As for the optimal economic requirements of energy coordination of micro-grid

receding horizon control（RHC） method under model predictive control framework is used traditionally and the support vector machines（SVMS） is sued to predict power of fan and load status. An improved complex process global optimization algorithm（IEACOP） with flexible frame structure is proposed in this paper with combination of feature of the micro-grid model. In this algorithm

each type of sub-method with search mechanism is embedded into the general framework. On the premise of giving full consideration to outer characteristic mathematical model and operational constraints of the wind generator and battery

the strategy of using infinite refraction mapping chaotic model to improve the diversity of the initial population generation is adopted. The local search method to deeply search the hyper-rectangle area caused by "transcendence method" is used to improve the solution efficiency of local optimal solution. It is shown by the simulation result that the method makes full use of advantage of the RCH and IE ACOP and meets the requirements of micro-grid energy coordination control. The IE ACOP method can use less adjustment parameters to finish the feasible solution of micro-grid coordinate optimization. The effectiveness of the proposed method is verified by the real example.