Electric freight vehicle team power distribution network collaborative scheduling method and system based on node electricity price

Abstract

The application discloses a node electricity price-based electric freight vehicle team power distribution network cooperative scheduling method and system, relates to the electric vehicle scheduling and power system optimization cross technical field, and through the establishment of the topological mapping relationship between the traffic network and the power distribution network, aggregates and introduces the charging and discharging behavior of the electric freight vehicle team at different nodes and time periods into the power distribution network market clearing model, and endogenously forms the time and space distributed node electricity price under the condition of meeting the safety constraints such as power flow, voltage and line capacity; meanwhile, the node electricity price is taken as the charging and discharging decision basis in the vehicle team scheduling model, the joint optimization of path planning, cargo loading and charging and discharging scheduling is realized. Through embedding the optimality condition of the power distribution network market clearing problem into the vehicle team scheduling model, the closed-loop coupling relationship between the vehicle team behavior and the electricity price formation is formed, so that the time and space arbitrage and global cooperative optimization of low-price charging and high-price discharging are realized while ensuring the safety constraint feasibility of the power distribution network.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle dispatching and power system optimization, specifically to a method and system for coordinated dispatching of electric freight fleets and distribution networks based on nodal pricing. Background Technology

[0002] With the advancement of global carbon neutrality goals and the acceleration of urban logistics electrification, electric freight trucks (EFTs) are being widely deployed in urban delivery, regional logistics, and cold chain transportation. Compared to traditional fuel-powered freight vehicles, electric freight trucks offer significant advantages such as zero emissions, low noise, and low operating costs. However, the large-scale promotion of electric freight trucks also faces challenges such as limited driving range, long charging times, and uneven distribution of charging infrastructure, making charging decisions one of the core issues in fleet scheduling.

[0003] Meanwhile, the deepening reforms of the electricity market have led to significant time and spatial dependence of charging prices. In the time dimension, prices exhibit marked intraday fluctuations due to factors such as system load levels and renewable energy output. In the spatial dimension, differences in network topology location, local load levels, and line congestion levels result in differentiated nodal marginal prices. This spatiotemporal distributed pricing structure provides electric freight fleets with opportunities for charging and discharging arbitrage, but also significantly increases the complexity of dispatching decisions.

[0004] Existing electric freight fleet dispatching technologies have the following shortcomings: First, the electricity pricing model is overly simplistic. Most existing route planning studies assume a fixed electricity price or use simplified time-of-use pricing models, neglecting the impact of distribution network constraints on nodal prices. This simplification fails to capture the price differences between nodes caused by factors such as network congestion and voltage overruns in real power systems, and also fails to reflect the feedback effect of the electric vehicle fleet's own charging and discharging behavior on prices. In reality, under the Distribution Locational Marginal Price (DLMP) pricing mechanism, nodal prices are not only related to time periods, but also closely related to the node's topological location, the load level carried by the node, and the degree of congestion of upstream lines.

[0005] Second, there is a lack of closed-loop coupling modeling. Existing methods typically treat grid operation and fleet dispatching as independent systems, failing to establish a two-way feedback mechanism between the charging / discharging behavior of electric freight fleets and the operating status of the distribution network. In reality, the centralized charging of large-scale electric freight fleets significantly affects the power flow distribution and node voltage levels of the distribution network. When the charging load exceeds line capacity or causes voltage overruns, system operating costs will rise sharply, and the marginal electricity price at each node will also increase accordingly. Ignoring this two-way coupling relationship will lead to fleet dispatching schemes either violating grid safety constraints or incurring higher-than-expected charging costs in actual implementation.

[0006] Third, the value of V2G has not been fully explored. Vehicle-to-Grid (V2G) technology allows electric freight trucks to generate revenue by discharging electricity into the grid during peak electricity price periods. Electric freight trucks are typically equipped with large-capacity batteries (100-300kWh), providing considerable redundancy for V2G services while meeting transportation needs. However, existing scheduling methods rarely consider the spatiotemporal arbitrage opportunities of V2G, failing to fully leverage the electricity price differences between different time periods and nodes to maximize the revenue of electric freight trucks.

[0007] Fourth, the decision-making processes for route planning, loading, and charging are fragmented. Electric freight fleet scheduling involves three tightly coupled sub-problems: route planning, cargo loading, and charging scheduling. Traditional methods often employ a phased solution strategy: first, determine the route structure; then, optimize the loading plan; and finally, arrange charging scheduling. This sequential approach ignores the interrelationships among these three elements—for example, cargo load affects energy efficiency, thus influencing charging demand; and charging time affects delivery time, thus constraining feasible routes. This fragmented decision-making leads to higher overall operating costs and makes it difficult to achieve global optimum.

[0008] Fifth, the scalability of the solution algorithm is insufficient. After co-modeling electric freight fleet scheduling with power grid operation, the problem size expands dramatically, with the number of variables proportional to the product of the number of time periods, nodes, and vehicles. Traditional mixed-integer linear programming (MILP) exact solution methods experience exponential growth in computation time when dealing with large-scale instances, making it difficult to meet the efficiency requirements of actual operation. Due to these technical problems, existing electric freight fleet scheduling methods struggle to achieve efficient, stable, and scalable collaborative decision-making under complex electricity price environments and power grid constraints, often resulting in excessively long computation times for scheduling schemes, difficulty in responding to actual operational needs in a timely manner, and low scheduling efficiency. Summary of the Invention

[0009] To address the shortcomings mentioned in the background section, the present invention aims to provide a method and system for coordinated dispatching of electric freight fleet distribution networks based on nodal pricing.

