Power distribution network-traffic network-charging station layered operation scheduling based on three-phase imbalance

Abstract

The invention discloses a power distribution network-traffic network-charging station layered operation scheduling method based on three-phase imbalance, relates to the technical field of electric vehicle charging station energy management and transaction, solves the layered energy scheduling and transaction problem of an EVCS by establishing a three-stage method, and firstly solves the day-ahead traffic distribution problem of a traffic system; determining the charging load of each electric vehicle; secondly, solving the profit maximization problem of an electric vehicle manufacturer so as to determine energy scheduling and market participation decisions of the electric vehicle manufacturer; and finally, ensuring network operation of production consumer transactions by using a dynamic operation envelope, expressing a power distribution system operator problem as a three-phase imbalance three-phase optimal power flow (OPF) model integrated with a bilateral bidding mechanism, and introducing grid characteristics such as network reconstruction and tapping transformation into the OPF model to ensure the effectiveness of the framework.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management and trading of electric vehicle charging stations, and in particular to an operation and dispatching strategy based on a three-phase unbalanced hierarchical distribution network-transportation network-charging station. Background Art

[0002] Due to the urgency of addressing global climate change, the deployment of distributed energy sources such as solar photovoltaics (PV) and battery energy storage systems (BESS) has grown rapidly in recent years. Similarly, electric vehicles (EVs) and their associated charging infrastructure, namely electric vehicle charging stations (EVCS), have become increasingly popular. With the increasing deployment of EVCS, the operation of power distribution networks (PDNs) and urban transportation networks (UTNs) has become increasingly closely linked. In fact, traffic conditions within the UTN will affect the path decisions of electric vehicle drivers, which will affect the spatiotemporal load distribution of the EVCS network and the scheduling of local distributed energy sources at EVCS sites, and will also affect the operation of the PDN. Therefore, an emerging research direction is to coordinate the services of PDN and UTN and use EVCS or EV as the connection point.

[0003] A two-layer model based on an overall framework is proposed in the prior art to enhance the coordination of the operation of PDN and UTN through EVCS. The proposed two-layer model not only determines the EVCS charging service fee but also reduces traffic congestion while considering the uncertainty of traffic demand and wind power generation. There is also a two-layer optimization model for obtaining the optimal pricing of an electric vehicle network managed by a charging network operator. Unlike other PDN-UTN coordination models designed from the perspective of system operators, this two-layer model is a coordination model from the perspective of profit-driven charging network operators. Due to the surge in distributed energy resources, the concept of interactive energy has gained great popularity. Trading energy advocates the establishment of local energy markets at the PDN level, using economic and control mechanisms to improve system operation performance. On the one hand, the interactive control framework promotes EV aggregators to participate in the energy trading market (TEM) to alleviate the flexibility of voltage and congestion problems in PDNs using electric vehicle charging and discharging. On the other hand, the TEM clearing mechanism constrained by PDN enables electric vehicles to participate in the real-time retail electricity market.

[0004] Although the existing technology has extensively studied the coordination of PDN and UTN, most of the existing research focuses on the coordinated operation of PDN and UTN from a system perspective to improve system operation. In addition, most PDN-UTN coordination research adopts a centralized coordination method, which will cause practicality and privacy issues in practice. On the topic of TEM, the interconnection ability of electric vehicles with the grid is emphasized, and electric vehicles or electric vehicle aggregators are regarded as direct market participants in the proposed TEM framework. However, compared with a single electric vehicle with very limited power generation capacity, electric vehicles with on-site distributed energy have more significant and valuable potential and contribution to TEM. Summary of the invention

[0005] In view of the deficiencies of the prior art mentioned in the above background technology, the present invention proposes a charging station operation scheduling strategy for coordinating the services of PDN and UTN by using electric vehicles with on-site distributed energy. The focus of this strategy is to separate the functions of the grid operator and the market operator. In reality, due to the dual iterative process and the support of the required communication infrastructure, the cost of implementing such a framework may be higher than expected. Therefore, the functions of the grid and market operator are undertaken by one entity, namely the distribution system operator (DSO).

[0006] The design scheme provided by the present invention:

[0007] A distribution network-transportation network-charging station hierarchical operation scheduling based on three-phase imbalance includes the following steps:

[0008] Step 1: Establish a hierarchical multi-agent framework model to solve the energy management problem of electric vehicle charging stations (EVCS) with on-site distributed energy resources (DER), including energy scheduling and bidding strategies for participating in the energy trading market (TEM);

[0009] Step 2: Based on the characteristics of the urban transportation network, the transportation operator (TO) solves the traffic assignment problem (TAP) and assigns electric vehicles (EVs) to designated electric vehicle charging stations (EVCS);

[0010] Step 3: Based on the allocation of EVs to designated EVCSs by transportation operators, a dynamic operating envelope (DOE) calculation algorithm is used to enable the distribution system operator (DSO) to calculate the dynamic operating envelope (DOE) of the import and export of prosumers on the three-phase unbalanced distribution network;

[0011] Step 4: The distribution system operator (DSO) solves the unbalanced three-phase optimal power flow (OPF) problem of the distribution network (PDN) while clearing the TEM based on the submitted bids and quotations; the DSO, as the PDN operator, ensures that the system requirements are met while satisfying the PDN operational constraints.

