Distributed Energy Management in Smart Grid With Wind Power and Temporally Coupled Constraints

This paper investigates the problem of distributed energy management for both generation and demand sides in a smart grid by formulating the economic dispatch and demand response in a united framework. Our main contribution is to formulate a social welfare maximization problem for a more practical scenario by taking wind power, and temporally coupled constraints of the demands into account. The complexity lies in the nonquadratic cost function of wind power, the temporally coupled constraints of the demands, and the nonconvexity of the optimization problem. Meanwhile, a smart grid has to guarantee privacy and accommodates plug-and-play features. Aiming at these challenges, we first relax an equality constraint to obtain a new convex optimization problem. Because of the coupling in the constraint, the Lagrange duality method is then adopted to decompose the problem into subproblems for generators and demands, which are regarded as agents. As a result, each agent solves its subproblem by exchanging information with only neighbor agents, and coordinates with others using the global information discovered by a distributed finite-time consensus algorithm. We also prove the convergence and optimality of the proposed distributed energy management algorithm (DEMA). Finally, simulations are performed on the IEEE 39-bus system to illustrate the performance of our DEMA.