A power distribution network reliability optimization evaluation method based on vehicle-network interaction

Abstract

The application relates to a power distribution network reliability optimization evaluation method based on vehicle-network interaction, which comprises an electric vehicle driving model, a vehicle-network coupling element reliability model, an electric vehicle charging and discharging incentive model and a fault recovery model with electric vehicle access, so as to solve the modeling problem of the influence of the travel characteristics and the charging and discharging characteristics of the electric vehicle under the guidance of the charging and discharging incentive model on reliability, thereby most effectively measuring the influence of the electric vehicle access to the power distribution network on the reliability index. The application can realize reliability optimization evaluation on the basis of considering vehicle-network interaction and has applicability when the operation scene changes, thereby making up for the problem of insufficient consideration of factors related to the electric vehicle access to the power distribution network in the existing method.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of reliability evaluation of power distribution network, and particularly relates to a reliability optimization evaluation method of power distribution network based on vehicle-to-grid interaction. BACKGROUND

[0002] The existing reliability optimization evaluation method of power distribution network considering electric vehicle access is as follows: a reliability model of power distribution network element and a probability model of electric vehicle access to power distribution network are established, the reliability parameters of electric vehicle access node are calculated by using the minimum path method or the fault mode and consequence analysis method, and the electric quantity shortage and other indexes of electric vehicle charging are obtained.

[0003] At present, there is a lack of management experience in China about large-scale electric vehicle access to power distribution network, and it is necessary to learn from international advanced model methods, carry out reliability research based on the prediction analysis of electric vehicle access location, access time and charging and discharging power, incentive guidance scheme and fault modeling of related elements. Due to large-scale access of electric vehicles to power distribution network, the reliability of the access node will affect the charging satisfaction of electric vehicle users, and the traditional reliability index is difficult to consider the differentiated characteristics of the users connected to the node through the indexes of power outage time and power outage frequency, and then quantify the influence of these users on the reliability index of power distribution network.

[0004] The existing reliability evaluation method of electric vehicle access to power distribution network lacks corresponding optimization model, and the regulation and control of electric vehicle charging power during fault and the discharge power of vehicle to power grid have great influence on the fault recovery process of power distribution network, thereby affecting the reliability index. SUMMARY

[0005] The purpose of the present application is to provide a reliability optimization evaluation method of power distribution network based on vehicle-to-grid interaction, which comprehensively considers the travel characteristics and charging and discharging characteristics of electric vehicles under the guidance of the charging and discharging incentive model, so as to measure the influence of electric vehicle access to power distribution network on the reliability index to the greatest extent, and realize the reliability optimization evaluation.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0007] A reliability optimization evaluation method of power distribution network based on vehicle-to-grid interaction, comprising the following steps:

[0008] Obtain first data, the first data is collected by a collection terminal, and the first data includes electric vehicle driving data, vehicle-to-grid coupling element fault information and power distribution network fault recovery information;

[0009] Based on the first data, obtain an automobile driving model, a vehicle-to-grid coupling element reliability model, an electric vehicle charging and discharging incentive model and a fault recovery model of electric vehicle access, respectively;

[0010] The fault recovery model for electric vehicle access includes a two-stage, three-layer distributed robust optimization model; the distributed robust optimization model is used to determine the power supply range of each electric vehicle discharge and to find the worst-case probability distribution that minimizes the costs of each item in the power supply recovery process.

[0011] Furthermore, the electric vehicle driving model index includes the driving mileage of the electric vehicle. The driving mileage of the electric vehicle is represented by a probability density function, which is expressed as follows:

[0012]

[0013] In the formula, d is the distance between the starting point and the destination, and μ and σ are the expectation and variance, respectively;

[0014] The round-trip time of an electric vehicle between its destinations follows a normal distribution, represented as:

[0015]

[0016] S d =S0d;

[0017] In the formula, μ t σ t These represent the expected value and variance, respectively, where S0 is the state of charge (SOC) required for a one-way trip of the electric vehicle; S... d The total SOC consumed for the total mileage traveled;

[0018] Wherein, SOC is represented as:

[0019] S i =1-αS d ;

[0020] In the formula, α is a random row vector uniformly distributed within (0,1), i is the electric vehicle number, and S... i Let SOC be the state of the i-th electric vehicle.

