A general control architecture of a real-time agent is presented for a distributed real-time multi-agent system from the point of view of design and implementation. Some mechanisms in the control architecture are studied， such as the perception of the environment， coordination and negotiation with messages， high-level perception based on wide-area information， integrated decision-making and scheduling. When the control architecture is applied to the design of a wide-area backup protection system， the working processes of some main modules in the protection agent are given. Some key implementing methods in designing the protection device are expatiated. The design scheme for the protection agent is intended to serve as a reference for future development of wide-area protection devices.