The growth of distributed generation or renewable generation resources and advanced information systems add uncertainties and complexity to the energy system and call for a “smart grid” solution. The mathematical models that are conventionally utilized in the energy management system usually take a long time to develop and execute and are not flexible enough to accommodate operational rules for generations (wind, solar and hydro) under varying conditions and loads. A rule-based system model would allow power system operators (who may lack experience in exploiting mathematical models and computer capabilities) to model their operations directly using their expertise, which enables the power system to make quick responses to real-time conditions. In this paper, a new rule-based system design methodology for power system automation and control is proposed. The methodology formulates the rule-based system as an optimal control problem in a feedback control architecture. The solution to the optimal control problem provides action rules for power system substation operations to achieve operator-configured preferences while minimizing cost. The proposed methodology, which integrates real-time information, optimal control, environmental constraints and human's expertise, provides a systematic tool for rule-based system design that gives more robust and realistic solutions.