Hierarchical optimization method and system for active-voltage stepped droop curve of charging station

Abstract

The invention discloses a layered optimization method and system for an active-voltage stepped droop curve of a charging station. The method comprises the following steps: constructing an SoC model of a single electric vehicle and a charging station SoC model after aggregation of the electric vehicles; establishing a charging station SoC interval condition value-at-risk optimization model, obtaining the SoC interval of the charging station and the corresponding power adjustable range through rolling optimization in each scheduling period, and reporting the SoC interval and the corresponding power adjustable range to a power distribution network scheduling center; on the level of the power distribution network, based on the determined power adjustable range and considering the source load uncertainty of the power distribution network, an active power-voltage step droop curve optimization model of the charging station is established, and the optimal step droop curve of each scheduling time period is solved and issued to the charging station; and the charging station decomposes the power of each segment of the stepped droop curve to obtain the charging power of each electric vehicle corresponding to different segments, and the charging station dynamically adjusts the charging active power of the electric vehicle according to the real-time voltage measurement value and the power decomposition result of the current scheduling time period.

Description

Technical Field

[0001] The present invention belongs to the field of power distribution networks, and in particular relates to a method and system for hierarchical optimization of active power-voltage step droop curves of charging stations. Background Art

[0002] The penetration of renewable energy within distribution networks is gradually increasing. However, renewable energy output is characterized by significant intermittency and volatility. Furthermore, distribution network load is highly uncertain due to factors such as user electricity consumption and industrial restructuring. Consequently, the dual uncertainties of renewable energy and load exacerbate the complexity of power flow distribution within distribution networks, leading to voltage excursions and fluctuations, threatening the safe operation of distribution networks.

[0003] Electric vehicles have also garnered widespread attention for their energy-saving and emission-reduction advantages. As sales of new energy vehicles continue to climb, the scale of electric vehicle charging stations is also expanding. However, it's noteworthy that most electric vehicle owners spend more time parked at charging stations than it would take to reach the required charge at maximum power. This phenomenon suggests that electric vehicles have significant potential as a regulated load resource during these periods. For example, when renewable energy output is in excess and the grid voltage is high, electric vehicles adaptively increase charging power to consume excess energy and lower the grid voltage. When renewable energy output is insufficient and the grid voltage is low, electric vehicles adaptively decrease charging power to reduce load and increase the grid voltage. To achieve this, it's necessary to accurately control the adjustable power range of charging stations so that they can participate in voltage regulation within the distribution network. No relevant solutions have been proposed in the prior art. Summary of the Invention

[0004] Purpose of the Invention: This invention proposes a hierarchical optimization method and system for charging station active power-voltage step droop curves, taking into account the aggregation of electric vehicle (EV) SoC intervals. This system provides a hierarchical coordination framework for coordinating SoC interval optimization with step droop curve optimization for electric vehicle charging stations. Furthermore, a comprehensive active power-voltage step droop curve model for the charging station is established, droop parameters are optimized, and the step droop curve is decomposed to obtain the charging power of each electric vehicle in different segments.

[0005] Technical solution: A method and system for hierarchical optimization of active power-voltage step droop curves at charging stations, comprising the following steps:

[0006] A method for hierarchical optimization of active power-voltage step droop curves of a charging station, characterized by comprising the following steps:

[0007] (1) Based on the charging behavior data of electric vehicle users, including the time of entering the charging station, the time of leaving the charging station, the initial state of charge (SoC) when entering the charging station, and the expected SoC when leaving the charging station, a single electric vehicle SoC model is constructed, and then a charging station SoC model after the electric vehicles are aggregated is constructed;

[0008] (2) Considering the number of electric vehicles newly arriving at the charging station and the uncertainty of charging behavior, a charging station SoC interval conditional risk value optimization model is established. The optimization goal is to minimize the charging cost and conditional risk value. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle of the charging station is calculated, and the power range is reported to the distribution network dispatching center.

[0009] (3) At the distribution network system level, considering the uncertainty of renewable energy output and load in the distribution network, based on the power adjustable range reported by the charging station in each scheduling period, an optimization model of the active power-voltage step droop curve of the charging station is established, and the optimal step droop curve of each scheduling period is obtained and sent to the charging station;

[0010] (4) Based on the step droop curve of the current scheduling period, the charging station decomposes the power of each segment of the droop curve to obtain the charging power of each electric vehicle corresponding to the different segment powers, and dynamically adjusts the charging active power of the electric vehicle based on the real-time voltage measurement value and the power decomposition result.

[0011] Furthermore, the SoC model of a single electric vehicle in step (1) includes:

[0012] (a) The charging power of electric vehicles during parking periods should meet the power range:

[0013]

[0014] Among them, t represents the time period index, m is the car index, i is the distribution network node index, and s is the electric vehicle scene that has just arrived at the charging station. is the charging power of electric vehicle m parked at charging station node i in time period t under scenario s, is the maximum charging power allowed for electric vehicles, It is the parking period of the car;

[0015] (b) The charging power of electric vehicles during non-parking hours is 0:

[0016]

[0017] (c) Each electric vehicle is dispatched only during the parking period, which is as follows:

[0018]

[0019] in, and are the arrival and departure time periods of the car respectively;

[0020] (d) The SoC constraints for electric vehicle parking periods are as follows:

[0021]

[0022] in, SoC for electric vehicles, Battery capacity for electric vehicles, SoC max and SoC min They are the maximum SoC and minimum SoC allowed for electric vehicles, represents the charging efficiency of electric vehicles, Δt is the length of the unit charging scheduling time; constraint (4) represents the charging efficiency of electric vehicles during the parking period according to The change process of SoC during charging; Constraint (5) represents the range constraint of the upper bound of SoC of electric vehicles during parking period;

[0023] (e) The SoC constraints for electric vehicles during entry and exit periods are as follows:

[0024]

[0025] in, The initial SoC for electric vehicles, is the SoC that the electric vehicle expects to achieve when leaving; constraint (6) means taking the initial SoC of the electric vehicle as the entry time t in Constraint (7) requires that the SoC when the electric vehicle leaves the vehicle should be greater than the SoC that the owner originally hoped to achieve.