[0021] Furthermore, the reliability model index of the vehicle-to-grid coupling element includes a discrete Copula function model for equipment fault edges. This Copula function model describes the correlation between charging pile faults, and the formula is as follows:

[0022]

[0023] In the formula, F(x) is the cumulative probability distribution function of the running state, u i P represents the charging pile's operating status, and P represents the charging pile's forced shutdown rate.

[0024] Further, the electric vehicle charging and discharging incentive model index includes flexibility supply of electric vehicles supporting power distribution network, and the supply capacity is mainly constrained by charging pile power and travel habits of vehicle owners, which can be expressed as:

[0025]

[0026]

[0027] In the formula, F1 and F2 represent the flexibility supply of electric vehicles supporting the power distribution network; P(t) and E(t) are the charging and discharging power and the battery capacity of the cluster EV at time t.

[0028] Further, the electric vehicle charging and discharging incentive model index also includes a virtual price, which guides the charging and discharging of electric vehicles through the virtual price, and the relationship between the virtual price and the power distribution network node load is:

[0029]

[0030] In the formula, is the virtual price of the k period when the electric vehicle l is connected; α0, α1 and α2 are virtual price adjustment coefficients; is the total load of the power distribution network in the k period when the electric vehicle l is connected, including electric vehicle load and other load; is the corresponding predicted total load; T is the electric vehicle connection time.

[0031] Further, the fault recovery model of the electric vehicle access is represented by a power supply reliability index, and the power supply reliability index includes an average power outage frequency index, and the formula is as follows:

[0032]

[0033] In the formula, λ represents the number of power outages of each user, L represents the total load power connected to the load point i, and a i represents the reliability demand coefficient of the user reliability price difference.

[0034] Further, the power supply reliability index includes a power outage time index, and the formula is as follows:

[0035]

[0036] In the formula, SAIDI L represents the user power outage time and power outage frequency weighted by load; U represents the power outage time of each user.

[0037] The power supply reliability index includes an energy shortage index, and the formula is as follows:

[0038]

[0039] Further, the robust optimization model is as follows:

[0040]

[0041] In the formula, P pk is the probability of scenario k, C los,t , C rep,t , C gen,t , C opr,t are outage cost, repair cost, generation cost and manual operation cost respectively;

[0042] The parameter δ R is defined to quantify the robustness degree of each scheme to different operation scenarios, and the formula is as follows:

[0043]

[0044] In the formula, S Dq represents the system average interruption time of scenario q, γ q ' represents the severity of scenario q, δ R reflects the severity of the change of the system reliability index with the change of the operation scenario, and the closer the value is to 0, the stronger the robustness is.

[0045] From the above technical scheme, the present application is used in the field of reliability optimization evaluation of distribution network, to solve the modeling problem of the influence of the travel characteristics and the charging and discharging characteristics of electric vehicles under the guidance of the charging and discharging incentive model on reliability, so as to measure the influence of the electric vehicles accessing the distribution network on the reliability index to the greatest extent. The model is composed of an electric vehicle driving model, a reliability model of a vehicle-to-grid coupling element, an electric vehicle charging and discharging incentive model, and a fault recovery model considering the access of electric vehicles. From the analysis results of the examples, it can be seen that the present application can realize reliability optimization evaluation on the basis of considering vehicle-to-grid interaction, and has applicability when the operation scenario changes, so as to make up for the problem of insufficient consideration of factors related to the access of electric vehicles to the distribution network in the existing method, and realize reliability optimization evaluation. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 The flow chart of the method model for reliability optimization evaluation of the distribution network considering vehicle-to-grid interaction;

[0047] Figure 2 The distribution network topology of the embodiment. DETAILED DESCRIPTION

[0048] The present application will be further described below in combination with the drawings:

[0049] As Figure 1As shown, the power distribution network reliability optimization evaluation method based on vehicle-network interaction of the embodiment is composed of an output voltage sampling circuit, a sampled voltage comparison amplification circuit and a feedback voltage superposition conversion circuit.

