A reliability assessment method for distribution network containing distributed power sources and electric vehicles

Abstract

A reliability assessment method for a distribution network containing distributed power sources and electric vehicles belongs to the field of distribution network reliability assessment. It includes: considering the uncertainty of wind and solar power output, constructing a probability model of wind turbine and photovoltaic generator output; analyzing EV user behavior characteristics, establishing an EV charging load model, and on this basis proposing an EV orderly charging and discharging control strategy based on dynamic time-of-use electricity prices; using the Monte Carlo sampling method to calculate the reliability index of each load point; performing multiple simulations, and statistically analyzing and accumulating the reliability index of each load point to calculate the reliability index of the entire system; finally, calculating the reliability index after connecting to the distribution network from the number of electric vehicles and different charging and discharging methods, and using the F4 main feeder of the improved IEEE RBTS BUS6 test system, which can better evaluate the reliability of the distribution network containing distributed power sources and electric vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of reliability assessment of distribution networks, and in particular to a reliability assessment method for distribution networks containing distributed power sources and electric vehicles. Background Art

[0002] With the proposal of the "dual carbon" goal, distributed generation (DG) based on wind and photovoltaic power generation has been vigorously developed. As of the end of October 2021, the installed capacity of wind and photovoltaic power generation reached 299 million kilowatts and 282 million kilowatts respectively. The National New Energy Absorption Monitoring and Early Warning Center released data on the grid-connected absorption of wind power and photovoltaic power in my country from January to July 2021, showing that the utilization rates of wind power and photovoltaic power reached 96.7% and 98% respectively, and the overall utilization rate was at a high level. Electric vehicles (EVs), as a green means of transportation with broad development prospects, are also an important way to achieve the "dual carbon" goals and have been vigorously developed. As of July 2021, the number of electric vehicles in my country reached 6.3 million, and a total of 66,000 charging stations had been built. However, the output of wind power and photovoltaic units is affected by natural conditions such as light intensity, temperature, and wind speed, and is highly random, intermittent, and volatile. Large-scale DG and EV access to the distribution network will inevitably affect the reliability of the distribution network. Therefore, it is necessary to evaluate the reliability of the distribution network containing DG and EV.

[0003] At present, the research on DG access to distribution network mainly focuses on the establishment of DG output model, optimal configuration, site selection and capacity determination. The literature "Xiong Xiaoping, Yang Lu, Li Ning, et al. Reliability analysis of distribution network with distributed generation based on affine minimum path method [J]. Automation of Electric Power Systems, 2017, 41(17): 43-50." established a multi-capacity state model of generators based on Markov process to simulate output fluctuations, and used the affine minimum path method to calculate the reliability of the distribution network with wind and solar power. The literature "Shi Min, Zhou Buxiang, Zhang Bing, et al. Reliability assessment of distribution network system considering the uncertainty of distributed generation [J]. Journal of Electric Power Systems and Automation, 2017, 29(12): 114-120." used the Cornish-Fisher series expansion method to generate a sample space that conforms to the standard normal distribution, and further analyzed the influence of the light intensity correlation coefficient on the reliability of the distribution network system. The literature "Zhaohong Bie, Zhang Peng, Li Gengfeng, et al. Reliability Evaluation of Active Distribution Systems Including Microgrids [J]. IEEE TRANSACTIONS ON POWER SYSTEMS, 2012, 27 (4): 2342-2350." introduces a multi-state model to characterize the output of distributed power sources, and uses a non-sequential Monte Carlo simulation method to evaluate the reliability of distribution networks. The literature "Li Juan, Zhou Honglian, Zhou Erbiao, et al. Reliability evaluation of distribution networks considering the correlation between wind speed and electric heating loads such as electric boilers [J]. Electric Power Automation Equipment, 2018, 38 (10): 26-31." considers the correlation between wind speed and electric heating loads, constructs the probability distribution function of the joint output based on the Copula theory, and uses Monte Carlo simulation for reliability calculation. These studies only consider distributed power sources in reliability evaluation, and do not calculate and evaluate the reliability of distributed power sources and electric vehicles when they are connected at the same time.

