Research On Distributed Control And Optimization In Smart Grid

With the rapid development of communication technology and the improvement of processor computing performance,distributed control techniques of multi-agent systems have been significantly developed.Those techniques establish the foundation for dealing with power control and optimization problems in a distributed manner,such as load frequency,voltage regulation and economic dispatch problem.As a key topic,consensus means that all agents' states reach a common value with each agent only sharing information with its neighbors over communication networks.This is exactly consistent with the requirements in which the load frequency voltages in the whole power system reach consensus and the incremental costs reach consensus in traditional power control and economic dispatch problems,respectively.In this regard,two types of economic dispatch models with transmission interface limitation,generator limitation,spinning reserve constraints are studied in this paper.With regard to those problems,two classes of distributed optimization algorithms without depending on initial values based on the consensus mechanism are proposed.The main contents are as follows:In chapter 1,the research background and significance of both smart grid and multi-agent systems are introduced.Then,we expound related works on distributed control and optimization theory as well as economic dispatch problem in smart grids.In chapter 2,the main theoretical tools used in this paper are introduced,including algebraic graph theory,convex analysis,projection operator and variational inequality.In chapter 3,we study a class of static economic dispatch problems,where the transmission interface limitation is introduced in a static economic dispatch problem.Regarding such a problem,a distributed consensus algorithm based on cone projection is proposed.The convergence of the algorithm is proved by the properties of the cone projection and Lyapunov stability theory.Finally,a simulation example is given to illustrate the effectiveness of the algorithm.In chapter 4,a class of dynamic economic dispatch problems is studied.The dynamic economic dispatch problem is to calculate power output to minimize total cost over planed time periods.This chapter proposes a distributed projected algorithm on account of the consensus mechanism for solving the dynamic economic dispatch problem with complicated constraints.Then,it is proved that the optimal variable associated with generators converges to the optimal solution of the considered problem by using the tools from projection properties and Lyapunov stability.Finally,numerical simulations further validate the theoretical results.In chapter 5,we give a summary of the full thesis and suggest the future work.