Research On Distributed Coordinated Predictive Control Of DC Micro-grid Group

As energy shortages and environment degrades,technology and society are also progressing and developing continuously,which makes micro-grids receive an increasing attention.It is flexible in control and simple in structure,which can better improve the utilization rate of renewable energy.With the increase of DC power equipment,the single form of DC micro-grid might not be able to meet the power demand of users.Therefore,the concept of DC micro-grid group was proposed and the distributed model predictive control method was used to control the system.The contributions as follows:Firstly,by establishing a mathematical model of the DC micro-grid,it could be known that the models of wind,photovoltaic and battery all had complex non-linear problem.Therefore,a neural network model for wind,photovoltaic,and battery subsystems were trained by using a neural network,and then linearized at a set point.Finally,the linearized distributed model of each subsystem was sorted out.Secondly,coordinated power generation for the entire DC micro-grid group in three modes.In mode 1,three sets of wind subsystems generated electricity,three sets of photovoltaic and battery subsystems were inactive.In mode 2,three groups of wind and photovoltaic subsystems generated electricity,three groups of battery subsystems were in an inactive state.In mode 3,three groups of wind,photovoltaic and battery subsystems generated electricity together.Simulation demonstrates the effectiveness of the method.Based on this,the distributed model predictive controller under three modes is designed and the quadratic programming method is used to find the optimal solution to meet the constraint problem.Finally,multiple single-port DC / DC converts were insteaded by a multi-port DC/DC converter,which was applied to the entire DC micro-grid group system.The result shows they cooperate well to maintain power balance in the system.Figure32;Table2;Reference 64...