An electric vehicle charging pricing strategy considering user characteristics and network constraints

Abstract

The present application relates to the field of electric vehicle charging, and discloses an electric vehicle charging pricing strategy that takes into account user characteristics and network constraints, including the following steps: S1. Receive energy consumption constraint information and network operation constraint information related to each user's electric vehicle; S2. Based on the fast dual decomposition algorithm, establish a sequential Stackelberg game model between the aggregator and the distribution network operator; S3. In the game model, the charging electricity price of the electric vehicle is comprehensively set according to the network constraints of the distribution network and the user's own preference characteristics. The present invention effectively coordinates the interaction between the distribution network operator and the aggregator by introducing the Stackelberg sequential game model. This model not only promotes the cooperation between the two parties on the electric vehicle charging strategy, but also realizes the optimal distribution of the power grid load, reduces the pressure of the power grid during peak hours, and thus improves the operation efficiency and reliability of the entire power grid.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging, and in particular to an electric vehicle charging pricing strategy that takes user characteristics and network constraints into consideration. Background Art

[0002] As a green means of transportation to replace traditional fuel vehicles, the popularity of electric vehicles continues to increase. The huge charging demand and disorderly charging behavior brought about by this growth have posed new challenges to the stable operation of the distribution network. Especially during peak hours, unoptimized electric vehicle charging activities may cause a series of problems such as voltage fluctuations and line congestion in the power grid. Therefore, formulating a scientific and comprehensive electric vehicle charging price strategy is of great significance to maintaining the stable operation of the distribution network and improving the overall social benefits.

[0003] At present, the electricity pricing strategies of distribution networks, such as the distribution marginal price (DLMP)-based methods, take into account the power flow and voltage constraints of the power grid, but often ignore the behavioral characteristics and preference differences of electric vehicle users, as well as the potential impact of these factors on electricity pricing strategies. In addition, although there are some methods that allow electric vehicle users to interact with distribution network operators through bidding mechanisms, these methods usually involve multiple rounds of communication, which not only has high communication costs, but also lacks sufficient physical explanations, limiting their practical application scope and efficiency.

[0004] The deficiencies of existing technologies are mainly manifested in three aspects: first, many electricity price setting mechanisms cannot fully deal with the various constraints in the distribution network, such as voltage, peak load and line flow constraints; second, the communication mechanism relied on in the implementation of existing interactive energy systems is inefficient and lacks the necessary physical foundation, making it difficult for the strategy to be widely used in practice; finally, the existing electricity price design fails to fully consider the differences in individual user preferences and reactions, and thus fails to effectively utilize the potential flexibility of electric vehicle users.

[0005] To solve these problems, the present invention proposes a new electric vehicle charging pricing strategy. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention provides an electric vehicle charging pricing strategy that takes into account user characteristics and network constraints. The strategy is based on the fast dual decomposition algorithm and constructs a sequential Stackelberg game model between aggregators and distribution network operators. Through this model, not only the various physical and operational constraints of the distribution network are considered, but also the user's preferences and behavioral characteristics are integrated, so that users are guided to charge in an orderly manner by dynamically adjusting the electricity price, which not only ensures the stable operation of the power grid, but also maximizes social benefits.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: an electric vehicle charging pricing strategy considering user characteristics and network constraints, comprising the following steps:

[0008] Receive energy consumption constraint information and network operation constraint information related to each user's electric vehicle;

[0009] Based on the fast dual decomposition algorithm, a sequential Stackelberg game model is established between aggregators and distribution network operators.

[0010] In the game model, the charging electricity price of electric vehicles is set comprehensively according to the network constraints of the distribution network and the user's own preference characteristics.

[0011] Preferably, the user characteristics include user preference settings.

[0012] Preferably, the operational status data of the power grid includes predicted load flow data, voltage level and line load conditions, wherein the aggregator reports the energy consumption plan for a period of time in the future to the distribution network and operates according to the plan.

[0013] Preferably, the fast dual decomposition algorithm comprises:

[0014] Modeling the grid status and user demand in each decision cycle;

[0015] Using dual variables to solve voltage and line constraints in power grids;

[0016] Iteratively update electricity prices and charging plans until convergence.

[0017] Preferably, the iterative updating includes processing power grid constraints using Lagrangian relaxation techniques and updating dual variables.

[0018] Preferably, the sequential Stackelberg game model includes:

[0019] The distribution network operator takes the lead in setting the initial electricity price;

[0020] Aggregators act as followers and adjust their EV charging strategies based on the electricity prices received.

[0021] Preferably, the charging strategy adjusted by the aggregator includes managing the charging rate of the electric vehicle.

[0022] Preferably, the game model adjusts electricity prices to guide electricity users to charge in an orderly manner, and takes into account the charging flexibility of users and voltage fluctuations, line congestion and peak load constraints of the distribution network.

[0023] The present invention also provides an electric vehicle charging pricing system that takes into account user characteristics and network constraints, including:

[0024] a data receiving module configured to receive energy consumption constraints of power users and charging requirements of electric vehicles, as well as real-time power flow and voltage constraint data of the distribution network;

[0025] A preliminary electricity price setting module, configured to set a preliminary electricity price based on the operation status of the power grid according to the operation status data of the power distribution network;

[0026] Optimized processing modules, including:

[0027] a fast dual decomposition algorithm unit configured to apply a fast dual decomposition algorithm to process an electricity price and charging schedule optimization problem between a distribution network and an aggregator;

[0028] a sequential Stackelberg game model unit configured to optimize electricity prices and charging plans through a sequential Stackelberg game model;

[0029] an electricity price adjustment module, configured to adjust the electricity price according to the optimization result to guide the electricity users to charge in an orderly manner, and taking into account the charging flexibility of the users and the voltage fluctuation, line congestion and peak load constraint of the distribution network;

[0030] The control output module is configured to output optimized electricity price signals and charging plans to relevant electricity market participants and aggregators.