[0050] A power distribution network reliability optimization evaluation method based on vehicle-network interaction, comprising the following steps:

[0051] S1: obtaining first data, the first data being collected by a collection terminal, the first data including electric vehicle driving data, vehicle-network coupling element fault information and power distribution network fault recovery information;

[0052] S2: based on the first data, obtaining an automobile driving model, a vehicle-network coupling element reliability model, an electric vehicle charging and discharging excitation model and an electric vehicle access fault recovery model respectively;

[0053] S3: the electric vehicle access fault recovery model includes a two-stage three-layer distributed robust optimization model; the distributed robust optimization model is used to determine the power supply range of each electric vehicle discharging and find the worst scenario probability distribution that maximizes the minimum value of the power supply recovery process cost.

[0054] The electric vehicle driving model includes a driving mileage index of the electric vehicle, and the electric vehicle driving model represents the driving mileage distribution of each electric vehicle by a probability density function, as shown in the formula.

[0055]

[0056] In the formula, d is the distance between the starting point and the destination, μ and σ are the expectation and variance. The time of the electric vehicle between the destination is subject to a normal distribution, which is represented as:

[0057]

[0058] S d =S0d (3)

[0059] In the formula, μ t and σ t are the expectation and variance. S0 is the SOC required for a single trip of the electric vehicle; S d is the total SOC consumed by the driving mileage.

[0060] According to the single trip driving distance of the electric vehicle, in order to consider the difference of the SOC of each electric vehicle, the SOC of the electric vehicle can be represented as:

[0061] S i =1-αS d (4)

[0062] In the formula, α is a random row vector uniformly distributed in (0, 1), and i is the electric vehicle number. Si is the soc of the ith electric vehicle.

[0063] In the reliability model of the vehicle-grid coupling element, the charging or discharging of the electric vehicle to the power distribution network needs to rely on the charging and discharging equipment. Taking the charging pile as an example, the charging pile is exposed to the outdoor environment for a long time, and the failure risk is high. In addition to random failure, the failure of the charging pile is affected by factors such as weather, use frequency, and customer use habits, so the failure of the charging piles in the same charging station has correlation. For example, the charging piles in the same charging station are used with similar frequency, so their failure probabilities have positive correlation. The charging piles in the same area are affected by weather factors such as rainstorms and typhoons, so the failures in this aspect also have correlation.

[0064] When a two-state model is used, the operating state u i of the charging pile is a 0-1 discrete variable (0 represents normal operation, and 1 represents failure shutdown). An equipment failure edge discrete Copula function model is proposed to describe the correlation between the failures of the charging piles. Assuming that the forced outage rate of the charging pile is P, the cumulative probability distribution function of the operating state is as follows:

[0065]

[0066] In the formula, F(x) is the cumulative probability distribution function of the operating state, u i is the operating state of the charging pile, and P is the forced outage rate of the charging pile.

[0067] The cumulative distribution function CDF and the probability density function PDF of the Copula function are as follows:

[0068]

[0069]

[0070] In the formula: k is a correlation measure, and its range is (0, 1].

[0071] The probability density of the Copula function is higher near the maximum value, that is, the correlation of the simultaneous failure of the charging pile equipment is more obvious than other states, which conforms to the correlation characteristics of the failure of the charging pile.