[0026] Furthermore, in step (1), the charging station SoC model after electric vehicle aggregation includes:

[0027] (f) The charging station power constraints are as follows:

[0028]

[0029] in, is the power consumed by the charging station, is the overall maximum charging power of the charging station, M t,i,s is the set of electric vehicles m parked at the charging station at node i in time period t under scenario s; constraint (8) indicates that the power consumed by the charging station is the sum of the charging power of all electric vehicles; constraint (9) indicates that the overall power consumed by the charging station should meet the power limit;

[0030] (g) The charging station cost constraint is as follows:

[0031]

[0032] where, is the power The corresponding charging cost, Pr t is the charging price of time period t;

[0033] (h) The charging station SoC model after aggregation of electric vehicles is as follows:

[0034]

[0035] where, is the charging station SoC; constraint (11) calculates the charging station SoC after aggregation of electric vehicles according to the SoC of each vehicle in the charging station and the rated capacity of the battery.

[0036] Further, in step (2), the charging station SoC interval conditional risk value optimization model is as follows:

[0037]

[0038] where, is the conditional risk value of the upper limit of the SoC interval, is the conditional risk value of the lower limit of the SoC interval, T is the set of time period indexes, S is the set of electric vehicle scenarios s newly arriving at the charging station, N s is the total number of scenarios, β1 and β2 are weight coefficients of different objectives respectively, is the upper limit of the SoC interval, is the lower limit of the SoC interval, and θ respectively represent the maximum and minimum values of the SoC interval width, and are auxiliary variables; formula (12) indicates that the optimization objective is to minimize the charging cost and the conditional risk value; constraint (13) indicates the SoC interval width limit; constraint (14) indicates the range limit of the SoC interval value; constraints (15)-(17) are the conditional risk value constraints of the upper limit of the SoC; constraints (18)-(20) are the conditional risk value constraints of the lower limit of the SoC.

[0039] Further, in step (2), the SoC interval of the charging station is obtained through rolling optimization in each scheduling time period in turn, and the power adjustable range of the charging station corresponding to the first time period in the rolling optimization cycle is calculated based on the SoC interval, and the calculation method is as follows:

[0040]

[0041] in, It is the minimum power that the charging station can upload to the distribution network and will serve as the power lower limit of the step droop curve. The maximum power that the charging station uploads to the distribution network will serve as the upper limit of the step droop curve. is the SoC determined one hour before the charging station; formula (21) calculates the minimum power of the charging station; formula (22) calculates the maximum power of the charging station.

[0042] Furthermore, in step (3), based on the power adjustable range reported by the charging station in each scheduling period, an optimization model of the active power-voltage step droop curve of the charging station is established, including:

[0043] (a) The expression for the step droop curve of electric vehicle charging stations is as follows:

[0044]

[0045] in, is the charging active power of charging station node i in time period t under the uncertainty scenario c of renewable energy output and load of distribution network, V t,i,c is the node voltage value, n is an odd number used to represent the number of segments of the step droop curve, is the voltage value of the inflection point of the step droop curve of the charging station, P 1,2,…,n,t,i is the power value of each segment of the step droop curve of the charging station;

[0046] (b) The power constraints for each segment are as follows:

[0047]

[0048] Constraint (24) limits the value range of each segment power;

[0049] (c) The knee voltage constraint is as follows:

[0050]

[0051] in, V and Respectively represent the minimum and maximum voltage allowed, and ζ represents the length of each segment of the step droop curve

[0052] The minimum threshold value that must be greater than; constraint (25) indicates the order of magnitude and constraint range of each inflection point voltage; constraint (26) indicates that there must be a certain distance between two adjacent inflection point voltage values.

[0053] Furthermore, in step (3), the optimal step droop curve for each scheduling period is obtained and sent to the charging station, including:

[0054] Based on the reconstruction of the staircase droop curve expression by the big M method, first define n binary decision variables, constraints as follows:

[0055]

[0056] wherein, is the binary decision variable of the staircase droop curve; constraint (27) ensures that each charging station in each group of scenarios only has a certain segment of the droop curve activated in each time period by limiting the value of the binary variable; constraint (28) represents the charging station power;

[0057] Since constraint (28) contains a bilinear term, auxiliary variables and multiple inequality constraints are introduced to represent the charging station power as follows:

[0058]

[0059] wherein, is the auxiliary variable; M is a specified positive number, which is greater than the maximum charging power of the charging station; constraint (29) represents the charging station power by introducing auxiliary variables; constraints (30)-(31) represent the values of the auxiliary variables corresponding to different binary variable values;

[0060] The segments of the staircase droop curve corresponding to the voltage are represented by multiple inequality constraints as follows:

[0061]

[0062] wherein, v is a specified positive number, less than one thousandth of the maximum voltage value, used to represent the step of the segment inflection point; constraint (32) represents that the node voltage is between the two inflection point voltages of the corresponding segment of the staircase droop curve under different binary variable values;

[0063] The linear power flow constraint model of the distribution network is established as follows:

[0064]