[0031] The present invention also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the above method is implemented.

[0032] The present invention also provides a storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described above is implemented.

[0033] The present invention provides an electric vehicle charging pricing strategy that takes user characteristics and network constraints into consideration.

[0034] It has the following beneficial effects:

[0035] 1. The present invention effectively coordinates the interaction between distribution network operators and aggregators by introducing the Stackelberg sequential game model. This model not only promotes cooperation between the two parties on electric vehicle charging strategies, but also achieves optimal distribution of grid loads, reduces the pressure of the grid during peak hours, and thus improves the operating efficiency and reliability of the entire grid.

[0036] 2. The present invention significantly improves computational efficiency by using the fast dual decomposition algorithm for distributed optimization. This algorithm allows distribution network operators to quickly respond to market changes and demand fluctuations when facing large-scale networks and a large number of users, achieve real-time or quasi-real-time optimization decisions, optimize resource allocation, and reduce energy loss.

[0037] 3. This invention enhances the transparency and traceability of power grid dispatch strategies by integrating physically interpretable mathematical models. This model not only ensures the mathematical rigor of the algorithm, but also facilitates regulators and grid operators to understand and verify the physical and economic principles behind the algorithmic decisions, which helps to enhance the trust of users and the public in the power grid management strategy.

[0038] 4. The present invention ensures the stability of the market and the fairness of electricity prices by achieving Stackelberg equilibrium. The implementation of the strategy helps to form a fair market environment, in which electricity prices reflect the real supply and demand relationship and the economic impact of user behavior, thereby encouraging users to adopt more reasonable electricity usage habits and promote energy consumption towards a more efficient and environmentally friendly direction. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A diagram of the overall price design framework for electric vehicle charging according to the present invention;

[0040] Figure 2 The active and reactive power curves of all non-electric vehicles collected in the distribution network in an embodiment of the present invention are as follows; wherein: (a) non-electric vehicle power curve; (b) initial retail price curve;

[0041] Figure 3 It is a schematic diagram of the maximum and minimum line loads of the distribution network in all k∈K time periods under the condition of no management in the embodiment of the present invention;

[0042] Figure 4 This is a schematic diagram of three-phase voltage levels of a power distribution network in the period k=13 under unmanaged conditions in an embodiment of the present invention;

[0043] Figure 5 Schematic diagram of the maximum and minimum voltage levels of the distribution network in all k∈K time periods in an embodiment of the present invention;

[0044] Figure 6 is the a-phase child node of node 113 in the embodiment of the present invention. Convergence diagram;

[0045] Figure 7 Schematic diagram of the maximum line load of the distribution network in all k∈K time periods under the sequential Stackelberg model in an embodiment of the present invention;

[0046] Figure 8It is a schematic diagram of comparison between the distribution network retail price result and the initial retail price of the 123 nodes when k=13 in an embodiment of the present invention;

[0047] Fig. 9 It is a schematic diagram of the system structure of the present invention;

[0048] Fig.10 It is a schematic diagram of the computer device structure of the present invention.

[0049] Among them, 100, data receiving module; 200, preliminary electricity price setting module; 300, optimization processing module; 400, electricity price adjustment module; 500, control output module; 40, computer equipment; 41, processor; 42, memory; 43, storage medium. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] Please refer to the attached Figure 1 -Attached Figure 8 , an embodiment of the present invention provides an electric vehicle charging pricing strategy that takes into account user characteristics and network constraints, including the following steps:

[0052] S1. Receive energy consumption constraint information and network operation constraint information related to each user's electric vehicle

[0053] Information collection: Distribution network operators receive detailed energy constraints information about electricity users from aggregators at each node, including the maximum charging requirements and time preferences of electric vehicles. In addition, constraint information related to network operations is collected, such as current grid load conditions, line capacity, voltage levels, etc.

[0054] Aggregator role: In this embodiment, it is assumed that there is an aggregator at each node of the distribution network. These aggregators are responsible for managing all electricity users connected to the node and ensuring that the collected data is accurate and real-time.

[0055] S2. Based on the fast dual decomposition algorithm, a sequential Stackelberg game model is established between the aggregator and the distribution network operator;

[0056] Model establishment: The distribution network operator first establishes a sequential Stackelberg game model based on the collected information using the fast dual decomposition algorithm. In this model, the distribution network operator acts as a leader, sets the initial retail electricity price, and sends it to the aggregator.

[0057] Aggregator response: Aggregators, as followers, optimize the charging and discharging plans of EVs based on the received electricity prices and the needs of users at their respective nodes and network constraints.

[0058] S3. In the game model, the charging price of electric vehicles is set comprehensively based on the network constraints of the distribution network and the user's own preference characteristics.

[0059] Electricity price setting: Under the game model framework, electricity prices are adjusted through multiple rounds of iterations until a balance point is found, which not only meets the safe and stable operation of the distribution network, but also meets the electricity demand of users and optimizes social benefits. The setting of electricity prices takes into account the real-time status of the power grid (such as voltage fluctuations, line congestion and peak load) and the charging flexibility of users.

[0060] Iterative optimization: Distribution network operators continuously adjust electricity prices based on charging plans fed back by aggregators and real-time grid status checks until all network constraints and user needs are met, minimizing costs and maximizing efficiency.

