disabling of an event do not change their decision, but those agents reserving their decision in an undetermined way can change their decision to either enabling or disabling depending on the influence from other agents. We derive a condition for convergence of the decision of all agents. Based on the individual decision of each agent, the final decision for enabling or disabling of an event is made by majority rule. We present the network-controllability of a global objective with respect to a given mask set as a main condition with which the networked agents can meet the global objective while they pursue only their own private objectives.

中文翻译：

---

在离散事件系统的框架内与竞争的网络代理实现全球目标

摘要 本文提出了一个新的控制问题，即通过网络控制代理实现离散事件系统的全局目标，这些代理只追求自己的目标并相互竞争。在这个框架中，每个代理在系统状态观察期间使用给定的掩码做出决定以满足自己的私有目标，并且这种决定可以通过通过有向网络连接的其他代理的影响而改变。具有启用或禁用事件的初始决定的代理不会改变他们的决定，但那些以未确定方式保留其决定的代理可以根据来自其他代理的影响将其决定更改为启用或禁用。我们推导出所有代理的决策收敛的条件。根据每个代理的个人决定，启用或禁用事件的最终决定由多数规则做出。我们提出了关于给定掩码集的全局目标的网络可控性，作为主要条件，网络代理可以满足全局目标，而他们只追求自己的私人目标。