[0065] wherein, P t,hi,c and Q t,hi,c are the active power and reactive power flowing through branch hi, P t,ij,c and Q t,ij,c are the active power and reactive power flowing through branch ij, V0 is the reference voltage, and represent the active power and reactive power of the photovoltaic inverter, and represent the active power and reactive power of the load, r ij is the resistance of branch ij, and x ijis the branch ij reactance; constraint (33) represents the node active power balance; constraint (34) represents the node reactive power balance; constraint (35) represents the node voltage drop;

[0066] The voltage constraints of the distribution network nodes are established as follows:

[0067]

[0068] Constraint (36) represents the node voltage range limit;

[0069] The power constraints of the distribution network branches are established as follows:

[0070]

[0071] in, Indicates the maximum apparent power allowed by the branch; constraint (37) indicates the power range limit of the branch;

[0072] The PV inverter constraints are established as follows:

[0073]

[0074] in, is the minimum reactive power allowed by the photovoltaic inverter, is the maximum reactive power allowed by the photovoltaic inverter, is the apparent power capacity of the PV inverter; constraint (38) represents the reactive power output limit of the PV inverter; constraint (39) represents that the sum of the squares of the PV active power and the PV inverter reactive power cannot exceed the square of its apparent power capacity; the polygonal approximation method is used to linearize constraints (37) and (39);

[0075] The step droop curve optimization model is established as follows:

[0076]

[0077] Among them, N c is the number of source-load uncertainty scenarios, I is the set of distribution network nodes i, L is the set of branches ij, C is the set of source-load uncertainty scenarios c, For network loss, is the node voltage offset, N i is the total number of distribution network nodes, ω1 and ω2 are weight coefficients of different objectives respectively; Formula (40) indicates that the optimization objective is to minimize power loss and node voltage offset; Formula (41) is used to calculate power loss; Formula (42) is used to calculate node voltage offset;

[0078] The model is solved based on the stochastic optimization method to obtain the power value and inflection point voltage value of each segment of the step droop curve and send them to the charging station.

[0079] Furthermore, in step (4), the charging station decomposes the power of each segment of the step droop curve, and the model is as follows:

[0080]

[0081] in A binary indicator variable for regulating the charging power of electric vehicles, To optimize the power of each section of the step droop curve, The charging power of each charging station is evenly distributed to each electric vehicle according to the minimum segment power of the step droop curve. is the change in charging power of each electric vehicle when the charging station adjusts the power from the minimum segment power upward according to the step droop curve; the optimization target (43) represents the number of electric vehicles that minimize the change in charging power; the constraint (44) represents the charging power distribution of electric vehicles under the 1st to nth segment powers when the charging station adjusts the power in a graded manner according to the step droop curve, where (44a) represents the power decomposition of the minimum segment power of the droop curve, (44b) represents the power decomposition of the second segment power of the droop curve, and so on to the decomposition of the nth segment power; the constraint (45) represents the variable power constraint of the electric vehicle during the step droop curve adjustment process, and the variable power range is between the maximum charging power of the electric vehicle and difference.

[0082] A hierarchical optimization system for active power-voltage step droop curves of a charging station, comprising:

[0083] The charging station-side processing module is configured to construct a single electric vehicle SoC model based on the charging behavior data of electric vehicle users, including the time of entry and exit of the charging station, the initial state of charge (SoC) upon entry, and the desired SoC upon exiting the charging station, and then construct a charging station SoC model after the aggregation of electric vehicles;

[0084] Considering the number of newly arrived electric vehicles and the uncertainty of charging behavior, a charging station SoC interval conditional value-at-risk optimization model is established. The optimization goal is to minimize the charging cost and conditional value-at-risk. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle is calculated, and this power range is reported to the distribution network dispatch center.

[0085] Based on the step droop curve of the current scheduling period obtained from the distribution network processing module, the power of each segment of the droop curve is decomposed to obtain the charging power of each electric vehicle corresponding to the different segment powers, and the charging active power of the electric vehicle is dynamically adjusted according to the real-time voltage measurement value and the power decomposition result;

[0086] The distribution network processing module is configured to consider the uncertainty scenarios of renewable energy output and load in the distribution network at the distribution network system level. Based on the power adjustment range reported by the charging station in each scheduling period, it establishes an active power-voltage step droop curve optimization model for the charging station, solves the optimal step droop curve for each scheduling period, and sends it to the charging station.

[0087] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the above-mentioned layered optimization method for the active power-voltage step droop curve of the charging station is implemented.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] (1) Unlike battery energy storage systems, electric vehicle charging stations are fixed-capacity devices. Their capacity will change with the number of electric vehicles parked at different times. Moreover, charging stations only have charging functions, that is, they can only serve as loads on the distribution network, while battery energy storage can both charge and discharge. Therefore, it is different from the SoC interval optimization of battery energy storage. Under the premise of meeting the charging needs of each electric vehicle owner, the present invention optimizes the SoC interval of the charging station, obtains the adjustable power range of each charging station, and enables it to participate in the voltage regulation of the distribution network. At the same time, considering that electric vehicles are not adapted to frequent power changes, a step-by-step active power-voltage droop curve of the charging station is established to enable it to be adjusted between limited power values. This can improve the voltage quality of the distribution network to a certain extent, reduce power loss, and provide a new technical path for achieving safe and economical operation of the distribution network. The optimization of the SoC interval conditional risk value of the charging station is coordinated with the active power-voltage step droop control, and the parameters of the charging station SoC interval and step droop curve are optimized. Local droop control is then performed to minimize the power loss and voltage offset of the distribution network.