[0061] The present invention effectively manages the charging needs of electric vehicles and the operation needs of the power grid by integrating the resources of power grid operators and aggregators. By adjusting the electricity price in real time to cope with changes in the power grid state, the load management capability and operation efficiency of the power grid are improved. At the same time, the electricity price is flexibly adjusted according to user characteristics and power grid demand, helping users reduce charging costs and improve economic benefits. In addition, by optimizing the charging plan, the burden on the power grid is reduced, the power grid overload is avoided during peak hours, and the stability and reliability of the system are enhanced.

[0062] In a preferred embodiment of the present invention, a network model is proposed, and the model is designed as follows:

[0063] The scenario is set as a three-phase unbalanced distribution network with N+1 nodes. Represents a node set, 0 represents the first node, Represents the set of all non-first nodes. The distribution network has N unconnected lines connecting adjacent nodes. For each node make represents the node before node j. represents the set of all nodes strictly after node j. Represents a collection of distribution circuits.

[0064] For all decision sub-cycles And each line The unbalanced linearized power flow can be written as:

[0065]

[0066]

[0067]

[0068] in,

[0069] R ij for l j = (i, j) three-phase resistance matrix (pu)

[0070] X i for l j = (i, j) three-phase reactance matrix (pu)

[0071] a=[1,e -j2π / 3 ,e j2π / 3 ] T , a H is the conjugate transposed matrix of a

[0072]

[0073]

[0074] ⊙ indicates corresponding multiplication of elements in the same position

[0075] (1) In the formula, the 3×1 column vector Where φ={a, b, c} three phases, representing the line l j The three-phase active power flow, three-phase reactive power flow, square of the three-phase voltage value of node j, and three-phase active and reactive loads of node j. All quantities are in (pu) and are arranged in the phase sequence of {a, b, c}.

[0076] At the same time, define the following vectors: v(k)=[v j (k)] j∈N , p(k)=[p j (k)] j∈N , q(k)=[q j (k)] j∈N .for The power flow (1) can be written as follows:

[0077] AP(k)=-p(k); AQ(k)=-q(k) (2a)

[0078]

[0079] in,

[0080] Represents the standard incidence matrix of the distribution network, is a block diagonal matrix. v(k) can be expressed as follows:

[0081] v(k)=-[A T ] -1 A0v0(k)-2R D p(k)-2X D q(k) (3a)

[0082] R d =[A T ] -1 D r a -1 ,X D =[A T ] -1 D x A -1 (3b)

[0083] Through the above network model, the present invention can effectively simulate and optimize the power flow in the power grid, especially when considering electric vehicles as an emerging load source. This model not only improves the operation efficiency of the power grid, but also helps to achieve more accurate charging pricing strategies, thereby improving user satisfaction and network reliability.

[0084] In a preferred embodiment of the present invention, an aggregator model is proposed, which focuses on optimizing the charging strategy of electric vehicle users at the node to improve the operation efficiency of the power grid and meet the charging needs of users. The specific implementation of the model is as follows:

[0085] Let ψ=(u,φ,j) represent the user identity with preference and structural attributes, where φ∈{a,b,c} represents the phase of its connection, j represents the location of the user connection, u represents the user's personal preference and structural characteristics, and for each user ψ is composed of electric vehicles and other electrical equipment. Here, the electric vehicles of ψ are considered to be price-sensitive flexible distribution resources.

[0086] Each node j has an aggregator responsible for managing the power usage of all users on the node. j φ represents the user connected to node j, φ phase; Ψ j Represents all users connected to node j, with Ψ j =Ψ j a ∪Ψ j b ∪Ψ j c ; Ψ represents all users on the distribution network.

[0087] Let pΨ (k)(pu) and q Ψ (k)(pu) represents the active and reactive power of the electric vehicle of user ψ in time period k∈K; and represents the active and reactive power of the non-electric vehicle electrical appliances of user Ψ in time period k∈K. It is also defined as follows:

[0088] P ψ (K) = [p ψ (1),…,p ψ (NK)] T ,Q ψ (K) = [q ψ (1),…,q ψ (NK)] T

[0089]

[0090] set up represents the state of charge of the electric vehicle of user ψ at the beginning of period K, SOC ψ (K) represents the state of charge of the electric vehicle of user ψ at the end of time period k∈K, represents the state of charge of the electric vehicle required at the end of the K period (the charging target of the K period). j , the energy required to charge the electric vehicle of user ψ should be equal to the energy required at the end of charging period K. Equality, this constraint can be expressed as:

[0091]

[0092]

[0093] Among them, S base (kVA) is the reference power used to convert between unit (pu) and power (kVA) form; E ψ (kWh) is the electric vehicle capacity of user ψ; △t is the unit time of each sub-period k (such as 0.25h, 1.0h, 1.5h); α ψ (%): Charging efficiency of the electric vehicle of user ψ. SOC ψ (K) are defined as follows:

[0094] SOC ψ (K) = [SOC ψ (1),…,SOC ψ (NK)] T

[0095]

[0096]

[0097] The goal of the aggregator is to obtain the maximum possible net benefit (i.e., utility minus cost) during period K by adjusting the EV charging plans of its managed customers during period K, where the net benefit is generally:

[0098] NetBen ψ =Utility ψ -μ ψ Cost ψ (5)

[0099] Among them, μ ψ (utils / cent) is the marginal cost of user ψ, which can realize the unit conversion from cent to utils. ψ It reflects the charging satisfaction of user ψ in period K, Cost ψ represents the cost incurred by user ψ for its power usage, represents the maximum satisfaction of user ψ in time period K, and the dissatisfaction of user ψ is given by Measure, m ψ (utils / cent) is a positive conversion coefficient. The utility function represents the total satisfaction of user ψ under the decision SOCψ(K), and its analytical form is as follows:

[0100]

[0101] Formula (6) shows that SOC ψ (k) The closer The higher the utility obtained by user ψ in time period k, the higher the utility obtained by user ψ in time period k.