[0004] At present, there are also related studies on the reliability assessment of electric vehicles connected to the distribution network. The literature "Huang Junhui, Zhou Hao, Han Jun, et al. Reliability Assessment of Distribution Network Considering Vehicle-Grid Interaction [J]. Electric Power Construction, 2017, 38 (02): 77-83." established a spatiotemporal distribution model of electric vehicle charging and discharging, and used the Monte Carlo simulation method to conduct reliability assessment from three aspects: electric vehicle penetration rate, battery type, and battery capacity. The literature "ZHANG Q, ZHU Y, WANG Z, et al. Reliability Assessment of Distribution Network and Electric Vehicle Considering Quasi-Dynamic Traffic Flow and Vehicle-to-Grid [J]. IEEE ACCESS, 2019, 7: 131201-131213." proposed an efficient quasi-dynamic driving simulation method considering the impact of traffic congestion on path selection, and used the improved Monte Carlo simulation method to conduct reliability assessment on three aspects: electric vehicle penetration rate, discharge threshold, and battery capacity. The literature "CHENG S, WEI Z, SHANG D, et al. Charging Load Prediction and Distribution Network Reliability Evaluation Considering Electric Vehicles' Spatial-Temporal Transfer Randomness[J]. IEEE ACCESS, 2020, 8: 124084-124096." studied the spatiotemporal distribution prediction of electric vehicle charging load, and then used the trip chain technology, Monte Carlo method and Markov decision process (MDP) theory to evaluate the reliability of the distribution network with large-scale cluster electric vehicle penetration. The literature "GE Shaoyun, LI Jifeng, LI Teng, et al. Comprehensive reliability analysis of the distribution network and the urban road network association network[J]. Proceedings of the CSEE, 2016, 36(06): 1568-1577." used game theory to analyze the urban road network and distribution network, and calculated and analyzed the comprehensive reliability indicators of the urban road network and distribution network. These studies only evaluate the reliability of electric vehicles connected to the distribution network, and do not calculate the reliability of the distribution network containing both DG and EV. Especially with the in-depth research on the optimal scheduling of orderly charging and discharging of electric vehicles, choosing an orderly charging and discharging method that has the least impact on the reliability of the distribution network is of great significance to the popularization of electric vehicles.

[0005] The present invention comprehensively considers the impact of distributed power sources and electric vehicles connected to the distribution network, and the obtained reliability evaluation index is more in line with the actual situation, which can provide a basis for the future distribution network planning containing large-scale distributed power sources and electric vehicles. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a method for evaluating the reliability of a distribution network including distributed power sources and electric vehicles, and to model and evaluate the reliability of a distribution network including distributed power sources and electric vehicles.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0008] A reliability assessment method for a distribution network containing distributed power sources and electric vehicles, comprising the following steps: Step 1, considering the uncertainty of wind and solar power output, constructing a probability model of wind turbine and photovoltaic generator output; Step 2, analyzing EV user behavior characteristics, establishing an EV charging load model, and on this basis proposing an EV orderly charging and discharging control strategy based on dynamic time-of-use electricity prices;

[0009] Step 3, using the Monte Carlo sampling method to calculate the reliability index of each load point in the distribution network; Step 4, performing multiple simulations, and statistically summarizing and accumulating the reliability index of each load point in the distribution network to calculate the reliability index of the entire system;

[0010] Step 5: Using the optimized sequential Monte Carlo simulation method, the reliability of the distribution network containing distributed power sources and electric vehicles is calculated after connecting different numbers of electric vehicles and adopting different orderly charging and discharging control strategies.

[0011] The probability distribution function F(P w ) and the probability distribution function F(P v ) can be expressed as follows:

[0012] The probability distribution of wind turbine output under:

[0013]

[0014] Where: v ci is the cut-in wind speed; v co is the cut-out wind speed; v r is the rated wind speed; A, B, and C can be obtained through calculation, and their values ​​are related to the parameter configuration of the fan;

[0015] The probability density function of photovoltaic generator output:

[0016]

[0017] Where: P v is the output power of the photovoltaic generator; P vmax is the maximum output power of the photovoltaic generator;

[0018] Probability distribution function of photovoltaic generator output:

[0019]

[0020] In the above Step 2, the specific steps of the EV orderly charging and discharging control strategy based on dynamic time-of-use electricity prices are as follows:

[0021] Step 2.1, analyze the behavior characteristics of electric vehicle users and build a probability model of the behavior characteristics of electric vehicle users;

[0022] Step 2.2, construct the objective function with the goal of optimizing the economic efficiency of electric vehicle charging stations and minimizing the impact on the distribution network;

[0023] Step 2.3, consider the charging and discharging power of electric vehicles, power balance of the power grid, charging and discharging status, upper and lower limits of electricity prices, and construct constraint conditions;

[0024] Step 2.4, use particle swarm optimization algorithm to find the optimal solution.