[0012] Furthermore, the hierarchical multi-agent framework model consists of three agents: TO, MC (master controller) and DSO. The hierarchical multi-agent framework model solves the day-ahead energy scheduling problem of EVCS, in which each EVCS provides independent scheduling decisions and coordinates their actions through DSO; the Transportation Bureau, as the central entity supervising the transportation system, determines the EV flow within the UTN by solving the TAP (traffic assignment problem), thereby determining the schedule for EVs to arrive at the EVCS; the Ministry of Transport shares the forecast data of electric vehicle arrival with the traffic management company so that they can each solve the scheduling problem and ensure the reasonable distribution of vehicle loads per hour;

[0013] MC manages EVCS resources to maximize profits, each EVCS is a group of charging ports with some on-site DERs; each prosumer models its expected imported and exported electric energy and sends it to the DSO to obtain acceptance / limit DOE, arranging them to participate in TEM managed by the DSO to trade energy with their peers.

[0014] Furthermore, the step 2 specifically includes the following steps:

[0015] Step 2.1: Use the origin-destination (OD) model in UTN to model vehicle mobility for electric vehicles and light trucks under their respective traffic demands (in terms of vehicles per hour);

[0016] Step 2.2: The Transportation Bureau uses the TAP solution to determine road vehicle flow and EVCS allocation;

[0017] Step 2.3: TAP is described as a user equilibrium problem that captures the autonomous decisions of EV drivers. The TAP model is formulated based on the primal-dual optimality condition, where the dual problem allows the inclusion of road congestion.

[0018] Furthermore, the TAP objective function given in the TAP model contains four terms, among which the heterogeneity of electric vehicle charging prices, driving distance and congestion will affect the path decision of electric vehicles; the first two terms represent the costs related to free-flow travel time and travel delays (due to congestion) in the traffic system; it is assumed that the free-flow travel time on a particular road (provided as input) represents the distance traveled on that road; this statement is essential for representing road congestion, as road congestion is measured as a function of time, the third term represents the cost associated with the waiting time of electric vehicles in the queue, and the last term represents the charging cost of all electric vehicles at the EVCS station; therefore, when solving TAP, the goal of TO is to achieve the optimal trade-off of minimizing these four terms.

[0019] Furthermore, the step 3 specifically includes the following steps:

[0020] Step 3.1: The framework is divided into two layers. In the lower layer, the prosumers equipped with EVCS, DER and household loads (HLs) manage their energy, and the energy management output of the prosumers in the previous time slot obtain their expected net demand (import) and net generation (export) and send them to the DSO to obtain the DOE. In the upper layer, the DSO monitors the network constraints and manages the expected energy input / output of the prosumers by calculating the DOE.

[0021] Step 3.2: Combine the proposed framework with an algorithm for DSO calculation of DOE in three-phase unbalanced networks. The algorithm includes voltage and line congestion constraints.

[0022] Step 3.3: In the presence of DERs at the EVCS site, multiple EVCS resources need to be dispatched to maximize the benefit to the EVCS owner; in addition, the EVCS can participate in TEM to earn additional revenue by selling energy or by reducing costs by purchasing from other EVCS instead of purchasing from the grid.

[0023] Further, in step 3.2 in the DOE algorithm, if the expected input / output of the prosumer satisfies the network constraints through power flow analysis, the DSO accepts the input / output of the prosumer.

[0024] Further, the EVCS dispatching and trading problem in step 3.3: the presence of DER at the EVCS site leads to the need to dispatch multiple EVCS resources to maximize the benefits for the EVCS owner; in addition, EVCS can participate in TEM and obtain additional income by selling energy or by purchasing from other EVCS instead of purchasing from the grid to reduce costs.

[0025] Furthermore, the step 4 specifically includes the following steps:

[0026] Step 4.1: The DSO determines the optimal operating settings of PDN equipment such as tap switches and determines a feasible PDN topology through network reconstruction. In addition, the DSO acts as a market operator to clear TEMs and determine transaction prices and quantities.

[0027] Step 4.2: The hierarchical scheduling problem is reformulated by linearizing the nonlinear terms in the TO and DSO models.

