Research On Dynamic Economic Dispatch Of Power System Based On Multidisciplinary Collaborative Optimization Theory

The integration of large-scale wind power brings new challenges to the dispatch operation of power system because of its randomness,fluctuation and intermittence.In this paper,the Multidisciplinary Collaborative Optimization(MCO)theory is introduced to the dynamic economic dispatch(DED)problem of power system,and the DED problem of power system with stochastic wind power integrated is studied.A DED model based on MCO is established to minimize the sum of the coal cost and the energy cost because of the valve point effect of steam turbine in this paper.According to the different modeling methods of stochastic wind power,we study in the following two aspects:(1)Use the predicted output of wind power: that is,the predicted output of wind power is used as the actual output of wind power to find the optimal dispatching plan.Combined with the core idea of MCO theory,the constraints decoupling DED model based on MCO is established.The original complex,large-scale DED problem is decomposed into three relatively simple optimization problems,which are system-level optimization problem,system basic constraint optimization problem and characteristic constraint of conventional unit optimization problem.And the dynamic relaxation algorithm is introduced to solve the system-level optimization problem,which effectively overcomes the shortcomings of the traditional MCO algorithm.(2)Consider the randomness of the predicted error of wind power based on the scenario method: assuming the predicted error of wind power obeys the normal distribution,the Latin hypercube sampling is used to simulate the possible output of the wind power by sampling scenarios.The sampling scenarios will be cut down by using the fast forward reduction technology,and the finally error scenarios set is formed.Based on that,the scenarios decoupling DED model based on MCO is established.Each sub-problem corresponding to each error scenario is optimized in parallel by using the grid computation tool,which greatly improves the computational efficiency.The simulation examples based on IEEE 39-bus system verify the proposed model and method in this paper.