[0102] SOC ψ (K) can be expressed as:

[0103]

[0104] in,

[0105]

[0106] Substituting (7) into (6), U ψ (SOC ψ (K)) can use U ψ (P ψ (K))Alternative:

[0107]

[0108] in,

[0109]

[0110]

[0111]

[0112] Cost ψ The function represents the cost paid by user ψ in period K due to the electric vehicle charging plan. Let π ψ (K) = [π ψ (1),…,π ψ (K)] represents the retail price curve of user ψ in period K. The cost analysis form is as follows:

[0113] Cost ψ =π ψ (K)P ψ (K)S base Δt (10)

[0114] Define X as follows ψ (K), X j (K) and P j (K):

[0115] X ψ (K) = {P ψ (K)∈R K ∣P ψ (K) satisfies (4)} (11a)

[0116]

[0117] P j (K) = {P ψ (K)|ψ∈Ψ j} (11c)

[0118] At this point, the mathematical expression for the maximum network benefit problem for each aggregator in the network is as follows:

[0119]

[0120] It is worth mentioning that the optimal solution of (12) depends on the price curve π received by the aggregator j (K), and there is

[0121] π j (K) = {π ψ (K)|ψ∈Ψ j} (13)

[0122] Let Pj(π j (K)) is given by π j (K) is the optimal solution of (12), then:

[0123]

[0124] Proposition 1: The benefit maximization problem in (12) is a strict concave optimization problem, where M ψ is a real symmetric negative definite matrix, indicating that for a given π l (K), P j (π j (K)) is unique.

[0125] Proof of Proposition 1: (12) About P ψ The second-order partial derivative of (K) is as follows:

[0126]

[0127] For any non-zero vector z∈R 3N ,have:

[0128] z T M ψ z=-m ψ [J ψ z[ T J ψ z

[0129] From (8), we can see that J W is positive definite and reversible, assuming that J ψ z=0, then This contradicts the fact that z is a non-zero vector, so J ψ z≠0, so we get [J ψ z] T J ψ z>0. ψ >0, we know z T M ψ z<0, so we can deduce M ψ is a real symmetric negative definite matrix, so the objective function (12) is a strictly concave function. In addition, from (4), we can see that the constraint X i is convex. Therefore, the maximum network benefit problem (12) of the aggregator is a strictly concave optimization problem, so for any given π i (K), Pj(π j (K)) is unique, proof complete.

[0130] Through this aggregator model, the present invention not only optimizes the load management of the power grid, but also improves the charging experience and satisfaction of users, while further enhancing the economic efficiency and environmental sustainability of the power grid through accurate demand forecasting and electricity price response. This model is a key tool for achieving efficient electric vehicle charging management, especially for urban or regional power grids with large-scale electric vehicle access.

[0131] In a preferred embodiment of the present invention, a centralized optimization model for distribution network operators is proposed, which focuses on optimizing the electric vehicle charging strategy of the entire power grid to maximize the grid operation efficiency and meet the charging needs of a wide range of users. The specific implementation of the model is as follows:

[0132] The goal of the distribution network operator is to maximize the network benefits of all aggregators and users by adjusting the EV charging plan of all users in period K, subject to local user constraints and distribution network constraints. Let P(K) and X(K) represent the set of EV charging curves and all local user constraints, respectively:

[0133] P(K)={P j (K)|j∈N}(17a)

[0134] X(K)=∏ j∈N X j (K)(17b)

[0135] Then, the maximum network benefit problem of the distribution network operator can be expressed as follows:

[0136]

[0137] Subject to (2)(18b)

[0138]

[0139]

[0140]

[0141]

[0142]

[0143]

[0144] In (18), π0(K) = [π0(1),…,π0(K)] represents the initial retail price, which is also the price at which the distribution network operator purchases electricity from the regional transmission operator / independent system operator or utility company, (18a) is the objective function that maximizes the benefits of all aggregators, and (18b) is the unbalanced distribution power flow constraint. Three different network-level constraints are also considered, (18c) (18e) are peak demand constraints, voltage constraints, and line flow constraints, respectively. Among them, and represents the active and reactive loads of the φ phase on node j. (18f) and (18g) are the active and reactive connections between the node and the user, and (18h) is the active and reactive relationship of the user ψ electric vehicle, where the power factor of the electric vehicle of ψ remains constant during the k∈K period.

[0145] Due to user privacy, U ψ (P ψ (K)) and X(K) cannot be obtained directly, so the optimal problem (18) cannot be solved directly by the distribution network operator. The operator can resort to the sequential Stackelberg model to use the retail price curve to induce the aggregator to change the EV charging plan, thereby achieving the same optimal solution as the optimal problem (23).

[0146] Through the centralized optimization model, the present invention effectively integrates the resources of grid operators and aggregators, optimizes the load management of the grid, and enhances the economic efficiency and environmental sustainability of the grid through accurate demand forecasting and electricity price response. This model provides a key tool for achieving efficient and sustainable electric vehicle charging management, especially for urban or regional grids facing the challenge of large-scale electric vehicle access.

