Research On Multi-level Optimal Dispatch Method Of Renewable Energy Cluster

In recent years,China has achieved great leap forward growth in the development of renewable energy such as wind and light,and its new installed capacity has ranked first in the world for many years.With the increasing proportion of installed capacity,the uncertainty of renewable energy output and the trend of miniaturization and decentralization of power stations,the dispatching operation of power system has brought challenges that can not be ignored.Most of the current dispatching modes are based on the station level,which makes the dispatching center directly interact with wind farms and photovoltaic power stations,and make centralized optimization calculation in the dispatching center.However,this mode gradually shows the problems of insufficient communication risk resistance ability and poor system expansion flexibility,which is difficult to adapt to the future development pattern.At the same time,the traditional model used in the scheduling optimization is not reasonable enough to consider the uncertainty of renewable energy output,the scheduling method is extensive,and the output deviation and the problems of wind and light abandonment in the process of regulation and control still need to be improved.Facing the above problems,this paper systematically constructs a multi-level optimal scheduling method of renewable energy from the three-dimensional level of unit station cluster system.In the cluster,a two-level optimal scheduling model is constructed to adapt to the multi-level coordination of machine field group,and the prediction error of renewable energy output is considered in the cluster station scheduling.At the same time,the distributed optimization calculation is carried out based on the multi-agent system theory design algorithm,and the autonomous coordination and overall observable controllability of the cluster are realized.At the same time,the cluster is taken as the whole object to participate in the economic dispatch of power system,and the cluster division method considering the spatiotemporal smoothing effect is constructed.Based on t-copula function,the joint distribution model of the uncertainty of the output of the cluster is established,and then the penalty cost of the uncertainty of the cluster is introduced into the economic dispatch model,which effectively weakens the influence of the randomness and volatility of the renewable energy output on the economic dispatch of power system The purpose of improving the scheduling accuracy and economy is realized.Firstly,in view of the shortcomings of centralized scheduling mode,this paper proposes a multi-agent system theory based distributed bi-level optimal scheduling method for renewable energy.The upper optimization constructs a coordination model for cluster station level,and the lower optimization realizes dynamic control for station unit level.At the same time,a distributed algorithm is designed based on the consistency theory to solve the model.The effectiveness,robustness and flexibility of the proposed method in a variety of scenarios are verified by numerical examples.Then,aiming at the problem that the current cluster internal coordination model does not consider the uncertainty of renewable energy output,this paper constructs a probability density function based on t-location-scale distribution to represent the prediction error of station output,and then brings the prediction error distribution characteristics into the cluster optimal scheduling model for distributed calculation.In the analysis of the example,the weakening effect of the proposed method on the random characteristics of the cluster output is verified through the comparison of several groups,which is conducive to the realization of the overall controllable and friendly grid connection of the cluster.Finally,aiming at the deficiency of traditional power system economic dispatch methods,which can not make full use of the spatio-temporal smoothing characteristics under the background of high proportion of renewable energy,this paper analyzes the spatio-temporal smoothing effect of wind power and photovoltaic in machine field group multi-level,and then extracts the spatio-temporal characteristics of station output.Combined with the smoothing effect,a field group division method based on linear iterative clustering is formed.Then,t-copula function is used to construct the joint probability distribution of the output uncertainty of the field group,and the uncertainty penalty cost is introduced into the economic dispatch model of the power system,and the penalty coefficient is used to represent the dynamic demand of the system for reserve capacity.The results of example analysis show that compared with the traditional model,the proposed method improves the scheduling accuracy and economy,and has good application value.