Distributed Price-based Optimal Control of Power Systems

This paper proposes a new control scheme for achieving optimal power balancing in electrical power systems. Optimality is defined via economical criteria related to steady-state operation requirements. Limits on the capacity of the power lines are taken into account via inequality constraints that are added to the optimization problem. Due to these inequality constraints, the resulting control law has a piecewise affine structure. Furthermore, the control action is distributed in the sense that a controller is assigned to each node and each controller communicates only with the controllers associated with the neighboring nodes. We prove that the resulting steady-state solution is optimal. The proposed control structure is a relaxation and an extension of classical "automatic generation control" in power systems, and can be interpreted as a fast acting real-time market for electrical power.