[0147] In a preferred embodiment of the present invention, a centralized optimization model for distribution network operators is proposed, which is combined with the operator-aggregator sequential Stackelberg model to achieve efficient management and optimization of electric vehicle charging pricing. The specific implementation of the model is as follows:

[0148] 1) Equivalent statement of the optimal problem of distribution network operators

[0149] In order to facilitate the application of the operator-aggregator sequential Stackelberg model, the distribution network operator optimization problem is first reformulated. The objective function (18a), constraints (18f) to (18h), and X(K) can all be decomposed to the level of a single aggregator, but (18b) to (18e) are coupling constraints between various aggregators.

[0150] Using (2) and (3), v(k) and P(k) can be reformulated as follows:

[0151]

[0152]

[0153] in,

[0154] v nin (k)=-[A T ] -1 A0v0(k)(20a)

[0155]

[0156]

[0157]

[0158] or 3 if φ=a,b,orc(20e)

[0159] First, F ψ (P(K)) is defined as follows:

[0160] F ψ (P ψ (K))=U ψ (P ψ (K))-μ ψ π0(K)P ψ (K)S base Δt (21)

[0161] Similarly, y j (K),y(K),Y j (K) and Y(K) are defined as follows:

[0162]

[0163] y(K)={y j (K)|j∈N} (22b)

[0164] Y j (K) = {y j (K)|y j (K) satisfies (18f) to (18h) and X j (K)} (22c)

[0165] Y(K)=∏ j∈N Y j (K) (22d)

[0166] in,

[0167] p j (K) = [p j (1),…,p j (NK)] T ,q j (K) = [q j (1),…,q j (NK)] T

[0168] The optimal problem of the distribution network operator can then be equivalent to the following:

[0169]

[0170] Subject to

[0171]

[0172]

[0173] 2) Distributed optimization based on fast dual decomposition algorithm

[0174] For each sub-period k∈K, let represents the non-negative dual variable (utils / pu) associated with the peak demand constraint (23b). Also, let the 1×3N row vector λv max (k) and λv min (k) represents the non-negative dual variable (utils / pu) associated with the three-phase voltage amplitude inequality constraint (23c). Let the row vector λl of 1×3N be max (k) and λl min (k) represents the non-negative dual variable (utils / pu) associated with the line blocking constraint (23d).

[0175] Row vector It consists of all non-negative dual variables and can be expressed as:

[0176]

[0177] Among them, the elements of λ(K) are expressed as follows:

[0178]

[0179] λv max (K)=[λv max (1),……,λv max (NK)] 1×(3N·K)

[0180] λv mim (K)=[λv min (1),……,λv min (NK)] 1×(3N·K)

[0181]

[0182]

[0183] The Lagrangian function L of the distribution network operator's optimal problem (23) is written as:

[0184]

[0185] Next, based on the fast dual decomposition algorithm, the distribution network operator optimal problem can be solved in a distributed manner.

[0186] Distributed Optimization Solutions for Distribution Network Operator Problems:

[0187] 1) Initialization: At the initial iteration time t = 0, IDSO sets λ 0 (K) = 0.

[0188] 2) Update y(K) as follows:

[0189]

[0190] L(y(K),λ t (K)) can be expressed as:

[0191] L(y(K),λ t (K))=∑ j∈N f j (y j (K))+C(K) (27)

[0192] have

[0193]

[0194] Among them, C(K) does not depend on y(K):

[0195]

[0196] In addition, Y(K) is decomposed to each aggregator level. Then (27) can be solved by each aggregator in a distributed manner:

[0197]

[0198] 3) Update λ(K). First, define the m×1 column vectors g(y(K)) and c as follows:

[0199]

[0200]

[0201] Then, the distribution network operator updates λ(K) as follows:

[0202] λ t+1 (K) = [λ t (K)+[g(y t (K))-c] T B] + (31)

[0203] in,[·] +represents the projection on [+,∞), N is a diagonal matrix whose main diagonal elements are the step size (i.e., the amount of change in the variable in each iteration). Specifically, the diagonal elements of B are v min (k), Related, the iterative variables are β1, β2, β3, β4, and β5 respectively.

[0204] Although distributed implementations of the distribution network operator optimization problem are developing, there is a lack of intuitive physical information to link it with the aggregator optimization problem. How to establish a connection between the two in a verifiable and admissible way remains to be solved. The flexibility of the algorithm can be increased and the convergence rate can be increased by designing a suitable B, that is, choosing different step sizes for different constraints.

[0205] Through this optimization model and algorithm, distribution network operators can more accurately adjust the balance between electric vehicle charging demand and the power supply capacity of the power grid, improve the operating efficiency and reliability of the power grid, and optimize the cost-effectiveness of electricity. In addition, the model supports real-time feedback and adjustment of power grid operation, enhancing the ability to adapt to uncertain and dynamically changing conditions.

[0206] Through this method, the present invention not only improves the economic operation efficiency of the power grid, but also increases the utilization rate of renewable energy, bringing significant economic and environmental benefits to power grid operators and electric vehicle users.