[0010] Firstly, the objective of this invention can be achieved through the following technical solution: a method for coordinated dispatching of electric freight fleets across power distribution networks based on nodal pricing, the method comprising the following steps: The system receives input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines. Based on a pre-built mapping function, the set of rechargeable nodes is mapped to the set of bus nodes to obtain the node set mapping relationship. The set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes, and the set of customer nodes with charging capabilities. The input data and the mapping relationship of the node set are input into the pre-established collaborative optimization scheduling model, and the optimal collaborative scheduling scheme is output. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model. It is obtained by optimizing the two-level model into a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

[0011] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the set of customer nodes with charging capabilities is a set of customer nodes divided based on the configuration of charging facilities, wherein the set of customer nodes equipped with charging facilities is divided into a set of customer nodes with charging capabilities, and the set of customer nodes not equipped with charging facilities is divided into a set of customer nodes without charging capabilities.

[0012] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the upper-level electric freight fleet route loading and charging coordination model defines route decision variables including arc selection binary variables, vehicle activation binary variables and node access binary variables, defines time decision variables including node arrival time and node departure time, defines loading decision variables including cargo load at node departure, and defines energy decision variables including battery state of charge at node arrival, battery state of charge at node departure, charging amount at each time period and discharging amount at each time period; The objective function of the upper-level electric freight fleet route loading and charging coordination model minimizes the total operating cost, including fixed deployment cost, travel distance cost, time operating cost, time window violation penalty cost, and net charging cost. It also constructs route connectivity constraints, time dynamic constraints, cargo loading dynamic constraints, battery state of charge dynamic constraints, and spatiotemporal price response charging constraints.

[0013] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the objective function of the upper-level electric freight fleet route loading and charging coordination model is as follows: The fixed deployment cost is as follows: In the formula, K represents the set of vehicles. For vehicles k Whether a binary decision variable is enabled or not. This represents the fixed deployment cost coefficient for a single vehicle. The cost of the travel distance is: In the formula, Let be the set of nodes that can serve as the starting point of an arc segment. Let be the set of nodes that can serve as the endpoints of an arc segment. For vehicles k Does it pass through an arc segment? i,j binary decision variables, For arc segment ( i,j The distance, Cost coefficient per unit distance; The time operating cost is: In the formula, For vehicles k Return to arrival time at the terminal station. For vehicles k Time of departure from the starting station This is the operating cost coefficient per unit time. The penalty cost for violating the time window is: In the formula, C is the set of customer nodes. For vehicles k The amount of time window violation at customer C. The penalty coefficient for violations within a unit time window; Net charging cost is: In the formula, Let T be the set of rechargeable nodes and T be the set of time periods. For nodes i During the period t The marginal electricity price at the distribution network node For vehicles k At the node i time period t The amount of charge, For vehicle k at node i time period t The amount of discharge.

[0014] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the lower-level distribution network market clearing model adopts the linearized distribution network power flow model LinDistFlow, with the objective function being to minimize the system operating cost, including the main grid power purchase cost, photovoltaic power generation cost, and energy storage operating cost; constructing node active power balance constraints, node reactive power balance constraints, active power recursive constraints, reactive power recursive constraints, voltage drop equations, photovoltaic output constraints, energy storage charging and discharging power constraints, energy storage state of charge constraints, voltage amplitude constraints, and line capacity constraints; the dual variable of the node active power balance constraints is defined as the marginal electricity price of the distribution network node.

[0015] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the objective function of the lower-level distribution network market clearing model is as follows: In the formula, Main grid electricity prices The power purchased by the root node from the main grid during time period t. A collection of photovoltaic nodes. For nodes n The marginal cost of photovoltaic power generation For nodes n During the period t Photovoltaic power output, It is a collection of energy storage nodes. and They are nodes n The marginal cost of energy storage charging and discharging, and They are nodes n During the period t The energy storage charging and discharging power; The active power balance constraint of the node is: In the formula, For nodes n During the period t The normal load power, For nodes n During the period t The combined power of the electric freight fleet The dual variable of this constraint is the marginal electricity price at the distribution network node; The formula for calculating the aggregate power of the electric freight fleet is as follows: In the formula, Represents traffic network nodes i Mapped to the distribution network bus n ; The relationship between the electricity price at transportation network nodes and the marginal electricity price at distribution network nodes is defined as follows: The lower-level distribution network market clearing model employs linearized distribution network power flow (LinDistFlow) constraints, including: Active power recursion constraint: Reactive power recursion constraint: Voltage drop equation: In the formula, pa ( n ) represents a node n The parent node, ch ( n () represents the set of child nodes of node n. and They are respectively the lines ( pa ( n ) ,n During the time period t Active power and reactive power, and Let n be the resistance and reactance of the line (pa(n), n), respectively. For nodes n During the period t The square of the voltage amplitude; Voltage amplitude constraint: Line capacity constraints: .