[0090] (2) The output power of traditional droop modeling will change frequently with the change of voltage, which will have an impact on electric vehicles and is not desired by electric vehicle users. The present invention proposes a step droop curve model suitable for electric vehicle charging stations and a corresponding droop curve power decomposition method, which can efficiently adjust the charging power of electric vehicles while avoiding the impact of frequent power adjustments on electric vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0091] Figure 1 It is the overall framework diagram of the method of the present invention;

[0092] Figure 2 This is a 33-node system used in the embodiment of the present invention;

[0093] Figure 3The SoC interval for the entire day obtained by optimizing each charging station in the embodiment of the present invention;

[0094] Figure 4 This is the step droop curve obtained by optimizing each charging station during the period of 10:00-11:00 in the embodiment of the present invention;

[0095] Figure 5 This is the charging power allocated to each electric vehicle by each charging station based on the second segment power decomposition of the step droop curve during the 10:00-11:00 period in the embodiment of the present invention. DETAILED DESCRIPTION

[0096] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0097] This embodiment discloses a method and system for hierarchical optimization of active power-voltage step droop curves of charging stations. The specific framework is as follows: Figure 1 As shown, and using Figure 2 The 33-node system shown implements the method proposed in the present invention. The system is equipped with 8 photovoltaic groups, 6 wind turbines and 4 electric vehicle charging stations. The photovoltaic and wind turbine capacity parameters are shown in Table 1.

[0098] Table 1 PV and wind turbine capacity parameters

[0099]

[0100] The specific steps include:

[0101] Step 1: Based on the charging behavior data of electric vehicle users, including the time of entering the charging station, the time of leaving the charging station, the initial state of charge (SoC) when entering the charging station, and the expected SoC when leaving the charging station, a single electric vehicle SoC model is constructed, and then a charging station SoC model after the electric vehicles are aggregated is constructed. Specifically, it includes:

[0102] (a) Construct the charging power constraint of electric vehicles during parking periods as follows:

[0103]

[0104] Among them, t represents the time period index, m is the car index, i is the distribution network node index, and s is the electric vehicle scene that has just arrived at the charging station. is the charging power of electric vehicle m parked at charging station node i in time period t under scenario s, is the parking period of the car. Constraint (1) indicates that the charging power should meet the power range during the parking period.

[0105] (b) Construct the charging power constraint of electric vehicles during non-parking hours as follows:

[0106]

[0107] Constraint (2) indicates that the charging power of electric vehicles during non-parking periods is 0.

[0108] (c) The parking period intervals are as follows:

[0109]

[0110] in, and are the arrival and departure periods of the car respectively, and formula (3) represents the parking period interval of each electric car. Each car is only dispatched within the parking period.

[0111] (d) The SoC constraints for electric vehicle parking periods are as follows:

[0112]

[0113] in, SoC for electric vehicles, Battery capacity for electric vehicles, SoC max and SoC min They are the maximum SoC and minimum SoC allowed for electric vehicles, represents the charging efficiency of electric vehicles, and Δt is the length of the unit charging scheduling time. Constraint (4) indicates that electric vehicles are charged according to The change process of SoC during charging. Constraint (5) represents the range constraint of the upper bound of SoC of electric vehicles during parking period.

[0114] (e) The SoC constraints for electric vehicles during entry and exit periods are as follows:

[0115]

[0116] in, The initial SoC for electric vehicles, is the SoC that the electric vehicle is expected to achieve when leaving. Constraint (6) means that the initial SoC of the electric vehicle is taken as the entry time t in Constraint (7) requires that the SoC when the electric vehicle leaves the vehicle should be greater than the SoC that the owner originally hoped to achieve.

[0117] (f) The charging station power constraints are as follows:

[0118]

[0119] in, is the power consumed by the charging station, is the overall maximum charging power of the charging station, Mt,i,s is the set of electric vehicles m parked at charging station node i in time period t under scenario s. Constraint (8) indicates that the power consumed by the charging station is the sum of the charging power of all electric vehicles. Constraint (9) indicates that the overall power consumed by the charging station should meet the power limit.

[0120] (g) The cost constraints of charging stations are as follows:

[0121]

[0122] in, Power The corresponding charging cost, Pr t is the charging electricity price in time period t. Constraint (10) calculates the charging cost of the charging station.

[0123] (h) The SoC model of the charging station after electric vehicle aggregation is as follows:

[0124]

[0125] in, is the charging station SoC. Constraint (11) Calculate the charging station SoC for the aggregated electric vehicles based on the SoC and battery rated capacity of each vehicle in the charging station. Here, the numerator is the electric energy of each vehicle obtained by multiplying the SoC of each vehicle by the corresponding battery capacity, and then summing the electric energy of these vehicles; the denominator is the total capacity of the charging station obtained by summing the battery capacity of each vehicle.

[0126] Step 2: Considering the number of newly arrived electric vehicles and the uncertainty of charging behavior, a charging station SoC interval conditional risk value optimization model is established. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle of the charging station is calculated, and the power range is reported to the distribution network dispatch center in turn, including:

[0127] (a) Constructing a risk value optimization model for each charging station SoC interval condition based on the number of electric vehicles:

[0128]

[0129]

[0130] in, is the conditional value at risk of the upper bound of the SoC interval, is the conditional risk value of the lower bound of the SoC interval, T is the set of time period index t, S is the set of electric vehicle scenarios s that have newly arrived at the charging station, N s is the total number of scenes, β1 and β2 are weight coefficients of different targets, is the upper bound of the SoC interval, is the lower bound of the SoC interval, and θ Respectively represent the maximum and minimum values ​​of the SoC interval width, and is an auxiliary variable. Formula (12) indicates that the optimization objective is to minimize the charging cost and the conditional value at risk. Constraint (13) represents the width restriction of the SoC interval. Constraint (14) represents the range restriction of the SoC interval value. Constraints (16)-(17) are the conditional value at risk constraints for the upper bound of the SoC. Constraints (19)-(20) are the conditional value at risk constraints for the lower bound of the SoC.