[0207] In a preferred embodiment of the present invention, it is demonstrated how to optimize the charging pricing and charging planning of electric vehicles through the centralized optimization model of the distribution network operator combined with the sequential Stackelberg model of the operator-aggregator. This process not only takes into account user characteristics and network constraints, but also effectively integrates the mathematical model of physical network behavior to achieve efficient and sustainable grid management. The specific implementation is as follows:

[0208] Substituting (18f) to (18h) into (28), it can be expressed as follows:

[0209]

[0210] Among them, γ(K) does not depend on y j (K):

[0211]

[0212] By using (32), the distributed solution of the distribution network operator's optimal problem can be equivalent to (29) through a physical information process when updating y(K), that is, the response to the retail price curve:

[0213]

[0214]

[0215]

[0216]

[0217] And there is

[0218]

[0219] H ψ (K) = diag(η ψ (1),η ψ (2),……,η ψ (K)) (35b)

[0220] From Proposition 1, we know that for a given but (34a)~(34b) show that aggregator j receives retail price curve set π j (K) determines the active and reactive power curves of its users’ electric vehicles. (34c) shows that at the node connecting each user, the aggregator generates the total active and reactive power curve, namely P ψ (K)=[P ψ (1),…,P ψ (K)] T , It shows that at node j, the aggregator generates the total active or reactive power curve, i.e., p j (K) = [p j (1),…,p j (K)] T ,q i (K) = [q i (1),…,q i (K)] T .

[0221] By retail price curve collection (34a)-(34b) establish the connection between the distributed solution of the optimal problem of distribution network operators and the optimal problem of aggregators. From (35a), we can observe that π ψ (k) exhibits an information structure. It is the sum of π0(k) plus the deviation of the user characteristic price, which is caused by the fiduciary responsibility of the distribution network operator to meet different distribution network constraints, including peak demand constraints, voltage constraints and line flow constraints. By using (34) and (35), the distributed solution of the distribution network operator's optimal problem can be replaced by: the operator-aggregator sequential Stackelberg model, the steps are as follows:

[0222] Step 1: Initialization At the initial iteration time t=0, the distribution network operator receives the initial retail price curve π0(k) from the regional transmission operator / isolated system operator or the utility company.

[0223] Step 2: The distribution network operator updates the dual variable. When t = 0, the operator sets the initial dual variable value λ(K) = 0. When t > 0, the I operator updates the dual variable λ t (K) = [λ t-1 (K)+[g(y t-1 (K))-c] T B] + .

[0224] Step 3: The operator updates and sends the retail price curve set to the aggregator. The operator acts as a guide and updates and sends to each aggregator j∈N based on (35a).

[0225] Step 4: Aggregators respond to the retail price curve. Each aggregator j∈N, as a follower, responds to the retail price curve based on the received The active and reactive power curves of the user's electric vehicle are determined according to (34a) to (34b).

[0226] Step 5: Aggregators feed back node power curves. Each aggregator j∈N forms the active and reactive power curves of user ψ∈Ψj according to (34c), and then forms the active and reactive power curves of node j according to (34d).

[0227] Step 6: Update the iteration time. Let t = t + 1. If no aggregator's active and reactive power curves cause the distribution network constraints to exceed the limit, this process stops and π(K) and P(π(K)) are determined. Otherwise, go back to step 2.

[0228] This model is suitable for urban and regional power grids, especially in areas with high electric vehicle penetration, and can effectively manage and optimize the charging needs of large-scale electric vehicles and support the integration of renewable energy.

[0229] Through this highly integrated and intelligent approach, the present invention not only improves the efficiency and reliability of grid operation, but also provides electric vehicle users with a more economical and greener charging solution. The implementation of this approach can significantly improve the grid's adaptability to future energy demand changes and overall sustainability.

[0230] In a preferred embodiment of the present invention, a method for implementing Stackelberg balance is described in detail. This concept is a highly coordinated electric vehicle charging pricing and scheduling framework established between distribution network operators and aggregators. Through the Stackelberg model, the grid operation efficiency can be optimized while meeting the charging needs of users and the safety and stability of the grid. The specific implementation method is as follows:

[0231] From (34), we can observe that y defined in (22) j (K) In j (K) is an exogenous variable, Q j (K), pj(K) and qj(K) are endogenous variables, which means that Q j (K), pj(K) and qj(K) depend on P j (K). Therefore y j (K) Available j (P j (K)), y(K) can be expressed as y(P(K)):

[0232] y(P(K))={y j (P j (K))|j∈N} (36)

[0233] Let π(K) and P(π(K)) be the set of retail price curves and active power curves of all users respectively:

[0234] π(K)={π j (K)|j∈N}

[0235] P(π(K))={P j (π j (K))|ψ∈Ψ}

[0236] Next, define the Stackelberg equilibrium as follows:

[0237] Assume that y(P * (K)) is the optimal solution to the distribution network operator's optimal problem (23). If P * (K) In some π * (K) and P(π * (K)) is consistent, then the quantity-price pair (P * (K), π * (K)) is the Stackelberg equilibrium.

[0238] In other words, if the distribution network operator sets the retail price curve to π * (K), then the electric vehicle charging plan P(π * (K)) depends on π *(K), which leads to the optimal solution y(P(π * (K)))=y((P * (K)). The following Proposition 2 states that π * (K) Analytical expression.

[0239] Proposition 2: Let (y(P * (K)),λ * (K)) is the saddle point of the Lagrangian function L(y(K),λ(K)) in (25). Definition Each aggregator j∈N Take the following form:

[0240]

[0241] Then (P * (K), π * (K)) is a Stackelberg equilibrium.

[0242] From (37), we can see that It is related to three factors:

[0243] 1) Personal preferences and attributes: μ ψ ,η ψ (k)

[0244] 2) Personal location information:

[0245] 3) Network constraint conflicts: λv max (k),λv min (k)

[0246] For example, two users are at the same location in the distribution network but have different preferences μ ψ , their EV charging retail prices are different because they have different price sensitivities to changing their charging behavior.