[0016] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the processing procedure obtained after optimizing the two-layer to single-layer transformation using KTT conditions, as follows: The lower-level distribution network market clearing problem is replaced with its Karush-Kuhn-Tucker optimality conditions, including stationary conditions for each lower-level decision variable, complementary relaxation conditions for each inequality constraint, and nonnegativity conditions for each dual variable. The Big M method is used to linearize the complementary relaxation conditions, transforming the complementary constraints into mixed integer linear constraints. The strong duality theorem is used to eliminate bilinear terms in the objective function involving the product of upper and lower level variables.

[0017] In conjunction with the first aspect, in some implementations of the first aspect, the method further includes: the method of linearizing the complementary relaxation conditions using the Big M method is as follows: For the form of Complementary conditions, introducing binary auxiliary variables And establish the following linear constraints: In the formula, M is a sufficiently large constant; when z=1, the constraint forces s=0 and Positive values ​​are allowed; when z=0, the constraint is enforced. =0 and s can take positive values; thus realizing the equivalent linearized representation of the complementary condition; The method for eliminating bilinear terms using strong duality is as follows: Using the strong duality theorem for linear programming problems, the optimal value of the primal problem is equal to the optimal value of the dual problem: In the formula, The optimal objective value for the underlying primal problem. In addition to the dual objective The sum of all other dual terms except the term.

[0018] Secondly, in order to achieve the above objectives, this invention discloses a distribution network collaborative dispatching system for electric freight fleets based on nodal pricing, comprising: The data receiving module is used to receive input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines; The node mapping module is used to map the set of rechargeable nodes to the set of bus nodes based on a pre-built mapping function to obtain the node set mapping relationship. The set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes, and the set of customer nodes with charging capabilities. The collaborative scheduling module is used to input the data to be input and the mapping relationship of the node set into a pre-established collaborative optimization scheduling model, and output the optimal collaborative scheduling scheme. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model, and is obtained after optimization from a two-level to a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

[0019] In another aspect of the present invention, in order to achieve the above-mentioned objective, a terminal device is disclosed, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The memory stores the computer program capable of running on the processor. When the processor loads and executes the computer program, it employs the above-described method for coordinated dispatching of electric freight fleet distribution networks based on nodal pricing.

[0020] The beneficial effects of this invention are: This invention explicitly models the distribution network market clearing process through a two-layer optimization framework, ensuring that the marginal electricity price at each node is endogenously determined by the system's operating state, rather than by exogenous assumptions. This mechanism accurately reflects the feedback impact of electric freight fleet charging and discharging behavior on electricity prices, preventing the fleet scheduling plan from becoming disconnected from the actual operating state of the power grid. The spatiotemporal price response charging constraint of this invention precisely correlates continuous time with discrete electricity price periods, enabling the fleet to schedule charging and discharging decisions based on the actual marginal electricity price at each node and time period. This achieves time arbitrage of charging at low prices and discharging at high prices, as well as spatial arbitrage of charging at low-price nodes and discharging at high-price nodes. The lower-layer model of this invention fully considers the power flow constraints, voltage constraints, and line capacity constraints of the distribution network, ensuring the feasibility of the fleet scheduling plan on the grid side. When the fleet's charging demand may lead to network congestion or voltage exceedances, the marginal electricity price at the corresponding node will increase significantly, guiding the fleet to adjust its charging behavior from an economic signal perspective. This invention integrates the three sub-problems of traditional phased solutions into a unified optimization framework, fully considering the mutual influence between path selection, cargo loading, and charging scheduling to obtain a globally optimal or near-optimal scheduling scheme. Based on KKT conditions and strong duality, this invention transforms the bi-level optimization problem into a single-level MILP problem, which can be directly processed by commercial solvers, avoiding the iterative complexity of bi-level problems. This significantly reduces the model solution time and computational complexity, enabling the collaborative scheduling method to maintain good scalability and engineering feasibility even when the number of vehicles, node scale, and time dimension expand. It overcomes the shortcomings of existing methods in solving problems in a timely manner in large-scale scenarios. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a schematic diagram of the method flow of the present invention; Figure 2 This is a schematic diagram of the nodes of the transportation network of the present invention; Figure 3 This is a schematic diagram of the test results of an embodiment of the present invention; Figure 4This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Example 1: like Figure 1 As shown, a distribution network collaborative scheduling method for electric freight fleets based on nodal pricing is presented. The method includes the following steps: S101: Receive input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines; Specifically, the transportation network is defined as a directed graph. ,in V For a set of nodes, A Let V be the set of arc segments. The node set V contains three basic types of nodes: Station Node Set The depot is the departure and return point for vehicles, as well as the loading point for goods and the deployment point for charging facilities; Client node set The customer is the destination that requires delivery services, and each customer has a defined demand and service time window; Independent charging station node collection Independent charging stations are public facilities that specifically provide charging services.

[0024] Based on the charging infrastructure configuration, the customer node set is further divided into: the customer node set with charging capabilities. These customer locations are equipped with charging facilities, allowing vehicles to charge simultaneously during service; customer nodes without charging capabilities... These customer locations do not have charging facilities. The set of available charging nodes is defined as: Define the power distribution network as a radial diagram. ,in This is the set of busbar nodes, with node 0 being the root node (low-voltage side busbar of the substation). EThis is a collection of power distribution lines. The power distribution network adopts a radial topology, and each node (except the root node) has one and only one parent node.