[0072] In the electric vehicle charging and discharging incentive model, the scheduling of electric vehicles is generally flexibly scheduled by electric vehicle aggregators. Compared with traditional load and power types, electric vehicles have fast response speed and strong flexibility. The supply capacity is mainly constrained by the charging pile power and the travel habits of the vehicle owner, and can be represented as

[0073]

[0074]

[0075] In the formula: F1, F2 are the upper and lower flexibility of the power supply of the electric vehicle supporting the power distribution network; P(t) and E(t) are the charging and discharging power and the battery capacity of the cluster EV at time t.

[0076] When high-power large-scale electric vehicles access the power distribution network, accurate modeling of each electric vehicle cannot be performed due to the limitation of the calculation efficiency of the model solving tool. When the electric vehicle aggregator induces and incentivizes large-scale electric vehicles to charge and discharge, the virtual price can be used to guide the charging and discharging behavior of the electric vehicles, and this form of price can reflect the supply and demand relationship in real time. The virtual price has real-time, abstract, and targeted. In the electric vehicle incentive model, the virtual price is used to guide the formulation of the electric vehicle charging and discharging plan, and is not used to calculate the actual cost.

[0077] The relationship between the virtual price and the load of the power distribution network node is:

[0078]

[0079] In the formula, k is the virtual price of the k period when the electric vehicle l accesses; α0, α1, α2 are virtual price adjustment coefficients; is the virtual price of the k period when the electric vehicle l accesses; α0, α1, α2 are virtual price adjustment coefficients; is the total load of the power distribution network in the k period when the electric vehicle l accesses, including the electric vehicle load and other loads. is the corresponding predicted total load; T is the access time of the electric vehicle. Therefore, the virtual price and the load of the access node are positively correlated, because the peak shaving pressure of the high-load node is high, and a higher virtual price needs to be formulated to guide the electric vehicles to shift to other nodes for charging. In addition, the virtual price is set separately for electric vehicles, and each vehicle enjoys its own virtual price, respecting customer privacy. During fault recovery, a specific scheduling model is used to attract electric vehicles to restore power supply to critical nodes. The virtual price at this time is determined according to the load importance.

[0080] The conventional power supply reliability index is generally based on SAIDI and SAIFI, and is a power supply reliability index based on the number of users, also known as a user power supply reliability index, which belongs to a demand-side power supply reliability index. In the case of electric vehicle access, the traditional index cannot measure the difference between electric vehicle charging demand and other types of load, so a reliability index considering user differentiation is proposed. First, the improved system average interruption frequency index.

[0081]

[0082] In the formula: λ, L represent the number of power outages of each user and the total load power connected to the load point i. iThe reliability demand factor considering the user reliability price difference, which is normalized by the average reliability price of the load point.

[0083]

[0084]

[0085] Therefore, SAIFI L The user outage time and outage frequency weighted by the load power. In order to distinguish from the traditional user-based power supply reliability index, the subscript L is added.

[0086] Similarly, the improved system average interruption time is as follows.

[0087]

[0088] In the formula, U represents the outage time of each user. Therefore, SAIDI L The user outage time and outage frequency weighted by the load power. The improved power shortage index is as follows.

[0089]

[0090] Considering the fault recovery model of electric vehicle access, considering that the electric vehicle charging load has great uncertainty, in order to find the optimal fault recovery scheme under the worst scenario probability distribution, a two-stage three-layer distribution robust model is proposed. The first stage min problem determines the power supply range of each electric vehicle discharge, and the second stage max-min problem finds the worst scenario probability distribution that minimizes the maximum cost of the power supply recovery process, which is expressed as follows

[0091]

[0092] In the formula, p k is the probability of scenario k, C los,t , C rep,t , C gen,t , C opr,t are the outage cost, repair cost, generation cost, and manual operation cost respectively. The calculation method of each cost is as follows

[0093] C los = ξ F SAIFI L + ξ D SAIDI L + ξ E ENS (18)

[0094]

[0095]

