| With the global energy and environmental issues becoming increasingly prominent,countries around the world have begun to attach importance to energy crisis,environmental damage and other issues,and focus on finding clean energy to supplement and improve the current situation.In recent years,the proportion of renewable energy in China has increased year by year,and the distributed generation projects have been put into production in the islands and remote areas.In addition,composed of distributed generation unit in microgrid system and network control system with smart microgrid,but also greatly reduce the operating costs of the micro grid system,improve the stability and flexibility of the microgrid,widen the application field.However,there are some deficiencies in the microgrid system based on network control,such as the independent and stable operation of the distributed unit,the energy switching of each distributed unit,the low voltage crossing of the microgrid.If these problems can not be effectively solved in the actual operation of the microgrid,the microgrid system will fall into unstable operation or paralysis,but also harm to the user side of the electrical equipment.Besides,the network is controlled by the detection system periodically,then the data will be transmitted to the central controller control decision.Due to the inherent defects of network communication,Such as communication packet loss,delay,etc.,it will inevitably occur communication errors,while the main microgrid distributed generation units because of the impact of meteorological factors,the fluctuation and randomness also makes the power output with uncertainty.Therefore,it needs to introduce new control methods in precision control.In view of this,this paper analyzes the structure and control strategy of many experts and scholars at home and abroad.It presents a neural network predictive control for energy optimization management of Microgrid.In this paper,the predictive control is added to the existing foundation,which provides the reference for the reasonable scheduling and control decision.It introduces a new microgrid energy management structure which makes the resource allocation more reasonable and energy management more optimized.Finally,the specific innovative research work from the following three aspects:(1)Predictive control model used the main prediction tool is the neural network,and join the reverse transmission path,which is BP neural network.In order to shorten the period of convergence,the genetic algorithm is introduced in this paper,but still can not completely eliminate the convergence cycle is too long.Therefore,the regression analysis is introduced to compensate for the long period of convergence.Neural network and regression analysis are combined to form a predictive control mechanism to realize the optimization of predictive control.(2)Each distributed unit is the use of power electronic interface output,and the output are U,I detection,the test information sent to the controller feedback to adjust the power electronic devices,to achieve a single distributed unit of relatively stable control output.In addition,the text will also send these detection signals to the central controller,and coordinated control of distributed units,making microgrid energy management more reasonable.(3)The microgrid energy management structure is multilevel structure,and is divided into upper level decision maker and lower level decision makers.the upper decision level of multilevel architecture system has a set of fixed parameter(reference values)function,and adapts to the lower decision level by a specific structure of the control strategy.On the basis of the parameters,each decision maker at the lower level solves its own optimization problem by tracking the reference values provided by the upper level.Structural modeling also consider taking into account the minimum carbon emissions and reduce production costs,to improve market competitiveness,making this research more engineering practice value. |
PDF Full Text Request