[0247] From the operator-aggregator sequential Stackelberg model algorithm and Proposition 2, we can see that as long as λ t (K) converges to λ * (K), then P t (K) and π t (K) will converge to P * (K) and π * (K), thus reaching the Stackelberg equilibrium.

[0248] In a preferred embodiment of the present invention, the convergence analysis of the optimal solution to the distribution network operator is described in detail to ensure the balance and feasibility of the Stackelberg model. This method ensures the stability and efficiency of the solution through systematic mathematical modeling and rigorous mathematical verification. The specific implementation is as follows:

[0249] Using (30), the distribution network operator optimal problem (23) can be expressed in a refined form:

[0250] max x∈X(K) F(x) depends on g(y(x))≤cx∈X(K) (38)

[0251] in,

[0252]

[0253]

[0254] First, from (11) and (17), X(K) is a non-empty compact convex set, satisfying the conditions [7, P5.A]. At the same time, from (9), (21) and (23a), the objective function can be expressed as:

[0255]

[0256] in,

[0257]

[0258]

[0259] According to Proposition 1, W is also a symmetric negative definite matrix, satisfying the conditions [7, P5.B]. In addition, according to (18f)-(18h), y is an affine function of x. From (19) and (30), it can also be concluded that g(·) is an affine function of y. Therefore, g(y(x)) is an affine function of x. Without loss of generality, g(y(x)) can be expressed in linear affine form:

[0260] g(y(x))=Cx+b (40)

[0261] As long as B satisfies:

[0262] I-JB ≥ 0, that is, I-JB is positive semidefinite (41)

[0263] in,

[0264] I is the identity matrix, J = C[-W] -1 C T

[0265] Then P t(K) and π t (K) will converge to P * (K) and π * (K), thus the Stackelberg equilibrium is convergent.

[0266] Through this method, the present invention not only provides a theoretically reliable and practically feasible power grid operation optimization model, but also ensures the effective application of the model in the actual power grid system, providing an efficient solution for modern power grid management and electric vehicle charging scheduling.

[0267] Example:

[0268] 1. Embodiment Overview:

[0269] This embodiment selects the improved IEEE 123-node power distribution network for numerical simulation, and the specific simulation parameters are as follows:

[0270] S base =100(KVA). The parameters are set to:

[0271] v0(k)=[1.01 2 ,1.01 2 ,1.01 2 ] T , v min (k) = [0.95 2 ,0.95 2 ,0.95 2 ] T , v max (k) = [1.05 2 ,1.05 2 ,1.05 2 ] T ,and

[0272] The distribution network consists of two types of users: User I and User II. They all have the same E ψ =20(kWh), m ψ =5(utils / pu)

[0273] a ψ =90%,η ψ (k) = 0.48, but μ ψ are not the same, where user I’s μ u =1(utils / cent), μ of user II ψ= 0.5 (utils / cent). Assume that there is an aggregator at each node responsible for managing all users at that node. The active and reactive power curves of all non-electric vehicles collected in the distribution network are as follows: Figure 2 In this embodiment, the specified time is from 20:00 to 6:00 the next day, with an interval of 15 minutes, that is, K = 40, Δt = 0.25h, and the initial retail price curve π0(K) obtained by the distribution network operator from the regional transmission operator / independent system operator or public utility company is as follows: Figure 2 (b) as shown.

[0274] 2. Electric vehicle charging management:

[0275] Assume that the distribution network operator does not manage the power usage of customers but instead sends the initial retail price curve directly to the aggregator. For this unmanaged distribution system, the peak demand is And because the maximum and minimum loads of the entire K period are between [-12,12], there will be no line congestion exceeding the limit, such as Figure 3 However, in Figure 4 In the figure, the voltage constraint exceeds the limit when k=13, and the minimum voltage level is 0.948(pu)<0.95(pu).

[0276] Next, the distribution network operator uses the proposed sequential Stackelberg model to manage user power usage. Under this model, there are no peak load limit violations or line congestion constraints. In addition, Figure 5 The maximum and minimum voltage levels of the distribution network in all k∈K time periods are shown, and no voltage constraint is exceeded. In order to study the optimal solution of the model, the results are compared with the centralized optimal solution directly obtained in the optimal problem (23). Taking the a-phase subnode of node 113 when k=13 as an example, its voltage level With active load The convergence of Figure 6 As shown. Figure 6 , converge to the same value, which is determined by the optimal solution in the set.

[0277] Then, consider Case 2: the line blocking constraint is reduced from 12 (pu) to 8 (pu). Figure 3 As shown in Figure 1, when k = 13, the unmanaged distribution system also has line blocking problems. Figure 7 The maximum line load (pu) of the distribution network under the sequential Stackelberg model for all time periods k∈K is shown. These results show that the sequential Stackelberg model enables distribution network operators to successfully manage different types of distribution network constraints.

[0278] 3. Retail price analysis:

[0279] Take Case 1 (line blocking constraint is 12) as an example, Figure 8 It shows that at different locations in the distribution network, users with the same personal preferences have different retail prices. At different locations in the distribution network, the same power usage will also have different effects on the distribution network constraints, resulting in different retail prices.

[0280] Figure 8 It also shows that the distribution network has the same location but μ ψ Different users have different retail prices. User I has a higher μ y =1, indicating that ψ is more concerned about price than practicality / comfort, showing higher price sensitivity. On the contrary, user II has a higher μ w =1, indicating that ψ is more concerned with practicality / comfort than price. ψ Users of the same price change will react differently. w Different products generally have different retail prices, which explains the influence of user preferences and characteristics.

[0281] The electric vehicle charging pricing system considering user characteristics and network constraints described below and the electric vehicle charging pricing strategy considering user characteristics and network constraints described above can refer to each other.