[0096] C opr = v · n s (21)

[0097] where ξ F , ξ D , ξ E , are the reliability penalty coefficients, representing the cost of the power grid company to compensate the customers for each unit of reliability index exceeding. α represents the ratio of maintenance cost to the investment cost of the component. ΔP MG,mt , ΔQ MG,mt are the active and reactive power of the microgrid m supporting the distribution grid at time t. α P , α Q are the corresponding generation cost coefficients. v is the cost of each switch operation for fault, n s is the number of faults. x is the first stage variable, y is the second stage variable. h(x)≤0 is the constraint of the first stage. g(x,y k )≤0 is the constraint of the second stage. In order to ensure the rationality of the scenario probability, the norm constraint is constructed to limit the probability distribution of the uncertain scenario, which is centered on the initial probability distribution of K discrete typical scenarios (ξ1, ξ2,..., ξ K ). The formula is as follows

[0098]

[0099] where p' k is the initial scenario probability distribution value; κ is the allowed deviation of the norm constraint. The larger the value is, the more conservative the model is, and vice versa. When the reliability requirement is high, the capacity of the backup power supply can be discretized, and the cost-benefit analysis method can be used to find the worst scenario. That is, find the capacity of the backup power supply that makes the system reduce a point of transferable load, which requires the least capacity to be cut. Add the cut capacity and continue to find, until the cumulative cut capacity reaches the boundary of the uncertain set. Calculate in the scenario after the capacity of the backup power supply is cut, and the solution can be obtained. Define the parameter δ R to quantify the robustness of each scheme to different operating scenarios, as shown in the formula.

[0100]

[0101] where S Dq represents the system average interruption time of scenario q. γ q ' represents the severity of scenario q, as shown in the formula.

[0102]

[0103] Therefore δ RReflects the system reliability index with the change of the running scene changes the degree of severity, so the closer to 0, the stronger the robustness.

[0104] As Figure 2 shown, the node load adopts typical data, in order to verify the efficiency of the reliability optimization scheme formed by the model, the following two schemes are designed for comparative analysis. Scheme 1 uses the traditional fault recovery method for load transfer in the operation stage; scheme 2 is a reliability improvement scheme using robust optimization considering vehicle-network interaction. Four scenarios are selected, and the methods of the two schemes are used for optimization. Take γ as 220kw, and the calculation results are shown in Table 1.

[0105] Table 1 Robustness comparison of two schemes under four scenarios

[0106] Scenario 1 Scenario 2 Scenario 3 Scenario 4 Delta P EV / kW 2080 2160 2300 2480 S D / ha -1 ]]> 4.1250 4.0836 3.9675 3.8863 S D,R / ha -1 ]]> 4.1250 4.1250 4.1250 4.1250

[0107] In the table, S D and S D,R respectively represent the system average outage time under the two reliability optimization strategies of scheme 1 and scheme 2. The calculation parameter δ R is used to quantify the robustness of each scheme under different running scenarios, and the δ R of scheme 3 is-5.1646×10 -4 , and the δ R of scheme 4 is 0. It can be seen that the robustness of scheme 4 is stronger, and it has strong adjustment ability and load power supply recovery ability when the running scene changes.

[0108] The above-described embodiments are merely preferred embodiments of the present application and do not limit the scope of the present application. Without departing from the design spirit of the present application, various modifications and improvements to the technical solutions of the present application made by those of ordinary skill in the art shall fall within the protection scope of the present application as defined by the claims.