[0025] The specific steps of Step 3 above are:

[0026] Step 3.1, set the time series simulation period N of the distribution network, input various parameters of the distribution network, and initialize the data and simulation time T = 0;

[0027] Step 3.2, number each component of the distribution network, sample its status, and calculate the normal working duration of each component, i.e., time to fault, TTF, and fault repair time, i.e., time to repair, TTR;

[0028]

[0029] In the formula, λ i , μ i are the failure rate and repair rate corresponding to component i respectively; σ1 and σ2 are random numbers uniformly distributed between (0,1), so the normal working duration and fault repair time of the component obey exponential distribution;

[0030] Step 3.3, through Monte Carlo sampling, get the sequence of the operating status of each component of the system, find the minimum value in the TTF of each component, record the corresponding component number as i, the component is the faulty component, and accumulate the fault time T = T + TTF min ;

[0031] Step 3.4, find the location of the faulty component i, determine the load range affected by it, and calculate its fault repair time TTR i , and classify the load;

[0032] The unaffected load is Class I load, the load that can be restored to normal working state after the faulty component is replaced or repaired is Class II load, and the load in the isolated island is Class III load;

[0033] Step 3.5: For Class III load, perform the following processing: set a random number μ that follows a uniform distribution between (0, 1) m , judge μ m The probability of successful switching to an isolated island is P m If μ m >P m , then the island switching is considered successful, otherwise it is considered a failure and all loads in the island are powered off;

[0034] Step 3.6: Use the probability model of wind turbine and photovoltaic generator output established in Step 1 to sample the power P of DG. w , P v Then, using the EV charging load model established in Step 2, generate 8760 hours of charging load P according to the time series. cha , and superimposed with the original load in the island, the total load P L ,Determine whether the fault time is within the charging and discharging time of the electric vehicle. If it is not within this time period, the power outage situation of each load point in the island is counted, otherwise proceed to the next step;

[0035] Step 3.7. Set the output P of DG within the island range. w , P v With total load P L If there is a difference, the load is cut off and the corresponding fault time and fault number are accumulated. Otherwise, proceed to the next step.

[0036] Step 3.8, determine whether the simulation lifespan has been reached, if so, end the simulation process and proceed to the next step, otherwise return to Step 3.2;

[0037] Step 3.9, calculate the system reliability indicators: average number of system power outages SAIFI, average power supply availability ASAI, average system power outage time SAIDI, and average system power shortage AENS.

[0038] In the above Step 3.4, after the loads are classified, the number of failures and the failure time of each load are counted. For Class I loads, since they are not affected by the faulty components, the total number of power outages and the power outage time remain unchanged. For Class II loads, the outage time is added to the total power outage time, and the total number of power outages is increased by 1.

[0039] In the above-mentioned Step 2 and Step 5, the reliability of the distribution network is calculated and analyzed from the charging and discharging modes of electric vehicles using disordered charging and discharging, time-of-use electricity prices, and dynamic time-of-use electricity prices, respectively.

[0040] The invention provides a reliability assessment method for a distribution network containing distributed power sources and electric vehicles. The invention has the following beneficial effects: considering the uncertainty of wind and solar power output, a probability density function of wind power generation and photovoltaic power generation is established; analyzing the behavior characteristics of electric vehicle users during charging, and constructing a probability model of the behavior characteristics of electric vehicle users; on the basis of time-of-use electricity prices, a more effective orderly charging and discharging control strategy for electric vehicles is proposed, and the reliability indicators after the electric vehicles are connected to the distribution network are compared, and then the orderly charging and discharging mode with the least impact on the reliability of the distribution network is selected, which is helpful for the future large-scale electric vehicles to be incorporated into the distribution network; by conducting reliability assessment on the distribution network containing distributed power sources and electric vehicles, a theoretical basis can be provided for the connection of large-scale new energy with the distribution network and the planning of the distribution network in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0042] Figure 1 A schematic flow diagram of the method of the present invention;

[0043] Figure 2 It is the probability density map of daily mileage of electric vehicles;

[0044] Figure 3 The probability density map of the time when electric vehicles start charging;

[0045] Figure 4 It is the probability density map of the time when electric vehicles leave home;

[0046] Figure 5 This is the flow chart of particle swarm optimization algorithm;

[0047] Figure 6 It is the flow chart of reliability assessment of distribution network containing DG and EV;

[0048] Figure 7 To improve the F4 feeder diagram of the IEEE-RBTS Bus6 test system;

[0049] Figure 8 It is the daily load curve of different numbers of electric vehicles;

[0050] Fig. 9 This is the EV daily load curve when using the time-of-use electricity price charging method;

[0051] Fig.10 This is the EV daily load curve when using the dynamic time-of-use electricity price charging method. DETAILED DESCRIPTION