[0282] Please see attached Fig. 9 The present invention also provides an electric vehicle charging pricing system that takes into account user characteristics and network constraints, including:

[0283] The data receiving module 100 is configured to receive the energy consumption constraints of the power users and the charging requirements of the electric vehicles, as well as the real-time power flow and voltage constraint data of the distribution network;

[0284] The preliminary electricity price setting module 200 is configured to set a preliminary electricity price based on the operation status of the power distribution network according to the operation status data of the power distribution network;

[0285] The optimization processing module 300 includes:

[0286] a fast dual decomposition algorithm unit configured to apply a fast dual decomposition algorithm to process an electricity price and charging schedule optimization problem between a distribution network and an aggregator;

[0287] a sequential Stackelberg game model unit configured to optimize electricity prices and charging plans through a sequential Stackelberg game model;

[0288] The electricity price adjustment module 400 is configured to adjust the electricity price according to the optimization result to guide the electricity users to charge in an orderly manner, and take into account the charging flexibility of the users and the voltage fluctuation, line congestion and peak load constraints of the distribution network;

[0289] The control output module 500 is configured to output the optimized electricity price signal and charging plan to relevant electricity market participants and aggregators.

[0290] The system of this embodiment can be used to execute the above method embodiments, and its principles and technical effects are similar, which will not be repeated here.

[0291] Please see attached Fig.10 The present invention further provides a computer device 40, comprising: a processor 41 and a memory 42, wherein the memory 42 stores a computer program executable by the processor, and when the computer program is executed by the processor, the above method is executed.

[0292] The present invention further provides a storage medium 43 on which a computer program is stored. When the computer program is run by the processor 41, the above method is executed.

[0293] Among them, the storage medium 43 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0294] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A charging pricing method for electric vehicles considering user characteristics and network constraints, characterized in that: The following steps are involved: Receive energy consumption constraint information and network operation constraint information related to each user's electric vehicle; Based on the fast dual decomposition algorithm, a sequential Stackelberg game model is established between aggregators and distribution network operators. In the game model, the charging price of electric vehicles is set comprehensively according to the network constraints of the distribution network and the user's own preference characteristics; The sequential Stackelberg game model includes the following steps: Step 1: Initialization, at the initial iteration time step , the DNO receives an initial retail price curve from the RTO / ISO or utility ; Step 2: The distribution network operator updates the dual variables when When the distribution network operator sets the initial dual variable value ,when hour, Operator updates the dual variable ; Step 3: The distribution network operator updates and sends the retail price curve set to the aggregator. The distribution network operator acts as a guide based on: To each aggregator Update and send ; Step 4: Aggregators respond to the retail price curve. Each aggregator , as a follower, based on the received ,according to: Determine the active and reactive power curves of its users’ electric vehicles; Step 5: Aggregators feedback node power curves. Each aggregator according to: Forming users Active and reactive power curves, then follow: Aggregators Managed node location Active and reactive power curves; Step 6: Update the iteration time and set If there is no distribution network constraint exceeding the limit caused by any aggregator's active and reactive power curves, the process stops. and Otherwise, go back to step 2.

2. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 1 is characterized in that: The user characteristics include user preference settings.

3. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 1 is characterized in that: The network operation constraint information includes the operation status data of the power grid, and the operation status data of the power grid includes predicted load flow data, voltage level and line load conditions, wherein the aggregator reports the energy consumption plan for a period of time in the future to the distribution network and operates according to the plan.

4. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 1 is characterized in that: The fast dual decomposition algorithm includes: Modeling the grid status and user demand in each decision cycle; Using dual variables to solve voltage and line constraints in power grids; Iteratively update electricity prices and charging plans until convergence.

5. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 4 is characterized in that: The iterative updating includes processing power grid constraints using Lagrangian relaxation techniques and updating dual variables.

6. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 1 is characterized in that: The sequential Stackelberg game model includes: The distribution network operator acts as a guide to set the initial electricity price; Aggregators act as followers and adjust their EV charging strategies based on the electricity prices received.

7. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 6 is characterized in that: The charging strategy adjusted by the aggregator includes managing the charging rate of electric vehicles.

8. The electric vehicle charging pricing method considering user characteristics and network constraints according to claim 1 is characterized in that: The game model adjusts electricity prices to guide electricity users to charge in an orderly manner, taking into account the charging flexibility of users and the voltage fluctuations, line congestion and peak load constraints of the distribution network.

9. An electric vehicle charging pricing system considering user characteristics and network constraints, implementing the method according to any one of claims 1 to 8, characterized in that: include: a data receiving module configured to receive energy consumption constraints of power users and charging requirements of electric vehicles, as well as real-time power flow and voltage constraint data of the distribution network; A preliminary electricity price setting module, configured to set a preliminary electricity price based on the operation status of the power grid according to the operation status data of the power distribution network; Optimized processing modules, including: a fast dual decomposition algorithm unit configured to apply a fast dual decomposition algorithm to process an electricity price and charging schedule optimization problem between a distribution network and an aggregator; a sequential Stackelberg game model unit configured to optimize electricity prices and charging plans through a sequential Stackelberg game model; an electricity price adjustment module, configured to adjust the electricity price according to the optimization result to guide the electricity users to charge in an orderly manner, and taking into account the charging flexibility of the users and the voltage fluctuation, line congestion and peak load constraint of the distribution network; The control output module is configured to output optimized electricity price signals and charging plans to relevant electricity market participants and aggregators.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 8 is implemented.

Images