Claims

1. A power distribution network reliability optimization evaluation method based on vehicle-network interaction, characterized in that, The method comprises the following steps: Obtaining first data collected by a collection terminal, wherein the first data comprises electric vehicle driving data, vehicle-grid coupling element fault information and power distribution network fault recovery information; Based on the first data, obtaining an automobile driving model, a vehicle-grid coupling element reliability model, an electric vehicle charging and discharging incentive model and an electric vehicle access fault recovery model respectively; The electric vehicle access fault recovery model comprises a two-stage three-layer distributed robust optimization model, which is used to determine the power supply range of each electric vehicle discharging and find the worst scenario probability distribution that maximizes the minimum value of the power supply recovery process cost; The electric vehicle charging and discharging incentive model comprises a flexible supply index of electric vehicles supporting the power distribution network, and the supply capacity is represented by the following formula: In the formula, F 1. F 2 indicates that electric vehicles support the flexible supply of the power distribution network; P ( t ) and E ( t For cluster EVs in t The charging and discharging power and battery capacity at any given time; The electric vehicle charging and discharging incentive model further comprises a virtual electricity price index, which guides the electric vehicle charging and discharging through the virtual electricity price, and the relationship between the virtual electricity price and the power distribution network node load is as follows: In the formula, plk L k is the virtual electricity price of the period k when the electric vehicle 1 is connected; α 0, α 1, α 2 are virtual electricity price adjustment coefficients; L k is the total load of the distribution network in the period k when the electric vehicle 1 is connected, including the electric vehicle load and other loads;​​​ L ( k )'is the corresponding predicted total load; T is the electric vehicle access time; The robust optimization model is as follows: wherein: pk is the probability of scenario k, C los, t , C rep, t , C gen, t , C opr, t are the loss of service cost, repair cost, generation cost, and manual operation cost, respectively. Definition of parameters δ R is used to quantify the robustness of each scheme to different operating scenarios, and is given by wherein S D q represent the scenario q of the system average interruption duration, yq ' represents the scenario q of the system average interruption duration, δ R reflects the severity of the change of the system reliability index with the change of the operating scenario, the closer the value is to 0, the stronger the robustness.

2. The power distribution network reliability optimization evaluation method based on vehicle-network interaction according to claim 1, characterized in that, The electric vehicle driving model comprises a driving mileage index of the electric vehicle, and the driving mileage of each electric vehicle is represented by a probability density function, and the formula is as follows: In the formula, d is the distance from the departure point to the destination, μ , σ are the expectation and variance, respectively; The time of the electric vehicle between destinations is subject to a normal distribution, which is represented as follows: Sd = S 0 d ； wherein μ t, σ t are the desired and variance, respectively, S 0 is the SOC required for a single trip of an electric vehicle; Sd is the total SOC consumed for a trip. Wherein, SOC is represented as: Si = 1- αSd ； wherein α is a random row vector uniformly distributed in (0, 1), i is the electric vehicle number, Si is the soc of the i-th electric vehicle.

3. The power distribution network reliability optimization evaluation method based on vehicle-network interaction according to claim 1, characterized in that, The vehicle-grid coupling element reliability model comprises a device fault edge discrete Copula function model index, and the Copula function model is used to describe the correlation between charging pile faults, and the formula is as follows: In the formula, F (x) is a cumulative probability distribution function of the operating state, ui is the operating state of the charging pile, and P is the forced outage rate of the charging pile.

4. The power distribution network reliability optimization evaluation method based on vehicle-network interaction according to claim 1, characterized in that, The electric vehicle access fault recovery model is represented by a power supply reliability index, and the power supply reliability index comprises an average power outage frequency index, and the formula is as follows: In the formula, λ represents the number of power outages for each user, L represents the total load power connected to the load point i, ai represents the reliability demand factor of the user reliability price difference.

5. The power distribution network reliability optimization evaluation method based on vehicle-network interaction according to claim 4, characterized in that, The power supply reliability index comprises a system average power outage time index, and the formula is as follows: Wherein, SAIDIL represents the user power outage time and power outage frequency weighted by load, and U represents the power outage time of each user.

6. The power distribution network reliability optimization evaluation method based on vehicle-network interaction according to claim 5, characterized in that, The power supply reliability index comprises an electricity shortage index, and the formula is as follows: 。

Images