[0028] Through the above design scheme, the present invention can bring the following beneficial effects:

[0029] The beneficial effects of the present invention are:

[0030] The present invention proposes an interactive dynamic operation scheduling strategy for charging stations considering a three-phase unbalanced hierarchical distribution network-transportation network, which is used to solve the energy management problem of profit-driven EVCS with on-site DER; the strategy model determines the energy scheduling of EVCS and the bidding strategy for participating in TEM, which considers the interaction between TO and EVCS on the one hand, and the interaction between EVCS and DSO on the other hand. The proposed DOE-based TEM framework enables consumers to ensure the safe operation of the network when using a designated DOE for TEM transactions. The proposed DOE calculation algorithm is computationally efficient and can achieve near real-time market clearing. Prosumers can manage their energy consumption by coordinating the charging and discharging of electric vehicles with photovoltaic systems to generate bids for TEM transactions within their assigned DOEs. A practical unbalanced three-phase optimal power flow (OPF) model is developed, which allows DSOs to simultaneously assume the functions of grid operators managing grid reliability and market operators managing TEM. The OPF model includes network reconstruction and tap-changing functions to ensure reliable PDN operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0032] Figure 1 A schematic diagram of the structure of a hierarchical multi-agent framework model provided by the present invention;

[0033] Figure 2 Flow chart of the DOE calculation framework provided by the present invention;

[0034] Figure 3 The present invention provides a logic diagram for the operation of a distribution network-transportation network-charging station hierarchical operation and dispatching strategy based on three-phase imbalance. DETAILED DESCRIPTION

[0035] In order to facilitate the understanding of the present application, the specific embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thoroughly understood.

[0036] The present invention provides a distribution network-transportation network-charging station hierarchical operation scheduling strategy based on three-phase imbalance, comprising the following steps:

[0037] Step 1: Establish a hierarchical multi-agent framework model to solve the energy management problem of profit-driven EVCS with on-site energy storage system (DER, DER can be photovoltaic and / or BESS units). The model determines the energy dispatch of EVCS and the bidding strategy for participating in TEM. It considers the interaction between TO (transportation operator) and EVCS on the one hand, and the interaction between EVCS and DSO on the other hand. The specific hierarchical multi-agent framework model is as follows Figure 1 As shown in Figure 1, the framework model is mainly composed of three main agents, namely TO, MC (Master Controller) and DSO. The hierarchical framework solves the EVCS day-ahead energy scheduling problem, where each EVCS provides independent scheduling decisions while coordinating their actions through the DSO. The Transportation Bureau, as the central entity overseeing the transportation system, determines the EV flow within the UTN by solving the TAP (Traffic Assignment Problem) to determine the schedule for EV arrival at the EVCS. The Ministry of Transportation shares the forecast data of electric vehicle arrival with the transportation management company so that they can solve the scheduling problem and ensure the reasonable distribution of vehicle load every hour. MC manages EVCS resources to maximize profits. Each EVCS is a group of charging ports with some on-site DERs. Each producer and consumer models its expected imported and exported electric energy and sends it to the DSO to obtain the acceptance / restriction DOE (dynamic operating envelope), arranging them to participate in the TEM managed by the DSO to trade energy with their peers.

[0038] A prosumer bidding strategy is introduced in the framework, where prosumers obtain bids to participate in TEM within a specified DOE. Overall, the proposed framework ensures the integrity of the distribution network, ensuring that voltage and line congestion constraints are not violated.

[0039] Step 2: Based on the characteristics of the urban transportation network, TO solves TAP and allocates EVs to designated EVCSs.

[0040] The step 2 specifically includes the following steps:

[0041] Step 2.1: Use the origin-destination (OD) model in UTN to model vehicle mobility for electric vehicles and light trucks under their respective traffic demands (in terms of vehicles per hour);

[0042] The graphical representation of UTN consists of a set of nodes (representing road intersections) and lines connecting these nodes (representing road segments). EV refers to electric vehicles that need to be charged before the end of the journey, and GV refers to electric vehicles and any other type of vehicles (e.g., gas engines) that do not require charging. The origin-destination (OD) model in UTN is used to model the vehicle mobility of electric vehicles and light trucks under their respective traffic demands (in vehicles per hour).

[0043] Step 2.2: The Transportation Bureau uses the TAP solution to determine road vehicle flow and EVCS allocation;

[0044] Step 2.3: TAP is described as a user equilibrium problem that captures the autonomous decisions of EV drivers. The TAP model is formulated based on the primal-dual optimality condition, where the dual problem allows the inclusion of road congestion.

