Research On Dispatch Strategy Of Charge And Discharge Of Electric Vehicle Cluster Based On Trip Chain Simulation

The widespread application of distributed generation,energy storage systems,reactive power regulation devices,and other equipment has made the traditional distribution network evolve into an active distribution network with high controllability.Meanwhile,electric vehicles have shown great advantages in alleviating dependence on fossil energy.With the "Peak Carbon Dioxide Emissions" and "Carbon Neutrality" goals proposed and the "Vehicle-to-Grid" technology advancement,electric vehicles will usher in rapid development together with renewable energy.In this context,how to simulate the continuous travel scenarios of electric vehicle clusters based on actual travel data and analyze the charging and discharging capabilities of the clusters,and how to improve the economics of grid operation by combining the high controllability of the active distribution network under the premise of ensuring the charging interests of electric vehicle users is of important research value.This paper discusses as follows:A trip chain model is proposed to describe the continuous travel behavior of a single electric vehicle over a period of time,and it defines the characteristic variables of the trip chain such as the first departure time of the day,the transition probability of the travel purpose,the length of parking,the driving time,and the driving mileage.Probability distribution models of the characteristic variables are introduced,and the Gaussian mixture distribution model and its distribution parameter fitting method based on the maximum expectation algorithm are introduced emphatically.The selection method of charging and discharging behavior of electric vehicles based on the five-zone map model is introduced.The optimal power flow model of the distribution network is proposed.Through phase angle relaxation and second-order cone relaxation,the original AC optimal power flow model,which is strongly non-convex and difficult to converge,is transformed into a second-order cone-convex optimization model.The convergence and relaxation accuracy of the obtained relaxation convex optimization model is analyzed.In order to achieve the global optimization of the active distribution network in the dispatch period,the single-period static optimal power flow model is extended to a multi-period dynamic optimal power flow model by introducing the time dimension.This paper introduces the calculation process for evaluating the charging and discharging capacity of electric vehicle clusters.Based on the National Household Travel Survey dataset,the probability distribution parameters of the characteristic variables are fitted.The accuracy of the simulated trip chain is varified.The natural charging demand of electric vehicle clusters is analyzed,which shows that the natural charging demand curve shows obvious peak-to-valley differences.The charging demand of residential areas accounts for about 60% to 70% of the total charging demand,and the charging demand in the shopping area and entertainment area is the lowest.The cluster charging and discharging capacity of different proportions of electric vehicles participating in the Vehicle-to-Grid is analyzed,results show that the residential areas and work areas meet more than 90% of the charging demand scenarios,and contribute more than 93% of the discharged power.The influence of the cluster size on the charging and discharging capacity is analyzed,and the excellent parallel computing performance of the proposed model is verified.An optimization strategy based on second-order cone programming for the active distribution network to accept the charging and discharging of electric vehicle clusters is proposed.First,a two-stage optimization solution model architecture that takes into account the two main bodies of electric vehicle charging agents and distribution network dispatching centers is proposed.The first stage calculates the charging plan of each charging agent in the three scenarios of disorderly charging,orderly charging,and orderly charging-discharging.The second stage solves the multi-period optimal power flow model of the distribution network that takes into account the access of energy storage,photovoltaic,wind power,and reactive power regulation devices,and aims at minimizing active power loss.The case study based on the improved IEEE-33 model shows that the orderly charging and orderly charging-discharging strategies significantly reduce the charging cost of electric vehicle cluster users compared with the disordered charging scheme.The active distribution network can reduce the power loss of the network by controlling devices under the premise of ensuring the charging income of electric vehicle users.All calculation scenarios in the case study have passed the second-order cone relaxation accuracy verification,and the excellent parallel computing performance of the proposed model has been verified.