[0131] (b) Obtain the lower bound of the SoC interval of the charging station through rolling optimization in each scheduling period and upper bound Figure 3 The diagram shows the optimized all-day SoC intervals of four charging stations in an embodiment of the present invention. Based on the SoC intervals, the power adjustable range corresponding to the first scheduling period of the rolling optimization cycle is calculated, and the power range is uploaded to the distribution network dispatch center. The calculation method is:

[0132]

[0133] in, It is the minimum power that the charging station can upload to the distribution network and will serve as the power lower limit of the step droop curve. The maximum power that the charging station uploads to the distribution network will serve as the upper limit of the step droop curve. is the SoC determined by the charging station one hour before. Formula (21) calculates the minimum power of the charging station. Formula (22) calculates the maximum power of the charging station.

[0134] Step 3: At the distribution network system level, considering the uncertainty of renewable energy output and load in the distribution network, an optimization model for the active power-voltage step droop curve of the charging station is established based on the power adjustable range reported by the charging station in each scheduling period. The optimal step droop curve for each scheduling period is obtained and sent to the charging station, including:

[0135] (a) The expression for the step droop curve of electric vehicle charging stations is as follows:

[0136]

[0137] in, is the charging active power of charging station node i in time period t under the uncertainty scenario c of renewable energy output and load in distribution network, V t,i,c is the node voltage value, n is an odd number used to represent the number of segments of the step droop curve, is the voltage value of the inflection point of the step droop curve of the charging station, P 1,2,…,n,t,i is the power value of each segment of the step droop curve of the charging station. Formula (23) represents the mathematical expression of the step droop curve.

[0138] (b) The power constraints for each segment are as follows:

[0139]

[0140] Constraint (24) limits the range of power values ​​of each segment.

[0141] (c) The knee voltage constraint is as follows:

[0142]

[0143] Among them, V and represents the minimum and maximum voltages allowed, respectively. ζ represents the minimum threshold that the length of each segment of the step droop curve must be greater than. Constraint (25) represents the order and constraint range of the inflection point voltages. Constraint (26) indicates that there must be a certain distance between two adjacent inflection point voltage values.

[0144] (d) Reconstruct the step droop curve expression based on the big M method. First, define n binary decision variables with the following constraints:

[0145]

[0146] in, is the binary decision variable of the step droop curve. Constraint (27) ensures that only one section of the droop curve is activated for each charging station in each scenario and each time period by restricting the value of the binary variable. Constraint (28) represents the charging station power.

[0147] (e) Since constraint (28) contains bilinear terms, the charging station power is expressed as follows by introducing auxiliary variables and multiple inequality constraints:

[0148]

[0149] in, is an auxiliary variable. M is a large positive number that must be greater than the maximum charging power of the charging station. Constraint (29) introduces auxiliary variables to represent the charging station power. Constraints (30)-(31) represent the corresponding values ​​of each auxiliary variable under different binary variable values.

[0150] (f) The voltage corresponding to each segment of the step droop curve is expressed through multiple inequality constraints as follows:

[0151]

[0152] Where v is a small positive number, which can be less than one thousandth of the maximum per-unit voltage value, and is used to represent the step of the segment inflection point. Constraint (32) indicates that the node voltage is between the two inflection point voltages of the corresponding segment of the step droop curve under different binary variable values.

[0153] (g) The linear power flow constraint model of the distribution network is established as follows:

[0154]

[0155] Among them, P t,hi,c and Q t,hi,c are the active power and reactive power flowing through branch hi, P t,ij,c and Q t,ij,c are the active power and reactive power flowing through branch ij respectively, V0 is the reference voltage, and They represent the photovoltaic active power generation and photovoltaic inverter reactive power respectively. and Represent the load active power and reactive power, r ij is the resistance of branch ij, x ij is the branch ij reactance. Constraint (33) represents the node active power balance. Constraint (34) represents the node reactive power balance. Constraint (35) represents the node voltage drop.

[0156] (h) Establish the voltage constraints of the distribution network nodes as follows:

[0157]

[0158] Constraint (36) represents the node voltage range limit.

[0159] (i) Establish the power constraints of the distribution network branches as follows:

[0160]

[0161] in, It represents the maximum apparent power allowed in the branch. Constraint (37) represents the power range limit of the branch.

[0162] (j) Establish the PV inverter constraints as follows:

[0163]

[0164] in, is the minimum reactive power allowed by the photovoltaic inverter, is the maximum reactive power allowed by the photovoltaic inverter, is the apparent power capacity of the PV inverter. Constraint (38) represents the reactive power output limit of the PV inverter. Constraint (39) indicates that the sum of the squares of the PV active power and the PV inverter reactive power cannot exceed the square of its apparent power capacity. Constraints (37) and (39) are linearized using polygonal approximation.

[0165] (k) The step droop curve optimization model is established as follows:

[0166]

[0167] Among them, N c is the number of source-load uncertainty scenarios, I is the set of distribution network nodes i, L is the set of branches ij, C is the set of source-load uncertainty scenarios c, For network loss, is the node voltage offset, N i is the total number of distribution network nodes, and ω1 and ω2 are the weight coefficients for different objectives. Equation (40) indicates that the optimization objective is to minimize power loss and node voltage offset. Equation (41) is used to calculate power loss. Equation (42) is used to calculate node voltage offset.