[0045] The TAP model formula is as follows, where the TAP objective function given by equation (1) contains four terms, among which the heterogeneity of electric vehicle charging prices, driving distance and congestion will affect the path decision of electric vehicles. The first two terms represent the costs associated with free-flow travel time and travel delay (due to congestion) in the traffic system. It is assumed that the free-flow travel time on a particular road (provided as input) represents the distance traveled on that road. This statement is essential for representing road congestion, because road congestion is measured as a function of time. The third term represents the cost associated with the waiting time of electric vehicles in the queue, and the last term represents the charging cost of all electric vehicles at the EVCS station. Therefore, when solving TAP, the goal of TO is to achieve the best trade-off of minimizing these four terms.

[0046]

[0047] Where: pr tt ,pr ttd Prices associated with travel time and travel time delay; is the free-flow travel time (i.e. no congestion); is the traffic flow of electric vehicles (i.e., electric vehicles that need to be charged) on road a; Traffic flow of GVs (i.e., electric vehicles and any other vehicles that do not need to be charged) on road a; t w1 ,t w2 Waiting time for queue No. 1 and No. 2; is the number of EVs assigned to EVCS on bus i and the charging power requirement of EVCS on bus i; Charging price for EVCSi.

[0048] Constraints (2)-(12) are the original constraints of the TAP model. For all OD pairs, the relationship between the traffic flow on the route and the traffic flow on individual roads is shown in (2), where a route specifies the set of roads between the start and end points. (3) defines the road capacity limit, and (4) represents traffic flow conservation (that is, the traffic demand of the OD pair is equal to the sum of the traffic flows on all paths connecting the start and end points). (5)-(11) give the routing and charging constraints for electric vehicles. Constraint (5) indicates that the total number of electric vehicles allocated to all EVCSs is equal to the electric vehicle traffic demand. Constraint (6) represents the allocation balance of a single electric vehicle, where is the number of electric vehicles in the charging process, and The electric vehicles queued for the first and second positions respectively. Constraint (7) ensures that the EVCS is at full capacity before adding an EV to the queue. In addition, constraint (7) limits each position in the queue to accommodate only one EV. Constraint (8) limits the total electric vehicle flow through each EVCS node and at the same time limits the total amount of electric vehicle traffic flow allocated to each road segment. Constraint (10) represents the relationship between the charging power demand and the number of electric vehicles allocated to the EVCS at the corresponding UTN node. The charging capacity of each EVCS station is shown in (11). Constraint (12) ensures that all vehicle flows are positive.

[0049]

[0050]

[0051] Where: A a,rs,k , For OD (r, s), if path k contains path a, the incidence matrix (transposed incidence matrix) is defined as 1, otherwise it is defined as 0; is the traffic flow of OD for k-way electric vehicle (GV); c a is the road capacity; α and β are the penetration rates of electric vehicles and fast charging users; q rs,t is the traffic demand between OD pairs at time t (r, s); N cu is the number of charging ports; D i,rs,k , It is defined as follows: if the path k between the OD pair (r, s) does not pass through the electric bus, the incidence matrix (transposed incidence matrix) is defined as 1, and if it is defined as 0; p cu is the charging power rating; Δp is the conversion factor from electric vehicle to charging power demand.

[0052] Constraints (13)-(15) define the feasible domain of the dual variables. The dual multipliers of (2) and (4) are the variables and πrs,t The dual multiplier λ associated with (3) a,t The concept of congestion is introduced, where λ is a,t A non-negativity constraint is imposed. Finally, a strong dual constraint (16) is constructed to ensure that the primal problem and the dual problem have the same solution.

[0053]

[0054] λ a,t ≥0(15)

[0055]

[0056] Where: λ a,t , is the bivariate of the original TAP; E rs,k , It is defined as follows: if the path k corresponds to the OD pair (r, s), the corresponding position in the incidence matrix (or transposed incidence matrix) is defined as 1; otherwise it is defined as 0;

[0057] The TAP solved by the Transportation Department is briefly introduced as follows:

[0058]

[0059] Step 3: Based on the allocation of EVs to designated EVCSs by transport operators, the DOE calculation algorithm enables the DSO to calculate the DOE of imports and exports of prosumers on the three-phase unbalanced distribution network. The overview of the framework is as follows: Figure 2 shown.

[0060] Step 3.1: The framework model is divided into two layers. In the lower layer, the prosumers equipped with EVCS, DER and household loads (HLs) manage their energy, and the energy management output of the prosumers in the previous time slot is obtained to obtain their expected net demand (import) and net generation (export) and send it to the DSO to obtain the DOE. In the upper layer, the DSO monitors the network constraints and manages the expected energy input / output of the prosumers by calculating the DOE.