[0025] The power distribution network includes the following types of equipment and loads: Conventional loads: residential, commercial, and industrial loads distributed on each bus node; Photovoltaic power generation: distributed photovoltaic power generation systems deployed on some bus nodes; Energy storage systems: battery energy storage systems deployed on some bus nodes; Electric freight fleet aggregated charging / discharging loads: aggregated from the charging and discharging power of vehicles on the traffic network charging nodes mapped to this bus.

[0026] S102: Based on a pre-built mapping function, the set of rechargeable nodes is mapped to the set of bus nodes to obtain the node set mapping relationship, wherein the set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes, and the set of customer nodes with charging capabilities; The set of customer nodes with charging capabilities is divided based on the configuration of charging facilities. The set of customer nodes equipped with charging facilities is classified as the set of customer nodes with charging capabilities, and the set of customer nodes without charging facilities is classified as the set of customer nodes without charging capabilities.

[0027] Establish mapping function To connect every charging node in the transportation network Mapped to the corresponding bus node in the distribution network ,in This represents a subset of bus nodes that carry the load of the electric freight fleet. The mapping function implements the topological association between the traffic layer and the power layer.

[0028] S103: Input the data to be input and the mapping relationship of the node set into the pre-established collaborative optimization scheduling model, and output the optimal collaborative scheduling scheme. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model, and is obtained after optimization from a two-level to a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

[0029] The upper-level electric freight fleet route loading and charging coordination model defines route decision variables including arc selection binary variables, vehicle activation binary variables, and node access binary variables; time decision variables including node arrival time and node departure time; loading decision variables including cargo load at node departure; and energy decision variables including battery state of charge at node arrival, battery state of charge at node departure, charging amount at each time period, and discharging amount at each time period. The objective function of the upper-level electric freight fleet route loading and charging coordination model minimizes the total operating cost, including fixed deployment cost, travel distance cost, time operating cost, time window violation penalty cost, and net charging cost. The net charging cost is determined using the marginal electricity price of the distribution network nodes intrinsically determined by the lower-level model. The model constructs route connectivity constraints, time dynamic constraints, cargo loading dynamic constraints, battery state of charge dynamic constraints, and spatiotemporal price response charging constraints. Specifically, the route connectivity constraints ensure that each activated vehicle departs from the starting depot and returns to the ending depot, guaranteeing that each customer node is visited exactly once and maintaining the flow conservation relationship between vehicles and nodes, thus ensuring the feasibility of the route structure. The time dynamic constraints characterize the propagation relationship of vehicle travel time between nodes and the matching relationship between service time and dwell time, and are used to calculate the time window violation amount, thereby ensuring that vehicle scheduling meets the requirements of time continuity and service sequence. The cargo loading dynamic constraints describe the change in cargo load during the process of vehicles visiting each customer node, ensuring… The system ensures that the vehicle's cargo load at any given time does not exceed its rated load capacity and meets the dynamic deduction relationship of customer demand. The dynamic constraint on battery state of charge is used to characterize the energy change pattern of the vehicle during driving, charging, and discharging, ensuring that the battery state of charge is always within the allowable range and reflecting the energy consumption and replenishment process. The spatiotemporal price response charging constraint is used to establish a correlation between the vehicle's continuous dwell time at a charging node and discrete electricity price periods, enabling the vehicle to flexibly arrange charging and discharging amounts according to the marginal electricity price of the distribution network node at each node and time period, thereby achieving economic optimization scheduling based on spatiotemporal electricity price signals.

[0030] Construct a route-loading-charging coordination model for the upper-level electric freight fleet.

[0031] The decision variable system is defined as follows: Path decision variables: :vehicle k Does it pass through an arc segment? i,j binary variables; :vehicle k Whether a binary variable is enabled; :node i Was it by vehicle? k Accessed binary variables. Time-based decision variables: :vehicle k Reaching the node i Time; :vehicle k Leave node i Time. Load decision variables: :vehicle k Leave node i Cargo capacity at any given time. Energy decision variables: :vehicle kReaching the node i The state of charge of the battery at that time; :vehicle k Leave node i The state of charge of the battery at that time; :vehicle k At the node i time period t The amount of charge; :vehicle k At the node i time period t The amount of discharge.

[0032] Construct an objective function to minimize total operating costs: The fixed deployment cost is as follows: In the formula, K represents the set of vehicles. For vehicles k Whether a binary decision variable is enabled or not. This represents the fixed deployment cost coefficient for a single vehicle. The cost of the travel distance is: In the formula, Let be the set of nodes that can serve as the starting point of an arc segment. Let be the set of nodes that can serve as the endpoints of an arc segment. For vehicles k Does it pass through an arc segment? i,j binary decision variables, For arc segment ( i,j The distance, Cost coefficient per unit distance; The time operating cost is: In the formula, For vehicles k Return to arrival time at the terminal station. For vehicles k Time of departure from the starting station This is the operating cost coefficient per unit time. The penalty cost for violating the time window is: In the formula, C is the set of customer nodes. For vehicles k The amount of time window violation at customer C. The penalty coefficient for violations within a unit time window; Net charging cost is: In the formula, Let T be the set of rechargeable nodes and T be the set of time periods. For nodes i During the period t The marginal electricity price at the distribution network node For vehicles k At the node i time period t The amount of charge, For vehicle k at node i time period t The amount of discharge.