[0168] The model is solved based on the random optimization method to obtain the power value and inflection point voltage value of each segment of the step droop curve and send it to the charging station. Figure 4 shown.

[0169] Step 4: Based on the step droop curve of the current scheduling period, the charging station decomposes the power of each segment of the droop curve to obtain the charging power of each electric vehicle corresponding to the different segment powers, such as Figure 5 As shown, the charging active power of the electric vehicle is dynamically adjusted based on the real-time voltage measurement value and the power decomposition result.

[0170] The power decomposition model is as follows:

[0171]

[0172] in A binary indicator variable for regulating the charging power of electric vehicles, To optimize the power of each section of the step droop curve, The charging power of each charging station is evenly distributed to each electric vehicle according to the minimum segment power of the step droop curve. It is the change in charging power of each electric vehicle when the charging station adjusts the power from the minimum segment power upward in accordance with the step droop curve. The optimization objective (43) represents the number of electric vehicles that minimize the change in charging power. Constraint (44) represents the charging power distribution of electric vehicles under the 1st to nth segment powers when the charging station performs graded adjustment according to the step droop curve, where (44a) represents the power decomposition of the minimum segment power of the droop curve, (44b) represents the power decomposition of the second segment power of the droop curve, and so on to the decomposition of the nth segment power. Constraint (45) represents the variable power constraint of the electric vehicle during the step droop curve adjustment process, and the variable power range is the maximum value of the electric vehicle charging power and difference.

[0173] To verify the effectiveness of the proposed hierarchical optimization method for active power and voltage droop curves at charging stations, 1,000 randomly generated distribution network load and photovoltaic and wind power output scenarios were used to simulate the real-time uncertainty of the dispatch period. A non-step droop control method was used for comparison. This method optimizes and maintains a constant charging power during each dispatch period. The comparison results for the 10:00-11:00 period are shown below:

[0174] Table 2 Comparison results of different methods during the 10:00-11:00 period

[0175]

[0176] During the period of 10:00-11:00, the proposed method achieves the lowest average power loss and average voltage deviation. The results show that the proposed method has significant advantages for the safe and economic operation of the distribution network.

[0177] The present invention also provides a hierarchical optimization system for active power-voltage step droop curves of charging stations, comprising:

[0178] The charging station-side processing module is configured to construct a single electric vehicle SoC model based on the charging behavior data of electric vehicle users, including the time of entry and exit of the charging station, the initial state of charge (SoC) upon entry, and the desired SoC upon exiting the charging station, and then construct a charging station SoC model after the aggregation of electric vehicles;

[0179] Considering the number of newly arrived electric vehicles and the uncertainty of charging behavior, a charging station SoC interval conditional value-at-risk optimization model is established. The optimization goal is to minimize the charging cost and conditional value-at-risk. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle is calculated, and this power range is reported to the distribution network dispatch center.

[0180] Based on the step droop curve of the current scheduling period obtained from the distribution network processing module, the power of each segment of the droop curve is decomposed to obtain the charging power of each electric vehicle corresponding to the different segment powers, and the charging active power of the electric vehicle is dynamically adjusted according to the real-time voltage measurement value and the power decomposition result;

[0181] The distribution network processing module is configured to consider the uncertainty scenarios of renewable energy output and load in the distribution network at the distribution network system level. Based on the power adjustment range reported by the charging station in each scheduling period, it establishes an active power-voltage step droop curve optimization model for the charging station, solves the optimal step droop curve for each scheduling period, and sends it to the charging station.

[0182] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the above-mentioned layered optimization method for the active power-voltage step droop curve of the charging station is implemented.

[0183] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, apparatus (systems), computer devices, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0184] The present invention is described with reference to flowcharts of methods according to embodiments of the present invention. It should be understood that each process in the flowcharts and combinations of processes in the flowcharts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts. Figure 1 A device that specifies functions in a process or multiple processes.

[0185] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A function specified in a process or multiple processes.

[0186] These computer program instructions can also be loaded into a computer or other programmable data processing devices, so that a series of operational steps are performed on the computer or other programmable devices to generate a computer-implemented process, so that the instructions executed by the computer or other programmable devices provide operational steps for implementing the functions specified in the flowchart or multiple flowcharts. Figure 1 The flowchart and the associated discussion are just one embodiment of a method, apparatus, or manufacturing and processing implementation for performing the functions specified in the flowchart. The computer program in particular, and each line of code of the computer program in particular, can represent a step in the process and / or be implemented by hardware that executes the instructions or code in a manner based in logic to perform the steps.

Claims

1. A hierarchical optimization method for active power-voltage step droop curve of a charging station, characterized in that: The following steps are involved: (1) Based on the charging behavior data of electric vehicle users, including the time of entering the charging station, the time of leaving the charging station, the initial state of charge (SoC) when entering the charging station, and the expected SoC when leaving the charging station, a single electric vehicle SoC model is constructed, and then a charging station SoC model after the electric vehicles are aggregated is constructed; (2) Considering the number of electric vehicles newly arriving at the charging station and the uncertainty of charging behavior, a charging station SoC interval conditional risk value optimization model is established. The optimization goal is to minimize the charging cost and conditional risk value. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle of the charging station is calculated, and the power range is reported to the distribution network dispatching center. (3) At the distribution network system level, considering the uncertainty of renewable energy output and load in the distribution network, based on the power adjustable range reported by the charging station in each scheduling period, an optimization model of the active power-voltage step droop curve of the charging station is established, and the optimal step droop curve of each scheduling period is obtained and sent to the charging station; (4) Based on the step droop curve of the current scheduling period, the charging station decomposes the power of each segment of the droop curve to obtain the charging power of each electric vehicle corresponding to the different segment powers, and dynamically adjusts the charging active power of the electric vehicle based on the real-time voltage measurement value and the power decomposition result.