[0052] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0053] like Figure 1 As shown in , a reliability assessment method for a distribution network containing distributed power sources and electric vehicles is provided, and the steps are as follows: Step 1, considering the uncertainty of wind and solar power output, a probability model of the output of wind turbines and photovoltaic generators is constructed;

[0054] Wind speed and sunlight are highly random, so the output of wind turbines and photovoltaic generators is also highly random;

[0055] The power of a wind turbine is related to the wind speed, so the mathematical expression of its power is P w Respectively expressed as:

[0056]

[0057] Where: v ci is the cut-in wind speed; v co is the cut-out wind speed; v r is the rated wind speed; P r is the rated output power of the wind turbine; A, B, and C can be obtained by calculation, and their values ​​are related to the parameter configuration of the wind turbine;

[0058]

[0059] The output of a photovoltaic generator is determined by the light intensity on its surface, so the mathematical expression of its power is P v for:

[0060] P v =A×η×r (3)

[0061] Where: A is the total area of ​​the photovoltaic generator panel; η is the photoelectric conversion efficiency; r is the actual light intensity;

[0062] Although wind speed and light intensity are highly random, their probability distributions have certain regularity. Wind speed follows Weibull distribution, and light intensity follows Beta distribution. Their probability density functions are:

[0063]

[0064] Where: k and c are shape parameter and scale parameter respectively;

[0065]

[0066] Where: r m is the maximum light intensity; α and β are shape parameters;

[0067] From equations (1) and (4), we can get the probability distribution of wind turbine output under different wind speeds F(P w ):

[0068]

[0069] The probability density function of the photovoltaic generator output can be obtained from equations (3) and (5):

[0070]

[0071] Where: P v is the output power of the photovoltaic generator; P vmax is the maximum output power of the photovoltaic generator;

[0072] Therefore, the probability distribution function F(P v ):

[0073]

[0074] Step 2: Analyze the EV user behavior characteristics, establish the EV charging load model, and on this basis propose an EV orderly charging and discharging control strategy based on dynamic time-of-use electricity prices;

[0075] After large-scale EVs are connected to the distribution network, the charging load is superimposed on the peak power load, which will lead to load tension in some local areas. The factors affecting EV charging load include objective factors such as EV type, quantity and charging method. In addition, the more critical factor is the user's behavioral characteristics, which are highly random, including the EV's daily mileage, the time to start charging, and the time away from home.

[0076] The state of charge (SOC) and charging power of an EV depend largely on its battery type and charging method. Currently, EV batteries mainly use lithium batteries, nickel-metal hydride batteries, and lead-acid batteries. Lithium batteries are widely used in EVs because of their high safety performance, long service life, large capacity, and high temperature resistance. Therefore, this article uses a Nissan Altra equipped with a 29.07kWh lithium battery for analysis.

[0077] Step 2.1, analyze the behavior characteristics of electric vehicle users and build a probability model of the behavior characteristics of electric vehicle users;

[0078] By analyzing the data from the U.S. Department of Transportation's survey of household vehicles across the United States through maximum likelihood estimation, and then fitting the historical data, it is found that the daily mileage approximates the parameters (μ, σ 2 ) is a log-normal distribution, and its probability distribution function is shown in formula (9), and the probability distribution diagram is shown in Figure 2 As shown;

[0079]

[0080] Where: x is the daily mileage of a single electric vehicle,, unit: km, μ x is the expected value, take 3.20, σ x is the standard deviation, which is 0.88, and the coefficient of determination R square is 0.9975;

[0081] The SOC can be obtained from the daily mileage:

[0082]

[0083] Where: x is the daily mileage, unit is km; E in SOC when charging an electric vehicle; E out W is the SOC of the electric vehicle when it was fully charged and left the grid the previous day; 100 Indicates the power consumption of EV per 100 kilometers; B is the battery capacity of EV;

[0084] The time when the user's vehicle returns, that is, the time when the electric vehicle starts charging, satisfies the normal distribution, and its probability density function is shown in formula (11). The probability distribution is as follows: Figure 3 Shown; 0

[0085]

[0086] Where: t ch The moment when electric vehicles start charging, μ ch is the expectation, take 17.47, σ ch is the standard deviation, which is 3.4, and the coefficient of determination R square is 0.9899;

[0087] The time when the electric car leaves the home also satisfies the normal distribution. Its probability density function is shown in formula (12). The probability distribution diagram is shown in Figure 4 As shown;

[0088]

[0089] Where: t d is the time when the electric vehicle leaves home for the first time, μ dis the expected value, take 7.70, σ d is the standard deviation, which is 3.27, and the coefficient of determination R square is 0.9987;