[0033] Path connectivity constraints ensure that: each activated vehicle departs from the starting depot and returns to the ending depot; each customer is visited by exactly one vehicle once; intermediate nodes satisfy flow conservation, i.e., inflow equals outflow; and node access variables and arc selection variables maintain consistency.

[0034] Path connectivity constraints include: Vehicle activation and departure constraints ensure that activated vehicles depart from the originating depot: Vehicle return constraint ensures that activated vehicles return to the termination depot: Customer access constraints ensure that each customer is accessed exactly once: Flow conservation constraints ensure that the inflow and outflow of intermediate nodes are balanced. Node access and arc selection association constraints: ; Dynamic time constraints ensure that: arrival times between nodes satisfy the propagation relationship of travel time; the dwell time of customer nodes is not less than the service duration; the dwell time of rechargeable nodes is not less than the larger of the service duration and the charging duration; and violations of time windows are recorded for penalty in the objective function.

[0035] Time-based dynamic constraints include: Arrival time propagation constraints are linearized using the Big M method: In the formula, For vehicles k Reaching the node j Time, For vehicles k Leave node i Time, For arc segment ( i,j The travel time of M is a sufficiently large constant. Service duration constraints: In the formula, Service duration for customer C; Time window constraints and violation calculation: In the formula, l c The upper bound of customer c's time window; Service-charging parallel constraints at rechargeable nodes: In the formula, For vehicles k At the node i Charging time; Dynamic constraints on cargo loading ensure that: the cargo load does not exceed the vehicle's load capacity at any given time; the cargo load at customer nodes decreases according to demand; and the cargo load at terminal nodes is reset to full load.

[0036] Battery SOC dynamic constraints ensure that the SOC remains within the allowable range at any given time. Internal; SOC is set to the initial value when departing from the starting station. E 0 Energy consumption during arc driving is calculated proportionally to distance; SOC updates for rechargeable nodes take into account charging and discharging efficiency.

[0037] The spatiotemporal price response charging constraint discretizes continuous dwell time intervals into time period sequences, enabling vehicles to flexibly arrange charging and discharging amounts according to the electricity price of each time period.

[0038] Spatiotemporal price response charging constraints include: Discretize the planning time domain into There are several equal-length time intervals, each with a length of [length missing]. ; Introducing binary indicator variables and Depicting the charging time window, among which In the vehicle k Reaching the node i The corresponding time period and afterwards is 1. In the vehicle k Leave node i The corresponding time period and afterwards is 1; Establish monotonicity constraints for indicator variables: Establish the correlation constraint between continuous charging time and discrete time intervals: Establish upper bound constraints for time-segmented charging and discharging quantities: In the formula, For nodes i The rated power of the charging pile; Establish charging and discharging mutual exclusion constraints: The lower-level distribution network market clearing model adopts the linearized distribution network power flow model LinDistFlow. Optional distribution network modeling methods also include AC power flow models, second-order cone relaxation models (SOCP), and DC power flow models. While AC power flow models and second-order cone relaxation models offer high accuracy, they contain nonlinearities or cone constraints, making it difficult to achieve unified linear modeling with the upper-level path-load-charge-discharge coordination model, which contains a large number of integer decision variables. Although the DC power flow model has a simple structure, it ignores reactive power and voltage amplitude constraints, making it difficult to reflect the impact of voltage constraints on nodal electricity prices in low-voltage distribution networks. This invention selects the LinDistFlow linearized model, which, while maintaining key physical characteristics such as nodal power balance, voltage constraints, and line capacity constraints, can achieve a unified mixed-integer linear programming solution with the upper-level model through KKT conditions and the Big M method, thus balancing modeling accuracy and computational efficiency.

[0039] The objective function of the lower-level model is to minimize the total system operating cost. The selection of cost terms follows the principle of "reflecting the true marginal cost of the system and forming a reasonable price signal," and includes: grid purchase cost, used to characterize the marginal cost of purchasing electricity from the upper-level grid; photovoltaic power generation cost, used to reflect the operation and maintenance and opportunity costs of distributed photovoltaic output, enabling the system to prioritize the absorption of low-marginal-cost renewable energy; and energy storage operation cost, used to characterize the efficiency loss and lifespan reduction during the charging and discharging process of energy storage, avoiding excessive and frequent use of energy storage. Through the joint modeling of the above cost terms, the resulting nodal marginal electricity price can simultaneously reflect the differences in energy source structure and the system regulation costs.