2. The method according to claim 1, characterized in that The SoC model of a single electric vehicle in step (1) includes: (a) The charging power of electric vehicles during parking periods should meet the power range: Among them, t represents the time period index, m is the car index, i is the distribution network node index, and s is the electric vehicle scene that has just arrived at the charging station. is the charging power of electric vehicle m parked at charging station node i in time period t under scenario s, is the maximum charging power allowed for electric vehicles, It is the parking period of the car; (b) The charging power of electric vehicles during non-parking hours is 0: (c) Each electric vehicle is dispatched only during the parking period, which is as follows: in, and are the arrival and departure time periods of the car respectively; (d) The SoC constraints for electric vehicle parking periods are as follows: in, SoC for electric vehicles, Battery capacity for electric vehicles, SoC max and SoC min They are the maximum SoC and minimum SoC allowed for electric vehicles, represents the charging efficiency of electric vehicles, Δt is the length of the unit charging scheduling time; constraint (4) represents the charging efficiency of electric vehicles during the parking period according to The change process of SoC during charging; Constraint (5) represents the range constraint of the upper bound of SoC of electric vehicles during parking period; (e) The SoC constraints for electric vehicles during entry and exit periods are as follows: in, The initial SoC for electric vehicles, is the SoC that the electric vehicle expects to achieve when leaving; constraint (6) means taking the initial SoC of the electric vehicle as the entry time t in Constraint (7) requires that the SoC when the electric vehicle leaves the vehicle should be greater than the SoC that the owner originally hoped to achieve.

3. The method according to claim 2, characterized in that In step (1), the charging station SoC model after electric vehicle aggregation includes: (f) The charging station power constraints are as follows: in, is the power consumed by the charging station, is the overall maximum charging power of the charging station, M t,i,s is the set of electric vehicles m parked at the charging station at node i in time period t under scenario s; constraint (8) indicates that the power consumed by the charging station is the sum of the charging power of all electric vehicles; constraint (9) indicates that the overall power consumed by the charging station should meet the power limit; (g) The cost constraints of charging stations are as follows: in, Power The corresponding charging cost, Pr t is the charging electricity price in time period t; (h) The SoC model of the charging station after electric vehicle aggregation is as follows: in, is the charging station SoC; constraint (11) calculates the charging station SoC of the electric vehicles after aggregation based on the SoC and battery rated capacity of each vehicle in the charging station.

4. The method according to claim 3, characterized in that In step (2), the charging station SoC interval conditional risk value optimization model is as follows: st(1)-(11) in, is the conditional value at risk of the upper bound of the SoC interval, is the conditional risk value of the lower bound of the SoC interval, T is the set of time period indexes, S is the set of electric vehicle scenarios s that have newly arrived at the charging station, and N s is the total number of scenes, β1 and β2 are weight coefficients of different targets, is the upper bound of the SoC interval, is the lower bound of the SoC interval, and θ Respectively represent the maximum and minimum values ​​of the SoC interval width, and is an auxiliary variable; formula (12) indicates that the optimization objective is to minimize the charging cost and the conditional value at risk; constraint (13) indicates the width restriction of the SoC interval; constraint (14) indicates the range restriction of the SoC interval value; constraints (15)-(17) are the conditional value at risk constraints of the upper bound of SoC; constraints (18)-(20) are the conditional value at risk constraints of the lower bound of SoC.

5. The method according to claim 4, characterized in that In step (2), the SoC interval of the charging station is obtained through rolling optimization in each scheduling period, and the power adjustable range corresponding to the first period of the rolling optimization cycle of the charging station is calculated based on the SoC interval. The calculation method is: in, It is the minimum power that the charging station can upload to the distribution network and will serve as the power lower limit of the step droop curve. The maximum power that the charging station uploads to the distribution network will serve as the upper limit of the step droop curve. is the SoC determined one hour before the charging station; formula (21) calculates the minimum power of the charging station; formula (22) calculates the maximum power of the charging station.

6. The method according to claim 5, characterized in that In step (3), based on the power adjustable range reported by the charging station in each scheduling period, an optimization model of the active power-voltage step droop curve of the charging station is established, including: (a) The expression for the step droop curve of electric vehicle charging stations is as follows: in, is the charging active power of charging station node i in time period t under the uncertainty scenario c of renewable energy output and load of distribution network, V t,i,c is the node voltage value, n is an odd number used to represent the number of segments of the step droop curve, is the voltage value of the inflection point of the step droop curve of the charging station, P 1,2,…,n,t,i is the power value of each segment of the step droop curve of the charging station; (b) The power constraints for each segment are as follows: Constraint (24) limits the value range of each segment power; (c) The knee voltage constraint is as follows: Among them, V and They represent the minimum and maximum voltages allowed, respectively; ζ represents the minimum threshold that the length of each segment of the step droop curve must be greater than; constraint (25) represents the order of magnitude and constraint range of each inflection point voltage; constraint (26) indicates that there must be a certain distance between two adjacent inflection point voltage values.