[0061] Consider a distribution network with a set of n∈N@{1,L,n,L,N} nodes connected by a set of distribution lines represented by l∈L@{1,L,l,L,L}. A set of prosumers are connected to different nodes of the distribution network. Each prosumer can be configured with any of the following loads and sources: EVCS, DER, and HLs. At any time t∈T@{1,L,t,L,T}, according to the net power P of the prosumer in (1) j,t , prosumer j acts as a producer or consumer.

[0062]

[0063] Where: P j,t represents the net power, which refers to the total power demand at time t and total power generation The total power demand and total power generation of producer consumer j can be calculated by equations (19) and (20), respectively:

[0064]

[0065] Charging power for electric vehicles and discharge power Subject to the constraints of equations (21) and (22), they are:

[0066]

[0067] Where: Refers to the maximum rating of the electric vehicle charger, the state variable Used to indicate the charging / discharging status of electric vehicles, defined as when When the electric vehicle is in charging mode, When the electric vehicle is in discharge mode, When the electric vehicle is in idle mode, it is allowed to charge and discharge when it is within the constraints of equations (21) and (22). According to equation (23), the state variables are set according to the mobility mode of the electric vehicle. Thereby configuring the charging or discharging of electric vehicles.

[0068]

[0069] Where: They are the current state of charge (SOC), minimum SOC, maximum SOC and the expected SOC of the jth electric vehicle; They are respectively represented as the available time of the electric vehicle and the remaining time to reach the desired SOC, which can be calculated by formula (24); are respectively the local market price, the grid repurchase price and the electric vehicle discharge price difference. Equation (25) describes the Initial SOC Start charging / discharging and leave the Based on the required SOC Start charging / discharging. Current SOC The constraints and updates of equations (26) and (27) are:

[0070]

[0071] Where: and represents the action time and the maximum capacity of the electric vehicle battery; τ refers to the time slot duration, η ch and η ch Refers to the charging and discharging efficiency of electric vehicles.

[0072] Power generation for photovoltaic system prosumers Subject to the constraints of formula (28):

[0073]

[0074] Where is the peak power generation of the photovoltaic system. In this framework, the cost of photovoltaic power generation is negligible, so the photovoltaic power generation system is used as the priority power source for household loads and electric vehicle charging. It means that the battery is fully charged at the fastest speed without reducing any load.

[0075] Step 3.2: In the proposed framework model, the Designated Operating Entity is determined by the DSO to limit the expected import / export of the prosumers in the presence of EVs and PV systems, thereby operating the distribution network within its operational constraints. The DOE is used to ensure the security of network operation, taking into account the imports and exports of the prosumers, so that the prosumers can participate in TEM within the assigned DOE without any network challenges. The DOE can communicate directly with the prosumers, so the prosumers can manage their assets autonomously and limit the access of other agents to their information and assets. The proposed framework incorporates an algorithm for the DSO to calculate the DOE in a three-phase unbalanced network, which takes into account voltage and line congestion constraints, where the execution of the DOE through Node voltage V n,t and line current I l,t In the DOE algorithm, if the expected input / output of the producer / consumer satisfies the network constraints through power flow analysis, the DSO accepts the input / output of the producer / consumer. Otherwise, the input / output is restricted by the algorithm proposed below.

[0076] The algorithm includes an objective function as described in equation (29), which is designed to minimize the absolute difference between the expected input / output and the assigned DOE, which is formulated as the constraints of equations (30)-(37).

[0077]

[0078] Constraints:

[0079]

[0080]

[0081] Constraints (30) and (31) limit the input power and output power of the production and consumer, respectively, and neither of them can exceed the expected power P j,t The input / output current and phase φ of the producer-consumer j at node n can be obtained by equation (32), where φ is a set of phase values ​​of the distribution network, defined as φ∈Φ={a,b,c}, is the nominal voltage, α n,φ is the vector voltage angle at phase φ. Use equation (33) to calculate the line current I flowing on line l between the upstream node (n-1) and node n l,t , where k is a set of connecting lines connecting node n and downstream, defined as k∈K@{1,L,k,L,K}. The node voltage V is calculated by equation (34): n,t , where V n-1,t is the node voltage of the upstream node (n-1), R l and X l is the resistance and reactance matrix of the network. The voltage and line congestion constraints are shown in (35) and (36) respectively. The total reactive power of the PV system is Maximum reactive power As shown in formula (37), all parameters and variables are expressed in unit values.

[0082] Step 3.3: EVCS dispatch and trading problem: The presence of DER at the EVCS site leads to the need to dispatch multiple EVCS resources to maximize the benefits for the EVCS owner. In addition, EVCS can participate in TEM and earn additional revenue by selling energy or by reducing costs by purchasing from other EVCS instead of the grid. The objective function of the MC energy dispatch problem of EVCS is expressed as (38). The objective function includes the respective internal (39) and external (40) operating costs. The internal operating cost is the cost of using the EVCS asset. The first term in (39) represents the aging cost of the BESS, while the second term represents the opportunity cost loss caused by underutilizing the photovoltaic (PV) output. The internal cost in the MC problem only considers the asset utilization without considering the operating expenses. The external cost in (40) basically represents the cost of purchasing electricity from TEM or the grid. The revenue from selling electricity to TEM is defined as shown in (41).