[0040] The model constructs node active power balance constraints and node reactive power balance constraints to ensure power supply and demand balance at each bus node and serves as the basis for the formation of node marginal electricity prices. It also constructs active power recursive constraints and reactive power recursive constraints to characterize the transmission relationship of line power in a radial distribution network; a voltage drop equation to reflect the impact of line impedance on node voltage levels; photovoltaic output constraints to limit photovoltaic power generation from exceeding its available output range; energy storage charging and discharging power constraints and energy storage state of charge constraints to ensure the energy storage system operates within power and energy boundaries; voltage amplitude constraints to prevent node voltage from exceeding upper and lower limits and ensure power supply quality; and line capacity constraints to limit the operation of distribution lines within a safe thermal stability range. The dual variable of the node active power balance constraints is defined as the distribution network node marginal electricity price. The objective function of the lower-level distribution network market clearing model is as follows: In the formula, Main grid electricity prices The power purchased by the root node from the main grid during time period t. A collection of photovoltaic nodes. For nodes n The marginal cost of photovoltaic power generation For nodes n During the period t Photovoltaic power output, It is a collection of energy storage nodes. and They are nodes n The marginal cost of energy storage charging and discharging, and They are nodes n During the period t The energy storage charging and discharging power; The active power balance constraint of the node is: In the formula, For nodes n During the period t The normal load power, For nodes n During the period t The combined power of the electric freight fleet The dual variable of this constraint is the marginal electricity price at the distribution network node; The formula for calculating the aggregate power of the electric freight fleet is as follows: In the formula, Represents traffic network nodes i Mapped to the distribution network busn ; The relationship between the electricity price at transportation network nodes and the marginal electricity price at distribution network nodes is defined as follows: The lower-level distribution network market clearing model employs linearized distribution network power flow (LinDistFlow) constraints, including: Active power recursion constraint: Reactive power recursion constraint: Voltage drop equation: In the formula, pa ( n ) represents a node n The parent node, ch ( n () represents the set of child nodes of node n. and They are respectively the lines ( pa ( n ) ,n During the time period t Active power and reactive power, and Let n be the resistance and reactance of the line (pa(n), n), respectively. For nodes n During the period t The square of the voltage amplitude; Voltage amplitude constraint: Line capacity constraints: ; The processing procedure obtained after optimizing the two-layer to single-layer model using KKT conditions is as follows: KKT conditions are Karush–Kuhn–Tucker optimality conditions, which are used to characterize the stationary point conditions, original feasibility conditions, dual feasibility conditions, and complementary relaxation conditions that the optimal solution of a convex optimization problem must satisfy. This invention uses these conditions to equivalently replace the market clearing problem of the lower-level distribution network with its optimality constraints, thereby transforming the original two-layer collaborative optimization model into a single-layer constrained optimization model.

[0041] The lower-level distribution network market clearing problem is a convex linear programming problem with lower-level decision variables as optimization variables and upper-level aggregated power as parameters. According to convex optimization theory, the KKT conditions are a necessary and sufficient condition for the optimality of a convex problem. Replacing the lower-level problem with its KKT conditions includes: Stationary point condition: The gradient of the objective function equals a linear combination of the gradients of the constraint functions (with coefficients as dual variables); Primitive feasibility: All primal constraints are satisfied; Dual feasibility: The dual variables of the inequality constraints are non-negative; Complementary relaxation: The product of the relaxation amount of the inequality constraint and the dual variable is zero. The complementary relaxation condition is of the form: The product of zero constraints is nonlinear and nonconvex. Linearization is achieved using the Big M method: a binary variable z is introduced, and a... and When z=1, s is forced to be 0; when z=0, s is forced to be 0. Equivalent to achieving complementary conditions.

[0042] The method for linearizing complementary relaxation conditions using the Big M method is as follows: For the form of Complementary conditions, introducing binary auxiliary variables And establish the following linear constraints: In the formula, M is a sufficiently large constant; when z=1, the constraint forces s=0 and Positive values ​​are allowed; when z=0, the constraint is enforced. =0 and s can take positive values; thus realizing the equivalent linearized representation of the complementary condition; The method for eliminating bilinear terms using strong duality is as follows: Using the strong duality theorem for linear programming problems, the optimal value of the primal problem is equal to the optimal value of the dual problem: In the formula, The optimal objective value for the underlying primal problem. In addition to the dual objective The sum of all other dual terms except the term.

[0043] The net charging cost term in the upper objective function includes Bilinear terms of the form where electricity price It is the lower-level dual variable, the amount of charge. These are the upper-level decision variables. Utilizing strong duality, the optimal value of the lower-level primal problem is equal to the optimal value of the dual problem. Through algebraic transformations, the bilinear terms can be converted into a linear expression involving only the lower-level primal variables.

[0044] For large-scale problems, a decomposition strategy can be adopted: first, promising path structures are enumerated or searched using heuristic algorithms; then, for each given path structure, a path-fixed subproblem is solved. After the path structure is determined by the path-fixed subproblem, the remaining decision variables include time variables, energy variables, and lower-level primal-dual variables. This subproblem is a mixed-integer linear programming problem (binary variables come from complementary relaxation linearization), which can be efficiently solved by commercial solvers such as Gurobi and CPLEX.

[0045] The optimal scheduling scheme output includes: the access node sequence and arrival / departure time for each vehicle; the service vehicle and delivery time for each customer; the cargo loading change trajectory for each node; the battery SOC change trajectory for each vehicle; the time-of-use charging / discharging power of each vehicle at each rechargeable node; and the time-of-use marginal electricity price for each distribution network bus node.

[0046] Specifically, the present invention will be further illustrated below through embodiments: Consider a 4×4 grid transportation network, including 2 depot nodes, 12 customer nodes, and 2 independent charging station nodes, such as... Figure 2 The power distribution network uses a modified IEEE 33-bus test system, consisting of 33 bus nodes and 32 distribution lines. The fleet comprises 6 electric freight trucks, each with a 200kWh battery capacity and a payload capacity of 5 tons. The planned time domain is 24 hours, discretely divided into 96 15-minute time intervals.