7. The method according to claim 6, characterized in that In step (3), the optimal step droop curve for each scheduling period is obtained and sent to the charging station, including: The step droop curve expression is reconstructed based on the big M method. First, n binary decision variables are defined, and the constraints are as follows: in, is the binary decision variable of the step droop curve; constraint (27) ensures that only a certain section of the droop curve is activated in each charging station in each scenario and each time period by limiting the value of the binary variable; constraint (28) represents the power of the charging station; Since constraint (28) contains bilinear terms, the charging station power is expressed as follows by introducing auxiliary variables and multiple inequality constraints: … … in, is an auxiliary variable; M is a specified positive number, whose value is greater than the maximum charging power of the charging station; constraint (29) represents the charging station power by introducing auxiliary variables; constraints (30)-(31) represent the corresponding values ​​of each auxiliary variable under different binary variable values; The voltage corresponding to each segment of the step droop curve is expressed through multiple inequality constraints as follows: … Wherein, v is a specified positive number, which is less than one thousandth of the maximum per-unit voltage value, and is used to represent the step of the segment inflection point; constraint (32) indicates that the node voltage under different binary variable values ​​is between the two inflection point voltages of the corresponding segment of the step droop curve; The linear power flow constraint model of the distribution network is established as follows: Among them, P t,hi,c and Q t,hi,c are the active power and reactive power flowing through branch hi, P t,ij,c and Q t,ij,c are the active power and reactive power flowing through branch ij respectively, V0 is the reference voltage, and They represent the photovoltaic active power generation and photovoltaic inverter reactive power respectively. and Represent the load active power and reactive power, r ij is the resistance of branch ij, x ij is the branch ij reactance; constraint (33) represents the node active power balance; constraint (34) represents the node reactive power balance; constraint (35) represents the node voltage drop; The voltage constraints of the distribution network nodes are established as follows: Constraint (36) represents the node voltage range limit; The power constraints of the distribution network branches are established as follows: in, Indicates the maximum apparent power allowed by the branch; constraint (37) indicates the power range limit of the branch; The PV inverter constraints are established as follows: in, is the minimum reactive power allowed by the photovoltaic inverter, is the maximum reactive power allowed by the photovoltaic inverter, is the apparent power capacity of the PV inverter; constraint (38) represents the reactive power output limit of the PV inverter; constraint (39) represents that the sum of the squares of the PV active power and the PV inverter reactive power cannot exceed the square of its apparent power capacity; the polygonal approximation method is used to linearize constraints (37) and (39); The step droop curve optimization model is established as follows: st(24)-(27),(29)-(39) Among them, N c is the number of source-load uncertainty scenarios, I is the set of distribution network nodes i, L is the set of branches ij, C is the set of source-load uncertainty scenarios c, For network loss, is the node voltage offset, N i is the total number of distribution network nodes, ω1 and ω2 are weight coefficients of different objectives respectively; Formula (40) indicates that the optimization objective is to minimize power loss and node voltage offset; Formula (41) is used to calculate power loss; Formula (42) is used to calculate node voltage offset; The model is solved based on the stochastic optimization method to obtain the power value and inflection point voltage value of each segment of the step droop curve and send them to the charging station.

8. The method according to claim 7, characterized in that In step (4), the charging station decomposes the power of each segment of the step droop curve, and the model is as follows: st … … in A binary indicator variable for regulating the charging power of electric vehicles, To optimize the power of each section of the step droop curve, The charging power of each charging station is evenly distributed to each electric vehicle according to the minimum segment power of the step droop curve. is the change in charging power of each electric vehicle when the charging station adjusts the power from the minimum segment power upward according to the step droop curve; the optimization target (43) represents the number of electric vehicles that minimize the change in charging power; the constraint (44) represents the charging power distribution of electric vehicles under the 1st to nth segment powers when the charging station adjusts the power in a graded manner according to the step droop curve, where (44a) represents the power decomposition of the minimum segment power of the droop curve, (44b) represents the power decomposition of the second segment power of the droop curve, and so on to the decomposition of the nth segment power; the constraint (45) represents the variable power constraint of the electric vehicle during the step droop curve adjustment process, and the variable power range is between the maximum charging power of the electric vehicle and difference.

9. A hierarchical optimization system for active power-voltage step droop curves of charging stations, characterized in that: include: The charging station-side processing module is configured to construct a single electric vehicle SoC model based on the charging behavior data of electric vehicle users, including the time of entry and exit of the charging station, the initial state of charge (SoC) upon entry, and the desired SoC upon exiting the charging station, and then construct a charging station SoC model after the aggregation of electric vehicles; Considering the number of newly arrived electric vehicles and the uncertainty of charging behavior, a charging station SoC interval conditional value-at-risk optimization model is established. The optimization goal is to minimize the charging cost and conditional value-at-risk. The SoC interval of the charging station is obtained through rolling optimization in each scheduling period. Based on the SoC interval, the power adjustable range corresponding to the first period of the rolling optimization cycle is calculated, and this power range is reported to the distribution network dispatch center. Based on the step droop curve of the current scheduling period obtained from the distribution network processing module, the power of each segment of the droop curve is decomposed to obtain the charging power of each electric vehicle corresponding to the different segment powers, and the charging active power of the electric vehicle is dynamically adjusted according to the real-time voltage measurement value and the power decomposition result; The distribution network processing module is configured to consider the uncertainty scenarios of renewable energy output and load in the distribution network at the distribution network system level. Based on the power adjustment range reported by the charging station in each scheduling period, it establishes an active power-voltage step droop curve optimization model for the charging station, solves the optimal step droop curve for each scheduling period, and sends it to the charging station.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for hierarchical optimization of active power-voltage step droop curve of a charging station according to any one of claims 1 to 8 is implemented.

Images