Research On Optimal Operation Of Microgrids And Microgrid Clusters Based On Multi-agent Systems

With the development of renewable energy and distributed generation(DG)technologies,it has been a remarkable trend for the future energy system to reduce the impact of renewable energy connecting to the power system and to promote the utilization of DGs.Microgrids can integrate varieties of DGs and energy storage devices together.With advanced and flexible control strategies,microgrids can effectively reduce the DG access impact on power system,thus becoming the most efficient way for DG absorption.And a microgrid cluster can be formed by interconnecting multiple microgrids.Compared with single microgrid,the microgrid cluster can overcome the drawbacks like low inertia or weak anti-interference ability,and can improve the benefits and reliability of the system through coordination of multiple microgrids.The control and energy management is usually complicated due to the variety and quantity of distributed generations in the microgrid or microgrid cluster system.Therefore,an effective optimization energy management strategy is the key to achieve large-scale microgrid application and high operation efficiency.With features like autonomy,distribution and coordination,the multi-agent system(MAS)has strong consistency with the architecture of microgrid cluster systems.So MAS can effectively handle the operation scheduling and control problem of microgrid cluster systems with multiple distribution generations.In this paper,the optimal operation mechanism of microgrid and microgrid cluster systems based on MAS is studied.The main contents and contributions are as follows:1)The basic operation control framework of microgrid cluster systems is proposed based on the multi-agent system.And the mathematical models are built for all kinds of distributed generation devices in the system.2)The coordinated optimization strategies for combined cooling,heating and power microgrids are studied.Taking into account the uncertainty of renewable energy and loads,the scenario-based stochastic programming method and model predictive control is introduced.And a multi time scale microgrid optimal scheduling strategy is proposed to coordinate all kinds of distributed generators in the microgrid.3)The distributed optimization for islanded microgrid clusters is studied.an ADMM-based distributed optimization method for islanded microgrid clusters considering power flow constraints and transmission losses is proposed.The optimization problem is divided into two subproblems.In the upper level,the power exchange between different microgrids is coordinated with the objective of minimizing the operational cost and considering the power flow equations and transmission losses.The lower level considers the optimal operation of each microgrid.During the iteration,each microgrid communicates with its neighbors and solves the problem separately,which causes light communication and calculation burden.4)The interactive optimization operation for multi microgrids and active distribution network system is studied.The operating benefits and willingness of both microgrids and active distribution networks are considered.And with interaction between different entities,the bi-level optimization model for active distribution network with multi microgrids is formulated.Since each microgrid or the active distribution network has its own benefits and targets,the whole model is separated into several subproblems by copying the exchange power between the active distribution network and the microgrid or exchange power between adjacent microgrids.And the problem is solved using target cascading method to achieve the balance of interests of different entities.