Event-triggered Scheme For Finite-time Distributed Economic Dispatch In Smart Grids

With the energy crisis and environmental pollution becoming more and more serious,the development and use of renewable energy resources are essential to attain sustainable development in human society.In order to ensure the safe,reliable and economic operation in a power grid with renewable energy,scholars put forward a concept called the smart grid.The economic dispatch problem refers to find the most economical and safe power generation distribution to meet the power demand of system load under the premise of satisfying some physical constraints,which is of great importance to the operation planning of smart grids.Frequent and severe changes in operating conditions of the power grid require rapid convergence of economic dispatch algorithms.To realize this goal and reduce information exchange,this thesis studies finite-time distributed economic dispatch under event-triggered scheme in smart grids.First,aiming at the economic dispatch problem in smart grid,an optimization model is established to minimize the generation cost,which can be equivalently transformed into a consensus problem of the incremental cost of generation according to the KKT condition.To reduce information exchange on this basis,a novel distributed event-triggered finite-time algorithm is proposed based on the multi-agent system framework in this thesis and analyzes the convergence of the algorithm.The results show that the proposed algorithm can converge to the neighborhood of the optimal solution in finite time without Zeno behavior while always satisfying the supply-demand balance.Then,considering that the initial state of the generation is difficult to obtain in practical,the fixed-time economic dispatch problem is studied in this thesis.Based on fixed-time consensus,a distributed event-triggered fixed-time algorithm is proposed by choosing the incremental cost of generation as the consensus variable.Theoretical analysis shows that the algorithm can obtain an upper bound for the estimation of the settling time which is independent of the initial state,and the inter-event time between any two events has a positive lower bound.The algorithm is used in IEEE 14-bus systems and IEEE-162 bus systems respectively.Simulation results show that the distributed economic dispatch algorithm proposed in this paper can converge to the neighborhood of the optimal solution within 1 second.Finally,considering the fixed-time economic dispatch problem under any initial power allocation.Based on duality theory,the economic dispatch problem is equivalently transformed into a distributed unconstrained optimization problem.To solve this problem,a distributed event-triggered fixed-time dual gradient algorithm is proposed based on the dual gradient method.The algorithm can obtain the optimal power generation of each generating unit in a fixed time and does not need the initial power allocation to satisfy the supply-demand balance.Theoretical analysis shows that the designed event-triggered communication has a minimum time between events.The algorithm is used in IEEE-57 bus systems and IEEE-118 bus systems respectively.Simulation results verify the effectiveness and scalability of the proposed algorithm.