Multi-Objective Optimal Allocation Of DG-EV Charging Station Considering Space-Time Characteristics Model

With the promotion and development of new energy in the country, Distributed Generation (DG) and electric vehicle (EV) are welcoming development opportunities. The paper proposes a DG-EV charging station that considers space-time characteristics. Multi-objective coordination optimization configuration method, by extracting the road network topology in the city, monitoring the road network traffic, using the traffic planning software TransCAD to carry out the OD matrix back-pull, establishing the travel probability matrix to describe the user travel characteristics, considering the timing characteristics of EV, DG and conventional load, taking into account the grid side and the user side, and establishing a multi-objective joint planning model for the DG-EV charging station with the sub-goal of comprehensive benefit, system load fluctuation and charging time-consuming cost. Finally, the simulation analysis of the IEEE-33 node distribution network and the main road in a certain urban area is carried out to verify the validity and feasibility of the model.


Introduction
Nowadays, the research on DG and EV is mainly to plan the two separately, and less consider the coordinated planning of the two. Literature [1] designed a new type of regional electric vehicle charging station system including wind power, photovoltaic and energy storage systems, whichever is better the effect fully shows that the joint planning of DG-EV charging stations is beneficial to promote the complementary timing characteristics of intermittent DG output and charging load. Literature [2] considers the dispersion of electric vehicles, and uses the short-term interaction between electric vehicles and microgrids to stabilize Load fluctuations. It realizes the local consumption of DG by electric vehicles and improves the penetration rate of intermittent DG.

Construction of typical operating scenarios of "wind-light-load"
This paper selects the historical data of wind speed, light intensity and conventional load in each hour during a period of time in the planned area as the original samples, and generates the "wind-light-load" daily operation scenario in a unit of 24 hours to avoid substituting all data into the model. For the problem of excessive calculation, the original data is reduced by the K-Means algorithm [3] , and the typical daily operation scenario is extracted.one of that as shown in Fig. 1

EV charging demand forecast
Predicting the user's charging demand is a prerequisite for charging station planning. This chapter will combine electric vehicle parameters and user travel characteristics to simulate the user's travel process and predict the charging demand.

Electric vehicle type and battery parameters
Electric vehicles mainly include four types of buses, taxis, urban functional vehicles and private cars. The parameters of different types of electric vehicles are shown in Table 1. According to the battery capacity of different types of electric vehicles, the battery power at the time of travel is obtained. The remaining power Cap during driving can be obtained by equation (1).

TransCAD-based inverse model of OD matrix
TransCAD is a professional traffic planning software for traffic data management and analysis. It provides a single path allocation and multi-path allocation OD matrix inversion program. The OD inversion program provides All-or-Nothing and capacity limitation methods.and other traffic distribution models. Researchers can select the traffic distribution model based on the imported road network map information, combined with the actual situation, enter the prior OD matrix, and use the  Figure 2. Flowchart of OD matrix reverse push.

Temporal and Spatial Distribution of Electric Vehicle Charging Demand
The Monte Carlo method is used to simulate and generate the initial operating power Cap 0 and the initial travel time t c of private cars, taxis, and urban function vehicles. For vehicle k, according to the vehicle's initial position Oi and the initial travel time, call the travel probability and OD probability matrix corresponding to time t c , and use the stratified random sampling method to generate the destination D j corresponding to vehicle k, assuming that the driver will choose the shortest the shortest path search algorithm of Floyd is used to get the shortest path set to the destination D j . The vehicle passes through the roads in the set-in turn, and the remaining power Cap d of vehicle k is updated for each road section, corresponding to time t, if If the remaining power of the vehicle is lower than the charging threshold, the information of the electric vehicle that generates the charging demand is recorded, including the type of electric vehicle, the remaining power Cap d , the time t of the charging demand, and the location.

Planning model of DG-EV charging station considering time-space distribution
The Monte Carlo method is used to simulate and generate the initial operating power Cap 0 and the initial travel time t c of private cars, taxis, and urban function vehicles. For vehicle k, according to the vehicle's initial position Oi and the initial travel time, call the travel probability and OD probability matrix corresponding to time t c , and use the stratified random sampling method to generate the destination D j corresponding to vehicle k, assuming that the driver will choose the shortest the shortest path search algorithm of Floyd is used to get the shortest path set to the destination D j . The vehicle passes through the roads in the set-in turn, and the remaining power Cap d of vehicle k is updated for each road section, corresponding to time t, if If the remaining power of the vehicle is lower than the charging threshold, the information of the electric vehicle that generates the charging demand is recorded, including the type of electric vehicle, the remaining power Cap d , the time t of the charging demand, and the location.

Mathematical Model of Multi-Objective Optimal Allocation
When optimizing the configuration of DG-EV charging stations, the grid-connected location and installation capacity of DG and EV charging stations should be determined from both the power grid and the user. Based on this, the establishment is established with the largest comprehensive benefit,

Case analysis
This paper combines the IEEE-33 node system and 29-node road network transportation network to carry out the simulation analysis of the DG-EV charging station joint planning. the system parameters are shown in Reference [4]. 29-node traffic road network diagram 3, the coupling relationship between road network parameters and road network and distribution network see reference [5]. The travel probability matrix is used to simulate the travel trajectory of electric vehicles to obtain the time distribution of electric vehicle charging requirements as shown in Fig. 3. It can be seen from Figure 3 that the charging load has two peaks of charging demand at 13:00-14:00 and 17:00-18:00.  In order to verify the superiority of the joint planning of DG-EV charging stations, the number of charging stations is set to 5, and two schemes are established for simulation analysis. Option 1) First carry out DG independent planning, and carry out EV charging station planning on the basis of the DG planning scheme. Option 2) Joint planning of DG-EV charging stations. In the DG planning scheme, 13(6) indicates that 6 DGs are installed at node 13, and 20(690) in the charging station planning scheme indicates that the installed capacity of the charging station at node 20 of the road network is 690kW.Comparing scheme 1 and scheme 2, it can be found that compared with the independent planning of DG and EV, the planned capacity of DG and charging station is improved because the joint planning considers the mutual absorption capacity of EV charging load and DG.

Conclusions
This paper proposes a DG-EV charging station joint planning method that considers the timing and spatial characteristics of the charging load of electric vehicles. The effectiveness of the method is demonstrated by combining the IEEE-33 node and the main road in a certain urban area. The peak demand of charging load and the output of DG The power peaks are highly complementary, which can effectively achieve the effect of "peak cutting and valley filling" and increase the system's consumption of DG.