[0083]

[0084] Where: pr BESS , η c , η d is the BESS attenuation cost and charging and discharging efficiency coefficient; Charging and discharging the BESS of bus i; The price of electricity purchased from the grid; Purchase electricity from the grid for EVCS and dispatch PV on the i-bus; PB i,t ,PS i,t It is the planned electric energy that EVCS buys and sells from TEM on bus i.

[0085] The MC scheduling problem is subject to the following constraints. The initial charge state of the BESS unit and the charge state at the end of the scheduling time horizon are defined as (42). The BESS unit operates in one of the three operating modes (i.e., charging, idling, and discharging modes) shown in (43). Equation (44) is the BESS time-coupled energy balance constraint, and the operating range of the BESS charge state is shown in (45). Constraints (46) and (47) restrict the charging and discharging operations.

[0086]

[0087] Where: are the initial, minimum and maximum charge states of the BESS on bus i; is the minimum and maximum power of the unit; It is a binary variable for the BESS unit in i-bus charging, idling and discharging modes.

[0088] The power balance of the EVCS is shown in (48). Constraint (49) imposes a non-negativity constraint on the purchase of power from the grid and specifies restrictions on the dispatch of PV generation. The value of M in (50) ensures that the EVCS cannot be both a buyer and a seller at the same time. Constraint (51) will limit the price offered by the EVCS when it is a seller, which is limited to the total energy provided by the on-site PV and BESS units.

[0089]

[0090] In contrast, energy purchase is limited to the total energy demand of charging EVs and BESS units, as shown in Equation (52).

[0091]

[0092] Where: is a binary variable representing the transaction price of market participants in TEM.

[0093] The management committee’s scheduling problem is stated as follows:

[0094]

[0095] Step 3.4: Bidding and Quoting of EVCS: Market participants MC determine their respective EV charging station decisions by solving equation (53), so bids and quotes are formed based on the quantities determined by this equation. For sellers, the quote consists of the bid quantity and the bid price. The bid price is set as a function of the DER installation cost of the EV charging station and the number of EVs allocated (i.e., a time-varying coefficient), which is considered to reflect the scarcity of charging points. In essence, EV charging stations tend to set lower bid prices during periods when charging demand is expected to be lower, to avoid being unable to complete clearing in TEM and having to limit their on-site DERs.

[0096] EV charging stations, positioned as buyers, have two types of loads: fixed loads represent the required EV charging power, while flexible loads are used for BESS charging. We consider that EV charging stations are willing to accept higher TEM prices to satisfy their fixed hourly loads. However, EV charging stations intend to use lower TEM prices to satisfy flexible loads. We use a value-based pricing approach to set the prices for fixed and flexible loads. The flexible bidding quantity (corresponding to flexible loads) and the fixed bidding quantity (corresponding to fixed loads) are calculated by solving equation (53).

[0097] Step 4: DSO solves the OPF problem of PDN and clears TEM based on the submitted bids and quotations. As the PDN operator, DSO ensures that the system requirements are met while meeting the PDN operation constraints.

[0098] Step 4.1: The DSO determines the optimal operating settings of PDN equipment such as tap switches, and determines a feasible PDN topology through network reconstruction. In addition, the DSO acts as a market operator to clear TEMs and determine transaction prices and quantities. The DSO's objectives in equation (54) include the PDN loss terms described in (55). In equation (55), since a fixed demand must be met, minimizing the power supply is equivalent to minimizing the line loss (i.e., the difference between the power supply and the demand). In addition, the DSO maximizes buyer satisfaction and minimizes seller purchase costs through a two-sided auction provided by (56).

[0099] minC loss -C TEM (54)

[0100]

[0101] Where: is the flexible bidding price of EVCS on bus i; is the quote of EVCS on bus i; To clear the flexible amount of power in the TEM via the DSO.

[0102] The three-phase unbalanced model of the DSO with embedded TEM scavenging architecture is shown below. The active and reactive balance terms are expressed as Equations (57) and (58), respectively, the power injection of each EVCS bus node is shown in (59), and the power balance of TEM scavenging is shown in (60).

[0103]

[0104] Where: Pl i,f,t 、Ql i,f,t is the active (reactive) line flow; is the active (reactive) load of the bus i power distribution system (excluding charging load); It is the amount of electricity purchased (sold) specified by the DSO in the TEM; Adjust variables for imbalance of aggregate TEM power.

