Planning Of Distribution Network Containing Distributed Generation And Electric Vehicles

Nowadays,as traditional energy is reducing and the voice for environment protection is growing,distributed generation(DG)and electric vehicles(EV)are widely used as effective means to alleviate energy crisis and environmental pollution.However,the output of DG such as wind turbines and photovoltaics has timing characteristics related to season and time of day,and the charging load of EVs which can be regard as the moving load has special spatialtemporal distribution characteristics.Therefore,the dual uncertainty of DG output and EV charging load brings serious challenges to the traditional distribution network planning.As a result,how to obtain a reasonable and effective distribution network planning scheme with DG and EV has become an urgent problem to be solved.The timing characteristics of DG and conventional load are studied,and timing characteristics model for DG and conventional load is proposed with season-scene and timedivision method in this paper.The Monte Carlo simulation method and Origin-Destination(OD)analysis method are adopted to establish the spatial-temporal model of EV charging load,and the impact of EV charging load on distribution network is evaluated.Chanceconstrained programming method is used to solve the problem of power flow uncertainty caused by the connection of DG and EVs with distribution network.In addition,the effects of DG and EVs on distribution network planning are also analyzed.The main work includes:(1)The wind speed,light intensity and different types of conventional load in the study area are simulated using Homer.Based on the simulation results,the timing characteristics of DG and conventional load related to season of year and time of day are analyzed,and limitations of traditional model such as constant power model,probability model and typical season model are derived.By using the method of season-scene and time-division,a timing characteristics model of DG and conventional load with both accuracy and reasonable calculation is developed,which provides model support for subsequent distribution network planning research.(2)EVs are classified according to the use of transport and vehicle type.Then,based on the geographic data,EV data and distribution network data of the study area,the spatialtemporal model of EV charging load is constructed with the Monte Carlo method to simulate the uncertain factors and OD analysis method to describe the travel demand,which provides model support for subsequent distribution network planning research.At last,the effects of EV charging load on load curve,network loss,voltage of nodes,power of branches with different EV permeability and different charging modes such as disorder mode,intelligent mode and mixed mode,are analyzed with a test system.(3)Based on the timing model of DG and conventional load,and spatial-temporal model of EV charging load,a chance-constrained model considering timing characteristics of distribution network which contains DG and EV is presented with the maximization of distribution network operator's profits as the objective function,node voltage,branch power and topology as constraints,replacement and construction of branches and allocation of DGs as optimization variables.In order to slove the proposed model,an improved Monte Carlo simulation-embedded double-population co-evolution genetic algorithm-based approach(DCGA-MCS)is adopted.(4)The advantages of DCGA-MCS in terms of searching ability and convergence speed are analyzed.The comprehensive distribution network planning scheme of a given test system is obtained through simulation,and feasibility and effectiveness of the proposed model and algorithm is verified with the simulation results.Moreover,the effects of timing characteristics,DG optimal allocation and EV charging load under intelligent modes on distribution network are discussed.