[0090] Step 2.2, construct the objective function with the goal of optimizing the economic efficiency of electric vehicle charging stations and minimizing the impact on the distribution network;

[0091] Through the analysis of the probability model of the time when EV starts charging, large-scale EVs will charge at the same time during the peak period of electricity consumption, such as from 6 to 9 pm, which will cause the load to "peak on peak" and affect the reliability of the distribution network. Therefore, from the perspective of the reliability of the distribution network, it is necessary to control or guide the charging behavior of EVs. Under the premise of not affecting the travel of users, EVs can be guided to charge and discharge in an orderly manner through effective control strategies to avoid the peak period of load electricity consumption. At present, the means of guiding EVs to charge and discharge in an orderly manner is mainly in the form of electricity prices. Common electricity price forms are divided into three categories: fixed electricity prices, time-of-use electricity prices, and real-time electricity prices. With the in-depth study of the optimal scheduling of EV orderly charging and discharging, dynamic time-of-use electricity prices have gradually become a control method for EV orderly charging and discharging. This paper takes charging stations as the background, considers reliability and economy, and constructs an EV charging and discharging control strategy;

[0092] For charging stations, the goal is to maximize their revenue and minimize the impact on the distribution network. The mathematical expression is:

[0093]

[0094] Where: are the charging and discharging power of EV respectively; are the electricity selling and purchasing prices of the charging station, λ is the adjustment factor, is the interaction power between the charging station and the distribution network;

[0095] Step 2.3, consider the constraints such as electric vehicle charging and discharging power, power grid power balance, charging and discharging status, upper and lower limits of electricity prices, and construct constraint conditions;

[0096] Step 2.3.1. Charge and discharge power constraints

[0097]

[0098] Where: are the charging and discharging power of the mth electric vehicle at the same time, is the maximum charging power of electric vehicles. It is generally believed that the maximum charging and discharging powers of electric vehicles are equal;

[0099] Step 2.3.1. Power Constraint

[0100] The maximum load that the distribution network can bear To meet the basic load of the network EV charging and discharging power The sum is:

[0101]

[0102] Step 2.3.3, SOC constraint when off-grid

[0103] In order to meet user needs, electric vehicles must meet the following requirements when they are off-grid:

[0104]

[0105] Where: E o 2 ut It is the SOC of the electric vehicle when it is off-grid;

[0106] Step 2.3.4. Electric vehicle charging and discharging state constraints

[0107]

[0108] Where: t arr ,t dep The time when EV is connected to the grid;

[0109] Step 2.3.5: Dynamic electricity price upper and lower limit constraints

[0110] Taking into account the operating costs of charging stations and the consumption level of users, the upper and lower limits of the dynamic electricity price for charging and discharging are:

[0111]

[0112] Where: C min , C max They are the upper and lower limits of charging and discharging electricity prices respectively;

[0113] Step 2.4, use particle swarm optimization algorithm to find the optimal solution;

[0114] To solve the multi-variable and high-dimensional optimization problem in this model, particle swarm optimization (PSO) algorithm can be used. During the iteration, particle i tracks its individual optimal solution. and the global optimal solution To update your new position and speed:

[0115]

[0116] Where: k represents the current number of iterations, Respectively represent the speed and position of particle i in the d-dimensional process during the optimization process, ω is the inertia weight; C1 and C2 are learning factors; r1 and r2 are random numbers uniformly distributed between [0,1];

[0117] This paper takes 1 hour as the time scale of electricity price. The specific algorithm process is as follows Figure 5 As shown;

[0118] Step 3: Use the Monte Carlo sampling method to calculate the reliability index of each load point. The flow chart is as follows: Figure 6 As shown;

[0119] Step 3.1, set the time series simulation period N of the distribution network, input various parameters of the distribution network, and initialize the data and simulation time T = 0;

[0120] Step 3.2, number each component of the distribution network, sample its status, and calculate the normal working duration of each component, i.e., time to fault, TTF, and fault repair time, i.e., time to repair, TTR;

[0121]

[0122] In the formula, λ i , μ i are the failure rate and repair rate corresponding to component i respectively; σ1 and σ2 are random numbers uniformly distributed between (0,1), so the normal working duration and fault repair time of the component obey exponential distribution;

[0123] Step 3.3, through Monte Carlo sampling, get the sequence of the operating status of each component of the system, find the minimum value in the TTF of each component, record the corresponding component number as i, the component is the faulty component, and accumulate the fault time T = T + TTF min ;