[0105] Given the limited supply from sellers and the high charging requirements from buyers, TEMs are bound to experience energy imbalances between bids and offers for hours. is used to explain the market energy imbalance, where the corresponding shortage is supplied by the grid. The proposed model can be easily extended to implement EVCS to fix demand reduction and energy import restrictions from the grid. The market clearing rule and constraint variables are defined in (61). Given a switch state z l,t , then the voltage drop of line l between busbars i and j is expressed as (62) and (63). The operating limits of voltage and line power flow are given in (65) and (66), respectively.

[0106]

[0107]

[0108] Where: is a fixed amount of power removed from the TEM by the DSO; Φ, Φ l For electric buses and lines; Φ S ,Ω,T is the setting of substation bus, phase and time range; t,i is the time index and electric bus; Fixed and flexible bidding for transient electromagnetic power imbalance adjustment variables; U i,f,t is the voltage squared; V, are the minimum and maximum voltage limits; l,t It is the circuit switch status; is the line capacity.

[0109] Consider a ratio y connected to line l lFor the tap changer, assume that there is a fictitious bus m between buses i and j. The tap changer constraints are given by (67) to (70). The line flow conservation from bus i to bus j is shown in (67). The voltage drop from bus i to bus j is expressed as (68), where the line loss occurs between bus i and bus m. The voltage relationship across the tap changer is shown in (69), and the turns ratio and tap position limits are expressed in (70). The integer variable representing the position of the actual tap is The network reconfiguration constraints are given in (71)-(74), and the radiality of the PDN is implemented by (71) and (72). The relationship between the switch action variable and the switch is shown in (73), and the maximum number of switch operations is limited by (74).

[0110]

[0111] Where: S, l, f are substation bus, line and phase; Δy l 、Nt l is the minimum speed ratio, the speed ratio change of each tap and the total tap position; is a binary variable for the switch-on action (switch-on is 1) and the switch-off action (switch-off is 1); Z max is the maximum number of swap operations.

[0112] Step 4.2: The hierarchical scheduling problem is reformulated by linearizing the nonlinear terms in the TO and DSO models. The nonlinear terms in (1) Linearization is performed by applying the Karush-Kuhn-Tucker optimality conditions. Thus, the TAP problem (17) is transformed into a linear programming problem. Linearization techniques are used to reformulate (69) at the DSO level, replacing it with the linear equation (77). The big-M method is used to effectively replace the nonlinear equation (75) with the linear equations (78) and (79).

[0113]

[0114]

[0115] Where: u l,n,t , Expand binary variables for binary.

[0116] Therefore, the DSO problem is reformulated as a mixed integer second-order cone programming problem as follows:

[0117]

[0118] After the first iteration, the calculated transaction price and quantity are communicated to the participating EVCS. The scheduling problem of the new MC is then solved using the updated transaction price given in (61) to ensure the convergence of the iterative process. The convergence rate parameter σ is adjusted heuristically here.

[0119]

[0120] The updated settlement price is The MC of charging station usage will increase as the difference in consecutive prices decreases; conversely, it will decrease. When the latest clearing price is greater than the previous clearing price, the latest clearing price will get closer and closer to the buyer's "infeasible" area (i.e. ). In essence, TEM is only feasible when the clearing market price is lower than the grid offer; otherwise buyers have no incentive to participate in TEM. Sellers will therefore make a profit by participating in TEM (by minimizing their opportunity cost losses). Figure 3 The iterative process in will continue until no additional changes are observed in the cleared TEM prices.

[0121] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. Hierarchical operation and dispatch of distribution network, transportation network and charging station based on three-phase imbalance, characterized by: The following steps are involved: Step 1: Establish a hierarchical multi-agent framework model to solve the energy management problem of electric vehicle charging stations (EVCS) with on-site distributed energy resources (DER), including energy scheduling and bidding strategies for participating in the energy trading market (TEM); Step 2: Based on the characteristics of the urban transportation network, the transportation operator (TO) solves the traffic assignment problem (TAP) and assigns electric vehicles (EVs) to designated electric vehicle charging stations (EVCS); Step 3: Based on the allocation of EVs to designated EVCSs by transportation operators, a dynamic operating envelope (DOE) calculation algorithm is used to enable the distribution system operator (DSO) to calculate the dynamic operating envelope (DOE) of the import and export of prosumers on the three-phase unbalanced distribution network; Step 4: The distribution system operator (DSO) solves the unbalanced three-phase optimal power flow (OPF) problem of the distribution network (PDN) while clearing the TEM based on the submitted bids and quotations; the DSO, as the PDN operator, ensures that the system requirements are met while satisfying the PDN operational constraints.