[0047] Test results are as follows Figure 3 As shown: Compared with the benchmark method using a fixed electricity price, the charging cost of the method of the present invention is reduced; the charging behavior of the fleet exhibits obvious time-period clustering characteristics, mainly concentrated in the low electricity price period (early morning and noon when photovoltaic output is high); V2G discharge behavior mainly occurs during the peak electricity price period (morning and evening load peak periods); the voltage of the distribution network nodes and the power flow of the lines are both kept within a safe range.

[0048] Example 2: To achieve the above objective, based on Example 1, as follows... Figure 4 As shown, this invention discloses a distribution network collaborative dispatching system for electric freight fleets based on nodal pricing, comprising: The data receiving module 11 is used to receive input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines; The node mapping module 12 is used to map the set of rechargeable nodes to the set of bus nodes based on a pre-built mapping function to obtain the node set mapping relationship, wherein the set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes and the set of customer nodes with charging capabilities; The collaborative scheduling module 13 is used to input the data to be input and the mapping relationship of the node set into the pre-established collaborative optimization scheduling model, and output the optimal collaborative scheduling scheme. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model, and is obtained after optimization from a two-level to a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

[0049] Based on the same inventive concept, this invention also provides a computer device, comprising: one or more processors, and a memory for storing one or more computer programs; the programs include program instructions, and the processor executes the program instructions stored in the memory. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, used to implement one or more instructions, specifically for loading and executing one or more instructions stored in a computer storage medium to implement the above-described method.

[0050] It should be further explained that, based on the same inventive concept, the present invention also provides a computer storage medium storing a computer program, which, when executed by a processor, performs the above-described method. This storage medium can be any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0051] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0052] The foregoing has shown and described the basic principles, main features, and advantages of this disclosure. Those skilled in the art should understand that this disclosure is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of this disclosure. Various changes and modifications can be made to this disclosure without departing from its spirit and scope, and all such changes and modifications fall within the scope of this disclosure as claimed.

Claims

1. A method for coordinated dispatching of electric freight fleets based on nodal pricing distribution networks, characterized in that, The method includes the following steps: The system receives input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines. Based on a pre-built mapping function, the set of rechargeable nodes is mapped to the set of bus nodes to obtain the node set mapping relationship. The set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes, and the set of customer nodes with charging capabilities. The input data and the mapping relationship of the node set are input into the pre-established collaborative optimization scheduling model, and the optimal collaborative scheduling scheme is output. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model. It is obtained by optimizing the two-level model into a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

2. The method for coordinated dispatching of electric freight fleets based on nodal pricing in the distribution network according to claim 1, characterized in that, The set of customer nodes with charging capabilities is divided based on the configuration of charging facilities. Customer nodes equipped with charging facilities are classified as the set of customer nodes with charging capabilities, while customer nodes without charging facilities are classified as the set of customer nodes without charging capabilities.

3. The method for coordinated dispatching of electric freight fleets based on nodal pricing in power distribution networks according to claim 1, characterized in that, The upper-level electric freight fleet route loading and charging coordination model defines route decision variables including arc selection binary variables, vehicle activation binary variables, and node access binary variables; defines time decision variables including node arrival time and node departure time; defines loading decision variables including cargo load at node departure; and defines energy decision variables including battery state of charge at node arrival, battery state of charge at node departure, charging amount at each time period, and discharging amount at each time period. The objective function of the upper-level electric freight fleet route loading and charging coordination model minimizes the total operating cost, including fixed deployment cost, travel distance cost, time operating cost, time window violation penalty cost, and net charging cost. It also constructs route connectivity constraints, time dynamic constraints, cargo loading dynamic constraints, battery state of charge dynamic constraints, and spatiotemporal price response charging constraints.

4. The method for coordinated dispatching of electric freight fleets based on nodal pricing in accordance with claim 3, characterized in that, The objective function of the upper-level electric freight fleet route loading and charging coordination model is as follows: The fixed deployment cost is as follows: In the formula, K represents the set of vehicles. For vehicles k Whether a binary decision variable is enabled or not. This represents the fixed deployment cost coefficient for a single vehicle. The cost of the travel distance is: In the formula, Let be the set of nodes that can serve as the starting point of an arc segment. Let be the set of nodes that can serve as the endpoints of an arc segment. For vehicles k Does it pass through an arc segment? i,j binary decision variables, For arc segment ( i,j The distance, Cost coefficient per unit distance; The time operating cost is: In the formula, For vehicles k Return to arrival time at the terminal station. For vehicles k Time of departure from the starting station This is the operating cost coefficient per unit time. The penalty cost for violating the time window is: In the formula, C is the set of customer nodes. For vehicles k The amount of time window violation at customer C. The penalty coefficient for violations within a unit time window; Net charging cost is: In the formula, Let T be the set of rechargeable nodes and T be the set of time periods. For nodes i During the period t The marginal electricity price at the distribution network node For vehicles k At the node i time period t The amount of charge, For vehicle k at node i time period t The amount of discharge.