[0124] Step 3.4, find the location of the faulty component i, determine the load range affected by it, and calculate its fault repair time TTR i , and then divide the island according to the load range affected by the faulty component, and classify the load: the unaffected load is Class I load, the load that can still resume normal working state after the faulty component is replaced or repaired is Class II load, and the load in the island is Class III load;

[0125] Step 3.5. Count the number of failures and the failure time of each load. For Class I load, since it is not affected by the faulty component, its total number of power outages and power outage time remain unchanged. For Class II load, add the outage time to the total power outage time, and the total number of power outages increases by 1. For Class III load, perform the following processing: set a random number μ that follows a uniform distribution between (0,1) m , judge μ m The probability of successful switching to an isolated island is P m If μ m >P m , then the island switching is considered successful, otherwise it is considered a failure and all loads in the island are powered off;

[0126] Step 3.6: Use the probability model of wind turbine and photovoltaic generator output established in Step 1 to sample the power P of DG. w , P v Then, using the EV charging load model established in Step 2, generate 8760 hours of charging load P according to the time series. cha , and superimposed with the original load in the island, the total load P L ,Determine whether the fault time is within the charging and discharging time of the electric vehicle. If it is not within this time period, the power outage situation of each load point in the island is counted, otherwise proceed to the next step;

[0127] Step 3.7. Set the output P of DG within the island range. w , P v With total load P L If there is a difference, the load is cut off and the corresponding fault time and fault number are accumulated. Otherwise, proceed to the next step.

[0128] Step 3.8, determine whether the simulation lifespan has been reached, if so, end the simulation process and proceed to the next step, otherwise return to Step 3.2;

[0129] Step 3.9, calculate the system reliability indicators: average number of system power outages SAIFI, average power supply availability ASAI, average system power outage time SAIDI, and average system power shortage AENS.

[0130] Example:

[0131] In order to verify the effectiveness of the reliability assessment method for distribution networks containing distributed generation and electric vehicles proposed in this paper, the example in this paper adopts the F4 feeder under the improved IEEE-RBTS Bus6 test system. The system structure diagram is shown in Figure 7As shown in the figure, wind power and photovoltaic power generation systems are connected at node 1, and different numbers of EVs are connected at node 2. The reliability parameters of the components in the system are shown in the following table:

[0132] Table 1 Reliability parameters of components

[0133]

[0134] Solution 1: Connect a wind turbine and a photovoltaic generator with a capacity of 3.5MW at node 1, and connect 200, 400, 800, and 1,000 electric vehicles at node 2 respectively, and use disordered charging and discharging. Figure 8 As shown in the figure, it is a daily load curve diagram of different numbers of EVs using disordered charging. It can be seen from the figure that as the scale of EVs continues to expand, the load of the system also increases. As the number of EVs increases, the impact on the reliability of the distribution network needs to be evaluated through reliability index calculation. Table 2 shows the reliability index after different numbers of EVs are connected to the distribution network.

[0135] Table 2 Reliability indicators when different numbers of electric vehicles are connected

[0136]

[0137] As shown in Table 2, after the electric vehicles are connected, SAIFI and SAIDI increase from the original 2.7745 times and 14.0786h to 2.9317 times and 14.9640h respectively, and the changes are quite obvious, indicating that the reliability of the distribution network is reduced after the electric vehicles are connected to the distribution network; with the increase in the number of electric vehicles connected, compared with 200 vehicles, the reliability indicators SAIFI and SAIDI increase slightly, and ASAI also decreases from 99.84% to 99.83%, because these indicators are related to the failure rate of components; but the system's power shortage indicator changes significantly, from the original 75.0307MW to 78.3246MW, and EENS changes significantly with the increase in the number, because the EV charging is superimposed with the load peak time, which increases the peak-to-valley difference of the load and causes the system power outage load to increase. Therefore, after the EV is connected to the distribution network, the reliability of the distribution network is reduced, and the reliability index becomes worse with the increase in the number of electric vehicles.

[0138] Solution 2: A wind-solar combined power generation system with a capacity of 7MW is connected at node 1, and 200 electric vehicles are connected at node 2. The charging and discharging methods of time-of-use electricity price and the dynamic electricity price proposed in this paper are adopted respectively. In formula (13), λ=0.8. As shown in Table 2, it is the time-of-use electricity price parameter, which is the electricity selling and purchasing price adopted by the EV charging station.

[0139] Table 3 Time-of-use electricity price parameters

[0140]

[0141] like Fig. 9 , 10 The figure shows the daily load curve of EV using different charging methods. Although compared with disordered charging, these two orderly charging methods have the effect of "peak shaving and valley filling", the planners and builders of the distribution network are more concerned about how to better reduce the impact on the reliability of the distribution network. By comparing the reliability indicators of EV using different orderly charging and discharging control strategies, a reference basis can be provided for the planning of new distribution networks containing EVs.