2. The distribution network-transportation network-charging station hierarchical operation scheduling based on three-phase imbalance according to claim 1 is characterized by: The hierarchical multi-agent framework model consists of three agents: TO, MC (master controller) and DSO. The hierarchical multi-agent framework model solves the day-ahead energy scheduling problem of EVCS, in which each EVCS provides independent scheduling decisions and coordinates their actions through DSO; The Transportation Bureau, as the central entity overseeing the transportation system, determines the EV flow within the UTN by solving the TAP (Traffic Assignment Problem) to determine the schedule for EV arrival at the EVCS; the Ministry of Transportation shares the forecast data of EV arrival with the Transportation Management Company so that each can solve the scheduling problem and ensure the reasonable distribution of vehicle load every hour; MC manages EVCS resources to maximize profits, each EVCS is a group of charging ports with some on-site DERs; each prosumer models its expected imported and exported electric energy and sends it to the DSO to obtain acceptance / limit DOE, arranging them to participate in TEM managed by the DSO to trade energy with their peers.

3. The hierarchical operation scheduling of distribution network-transportation network-charging station based on three-phase imbalance according to claim 1 is characterized in that: The step 2 specifically includes the following steps: Step 2.1: Use the origin-destination (OD) model in UTN to model vehicle mobility for electric vehicles and light trucks under their respective traffic demands (in terms of vehicles per hour); Step 2.2: The Transportation Bureau uses the TAP solution to determine road vehicle flow and EVCS allocation; Step 2.3: TAP is described as a user equilibrium problem that captures the autonomous decisions of EV drivers. The TAP model is formulated based on the primal-dual optimality condition, where the dual problem allows for the inclusion of road congestion.

4. The hierarchical operation and dispatching of distribution network-transportation network-charging station based on three-phase imbalance according to claim 3 is characterized in that: The TAP objective function given in the TAP model contains four terms, among which the heterogeneity of electric vehicle charging prices, driving distance and congestion will affect the path decision of electric vehicles; the first two terms represent the costs related to free-flow travel time and travel delays (due to congestion) in the traffic system; it is assumed that the free-flow travel time on a particular road (provided as input) represents the distance traveled on that road; this statement is essential for representing road congestion, because road congestion is measured as a function of time, the third term represents the cost associated with the waiting time of electric vehicles in the queue, and the last term represents the charging cost of all electric vehicles at the EVCS station; therefore, when solving TAP, the goal of TO is to achieve the optimal trade-off of minimizing these four terms.

5. The hierarchical operation and dispatching of distribution network-transportation network-charging station based on three-phase imbalance according to claim 1 is characterized in that: The step 3 specifically comprises the following steps: Step 3.1: The framework is divided into two layers. In the lower layer, the prosumers equipped with EVCS, DER and household loads (HLs) manage their energy, and the energy management output of the prosumers in the previous time slot obtain their expected net demand (import) and net generation (export) and send them to the DSO to obtain the DOE. In the upper layer, the DSO monitors the network constraints and manages the expected energy input / output of the prosumers by calculating the DOE. Step 3.2: Combine the proposed framework with an algorithm for DSO calculation of DOE in three-phase unbalanced networks. The algorithm includes voltage and line congestion constraints. Step 3.3: In the presence of DERs at the EVCS site, multiple EVCS resources need to be dispatched to maximize the benefit to the EVCS owner; in addition, the EVCS can participate in TEM to earn additional revenue by selling energy or by reducing costs by purchasing from other EVCS instead of purchasing from the grid.

6. The distribution network-transportation network-charging station hierarchical operation scheduling based on three-phase imbalance according to claim 5 is characterized in that: In step 3.2 in the DOE algorithm, if the expected input / output of the prosumer satisfies the network constraints through power flow analysis, the DSO accepts the input / output of the prosumer.

7. The distribution network-transportation network-charging station hierarchical operation scheduling based on three-phase imbalance according to claim 5 is characterized in that: EVCS dispatch and trading problem in step 3.3: The presence of DER at the EVCS site leads to the need to dispatch multiple EVCS resources to maximize the benefit of the EVCS owner; in addition, EVCS can participate in TEM and obtain additional income by selling energy or reducing costs by purchasing from other EVCS instead of purchasing from the grid.

8. The distribution network-transportation network-charging station hierarchical operation scheduling based on three-phase imbalance according to claim 1 is characterized in that: The step 4 specifically comprises the following steps: Step 4.1: The DSO determines the optimal operating settings of PDN equipment such as tap switches and determines a feasible PDN topology through network reconstruction. In addition, the DSO acts as a market operator to clear TEMs and determine transaction prices and quantities. Step 4.2: The hierarchical scheduling problem is reformulated by linearizing the nonlinear terms in the TO and DSO models.

Images