5. The method for coordinated dispatching of electric freight fleets based on nodal pricing in a distribution network according to claim 1, characterized in that, The lower-level distribution network market clearing model adopts the linearized distribution network power flow model LinDistFlow. The objective function is to minimize the system operating cost, including the main grid power purchase cost, photovoltaic power generation cost, and energy storage operating cost. The model constructs node active power balance constraints, node reactive power balance constraints, active power recursive constraints, reactive power recursive constraints, voltage drop equations, photovoltaic output constraints, energy storage charging and discharging power constraints, energy storage state of charge constraints, voltage amplitude constraints, and line capacity constraints. The dual variable of the node active power balance constraints is defined as the marginal electricity price of the distribution network nodes.

6. The method for coordinated dispatching of electric freight fleets based on nodal pricing in power distribution networks according to claim 5, characterized in that, The objective function of the lower-level distribution network market clearing model is as follows: In the formula, Main grid electricity prices The power purchased by the root node from the main grid during time period t. A collection of photovoltaic nodes. For nodes n The marginal cost of photovoltaic power generation For nodes n During the period t Photovoltaic power output, It is a collection of energy storage nodes. and They are nodes n The marginal cost of energy storage charging and discharging, and They are nodes n During the period t The energy storage charging and discharging power; The active power balance constraint of the node is: In the formula, For nodes n During the period t The normal load power, For nodes n During the period t The combined power of the electric freight fleet The dual variable of this constraint is the marginal electricity price at the distribution network node; The formula for calculating the aggregate power of the electric freight fleet is as follows: In the formula, Represents traffic network nodes i Mapped to the distribution network bus n ; The relationship between the electricity price at transportation network nodes and the marginal electricity price at distribution network nodes is defined as follows: The lower-level distribution network market clearing model employs linearized distribution network power flow (LinDistFlow) constraints, including: Active power recursion constraint: Reactive power recursion constraint: Voltage drop equation: In the formula, pa ( n ) represents a node n The parent node, ch ( n () represents the set of child nodes of node n. and They are respectively the lines ( pa ( n ) ,n During the time period t Active power and reactive power, and Let n be the resistance and reactance of the line (pa(n), n), respectively. For nodes n During the period t The square of the voltage amplitude; Voltage amplitude constraint: Line capacity constraints: 。 7. The method for coordinated dispatching of electric freight fleets based on nodal pricing in a distribution network according to claim 1, characterized in that, The optimization process obtained after converting a two-layer to a single-layer structure using the KTT condition is as follows: The lower-level distribution network market clearing problem is replaced with its Karush-Kuhn-Tucker optimality conditions, including stationary conditions for each lower-level decision variable, complementary relaxation conditions for each inequality constraint, and nonnegativity conditions for each dual variable. The Big M method is used to linearize the complementary relaxation conditions, transforming the complementary constraints into mixed integer linear constraints. The strong duality theorem is used to eliminate bilinear terms in the objective function involving the product of upper and lower level variables.

8. The method for coordinated dispatching of electric freight fleets based on nodal pricing in a distribution network according to claim 7, characterized in that, The method for linearizing complementary relaxation conditions using the Big M method is as follows: For the form of Complementary conditions, introducing binary auxiliary variables And establish the following linear constraints: In the formula, M is a sufficiently large constant; when z=1, the constraint forces s=0 and Positive values ​​are allowed. When z=0, the constraint is enforced. =0 and s can take positive values; thus realizing the equivalent linearized representation of the complementary condition; The method for eliminating bilinear terms using strong duality is as follows: Using the strong duality theorem for linear programming problems, the optimal value of the primal problem is equal to the optimal value of the dual problem: In the formula, The optimal objective value for the underlying primal problem. In addition to the dual objective The sum of all other dual terms except the term.

9. A distribution network collaborative dispatching system for electric freight fleets based on nodal pricing, employing the distribution network collaborative dispatching method for electric freight fleets based on nodal pricing as described in any one of claims 1 to 8, characterized in that... include: The data receiving module is used to receive input data, which includes: traffic network data, distribution network data, fleet operation parameters, distribution network operation parameters, and main grid electricity purchase price parameters; wherein, the traffic network data includes a set of depot nodes, a set of customer nodes, and a set of independent charging station nodes; the distribution network data includes a set of bus nodes and a set of distribution lines; The node mapping module is used to map the set of rechargeable nodes to the set of bus nodes based on a pre-built mapping function to obtain the node set mapping relationship. The set of rechargeable nodes is the union of the set of station nodes, the set of independent charging station nodes, and the set of customer nodes with charging capabilities. The collaborative scheduling module is used to input the data to be input and the mapping relationship of the node set into a pre-established collaborative optimization scheduling model, and output the optimal collaborative scheduling scheme. The pre-established collaborative optimization scheduling model includes an upper-level electric freight vehicle fleet route loading and charging coordination model and a lower-level distribution network market clearing model, and is obtained after optimization from a two-level to a single-level model through KTT conditions. The optimal collaborative scheduling scheme is the optimal fleet route loading charging and discharging scheduling scheme and the marginal electricity price of the distribution network node.

10. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, The memory stores a computer program that can run on a processor. When the processor loads and executes the computer program, it employs the electric freight fleet distribution network collaborative scheduling method based on nodal pricing, as described in any one of claims 1 to 8.

Images