[0142] Table 4 Reliability indicators for different orderly charge and discharge control strategies

[0143]

[0144] The reliability calculation results are shown in Table 4. For the convenience of comparison, the reliability calculation results of the disorderly charging of 200 EVs in the above example are placed in Table 4 for comparison with the orderly charging and discharging method with dynamic electricity price adopted in this paper. It can be seen from the table that compared with disorderly charging and discharging, SAIFI and SAIDI are both smaller after adopting the charging and discharging methods of time-of-use electricity price and dynamic electricity price, and the power shortage index of the system is also reduced from 77.6662MW to 76.3983MW and 75.7556MW respectively. This is because compared with disorderly charging and discharging, the orderly charging and discharging method can reduce the power shortage index of electric vehicles. Managing the charging behavior of the electric vehicle can effectively smooth out load fluctuations, reduce the peak-to-valley difference of the distribution network, and thus reduce the impact on the reliability of the distribution network; due to accuracy issues, the ASAI of the three charging methods is the same at 99.84%. In fact, in the simulation calculation results, the latter two are slightly larger than the ASAI of disordered charging; therefore, although after comparison, the orderly charging and discharging methods using time-of-use electricity prices and the dynamic electricity prices of this article have the effect of "peak shaving and valley filling", and both can effectively reduce the impact on the distribution network, the dynamic electricity price method of this article has less impact on the reliability of the distribution network than the time-of-use electricity price.

Claims

1. A reliability assessment method for a distribution network containing distributed power sources and electric vehicles, characterized in that: The steps are: Step 1, construct the probability model of wind turbine and photovoltaic generator output; Step 2: Analyze the EV user behavior characteristics, establish the EV charging load model, and on this basis propose an EV orderly charging and discharging control strategy based on dynamic time-of-use electricity prices; Step 2.1, analyze the behavior characteristics of electric vehicle users and build a probability model of the behavior characteristics of electric vehicle users; The daily mileage is subject to the parameter The log-normal distribution of has a probability distribution function as shown below: ； Where: is the daily mileage of a single electric vehicle, unit: km, is the expected value, take 3.20, is the standard deviation, which is 0.88, and the coefficient of determination R square is 0.9975; The SOC can be obtained from the daily mileage: ； Where: is the daily mileage, in km; SOC when charging an electric vehicle; The SOC of the electric vehicle when it was charged and left the grid the previous day; Indicates the power consumption of EV per 100 kilometers; is the EV battery capacity; The time when the user's vehicle returns, that is, the time when the electric vehicle starts charging, satisfies the normal distribution, and its probability density function is shown as follows: ； Where: Start charging your electric car. As the expectation, take 17.47, is the standard deviation, taking 3.4, and the coefficient of determination R square is 0.9899; the network time, that is, the time when the electric car leaves home, also satisfies the normal distribution, and its probability density function is shown as follows: ； Where: The first time an electric car leaves home, is the expected value, take 7.70, is the standard deviation, which is 3.27, and the coefficient of determination R square is 0.9987; Step 2.2, construct the objective function with the goal of optimizing the economic efficiency of electric vehicle charging stations and minimizing the impact on the distribution network; For charging stations, the goal is to maximize their revenue and minimize the impact on the distribution network. The mathematical expression is: ； Where: are the charging and discharging power of EV respectively; are the electricity sales and purchase prices of the charging station, is the regulating factor, is the interaction power between the charging station and the distribution network; Step 2.3, consider the charging and discharging power of electric vehicles, power balance of the power grid, charging and discharging status, upper and lower limits of electricity prices, and construct constraint conditions; Step 2.3.1, Charge and discharge power constraints: ； Where: At the same time The charging and discharging power of electric vehicles, Maximum charging power for electric vehicles; Step 2.3.2, Power Constraint: The maximum load that the distribution network can bear To meet the basic load of the network EV charging and discharging power The sum is: ； Step 2.3.3, SOC constraint when off-grid: In order to meet user needs, electric vehicles must meet the following requirements when they are off-grid: ； Where: It is the SOC of the electric vehicle when it is off-grid; Step 2.3.4, Electric vehicle charging and discharging state constraints: ； Where: The time when EV is connected to the grid; Step 2.3.5, Dynamic electricity price upper and lower limit constraints: Taking into account the operating costs of charging stations and the consumption level of users, the upper and lower limits of the dynamic electricity price for charging and discharging are: ； Where: They are the upper and lower limits of charging and discharging electricity prices respectively; Step 2.4, use particle swarm optimization algorithm to find the optimal solution; Using particle swarm optimization, that is, particle swarm optimization, PSO algorithm, particles During the iteration process, by tracking its individual optimal solution and the global optimal solution To update your new position and speed: ； Where: Indicates the current iteration number. They represent the particles in the optimization process. exist d dimensional velocity and position, ω is the inertia weight; , is the learning factor; , is a random number uniformly distributed between [0,1]; Step 3: Use the Monte Carlo sampling method to calculate the reliability index of each load point in the distribution network; Step 4: Perform multiple simulations, and collect and accumulate the reliability indexes of each load point in the distribution network to calculate the reliability index of the entire system; Step 5: Using the optimized sequential Monte Carlo simulation method, the reliability of the distribution network containing distributed power sources and electric vehicles is calculated after connecting different numbers of electric vehicles and adopting different orderly charging and discharging control strategies.

2. A method for evaluating the reliability of a distribution network containing distributed power sources and electric vehicles according to claim 1, characterized in that: The probability distribution function of the wind turbine output model in Step 1 is and the probability distribution function of the photovoltaic generator output model It can be expressed as follows: The probability distribution of wind turbine output under: Where: is the cut-in wind speed; To cut out the wind speed; is the rated wind speed; A, B, and C can be obtained through calculation, and their values ​​are related to the parameter configuration of the fan; The probability density function of photovoltaic generator output: Where: is the output power of the photovoltaic generator; is the maximum output power of the photovoltaic generator; Probability distribution function of photovoltaic generator output: 。 3. A method for evaluating the reliability of a distribution network containing distributed power sources and electric vehicles according to claim 2, characterized in that: The specific steps of Step 3 are: Step 3.1, set the time series simulation period N of the distribution network, input various parameters of the distribution network, and initialize the data and simulation time T=0; Step 3.2, number each component of the distribution network, sample its status, and calculate the normal working duration of each component, i.e., time to fault, TTF, and fault repair time, i.e., time to repair, TTR; In the formula, , Components Corresponding failure rate and repair rate; , They are all random numbers uniformly distributed between (0, 1), so the normal working duration and fault repair time of the components obey the exponential distribution; Step 3.3, through Monte Carlo sampling, obtain the sequence of the operating status of each component of the system. TTF Find the minimum value and record the corresponding component number as , the component is the faulty component, and the fault time is accumulated ; Step 3.

4. Find the faulty component The location of the load is determined, the load range affected by it is determined, and the fault repair time is calculated. , and classify the load; The unaffected load is Class I load, the load that can be restored to normal working state after the faulty component is replaced or repaired is Class II load, and the load in the isolated island is Class III load; Step 3.5: For Class III load, perform the following processing: set a random number that follows a uniform distribution between (0, 1) ,judge The probability of successful switching to an isolated island If the size , then the island switching is considered successful, otherwise it is considered a failure and all loads in the island are powered off; Step 3.6: Use the probability model of wind turbine and photovoltaic generator output established in Step 1 to sample the power of DG , Then, using the EV charging load model established in Step 2, generate 8760 hours of charging load according to the time series , and superimposed with the original load in the island, the total load is obtained ,Determine whether the fault time is within the charging and discharging time of the electric vehicle. If not, count the power outages of each load point in the island, otherwise proceed to the next step; Step 3.7: DG output within the island range , With total load If there is a difference, the load is cut off and the corresponding fault time and fault number are accumulated. Otherwise, proceed to the next step. Step 3.8, determine whether the simulation lifespan has been reached, if so, end the simulation process and proceed to the next step, otherwise return to Step 3.2; Step 3.9, calculate the system reliability indicators: average number of system power outages SAIFI, average power supply availability ASAI, average system power outage time SAIDI, and average system power shortage AENS.

4. A method for evaluating the reliability of a distribution network containing distributed power sources and electric vehicles according to claim 3, characterized in that: In the above-mentioned Step 3.4, after the loads are classified, the number of failures and the failure time of each load are counted. For Class I loads, since they are not affected by the faulty components, their total number of power outages and power outage time remain unchanged. For Class II loads, the outage time is added to the total power outage time, and the total number of power outages is increased by 1.

5. A method for evaluating the reliability of a distribution network containing distributed power sources and electric vehicles according to any one of claims 1 to 4, characterized in that: In the above-mentioned Step 2 and Step 5, the reliability of the distribution network is calculated and analyzed from the perspective of the charging and discharging modes of electric vehicles using disordered charging and discharging, time-of-use electricity prices, and dynamic time-of-use electricity prices.

Images