Electric vehicle freight planning system based on charging and battery replacement hybrid strategy

Abstract

The invention discloses an electric vehicle freight planning system based on a charging and battery replacement hybrid strategy. The electric vehicle freight planning system comprises a network construction module, an optimization model module and an algorithm solving module. The network construction module conducts modeling through a space-time convolution network, time is dispersed into a plurality of time points, space is defined as freight station nodes, and a charging arc, a service arc and a convolution arc are constructed. And the optimization model module establishes a mixed integer programming model based on the network, wherein a target function minimizes the total transportation cost, and constraint conditions comprise commodity flow balance, transport capacity limitation, purchase budget and power conversion capacity constraint. The algorithm solving module adopts a column generation algorithm to generate a path ring, and solves a sub-problem in combination with multi-dimensional dynamic programming; the dynamic planning transmits and updates the electric quantity state through hierarchical labels, and applies a dominating rule to screen a non-inferior solution path. The system intelligently triggers battery replacement or charging according to the electric quantity state, optimizes an electric truck energy complementing strategy and a freight path, improves the transportation efficiency, reduces the energy consumption cost, reduces the unloaded driving time, and improves the utilization rate of a motorcade.

Description

Technical Field

[0001] The present invention relates to the field of freight planning technology, and specifically to an electric vehicle freight planning system based on a hybrid charging and battery replacement strategy, which is used to optimize the energy replenishment decisions and route planning of electric vehicles in urban freight, improve transportation efficiency and reduce operating costs. Background Art

[0002] With the rapid development of smart logistics systems, electric vehicles, as an important carrier of green transportation, are gradually gaining widespread attention in medium- and short-distance urban distribution and trunk logistics. Compared with traditional fuel vehicles, electric vehicles have significant advantages in energy consumption, carbon emissions and maintenance costs, but their promotion is limited by three core issues:

[0003] 1. Limited battery life: The range on a single charge is insufficient to meet the needs of long-distance freight transport, and frequent recharging can lead to transport interruptions.

[0004] 2. Inefficient recharging: Charging is time-consuming (fast charging still takes tens of minutes or even longer), limiting vehicle operational efficiency. Battery swapping relies on a high-density network of battery swap stations. A single model struggles to balance operational efficiency and infrastructure costs, especially in diverse and dynamic urban logistics scenarios, where adaptability is significantly limited.

[0005] 3. High scheduling complexity: The existing logistics system is designed based on fuel vehicles and lacks a collaborative optimization mechanism for the state of charge (SOC) of electric trucks, the selection of energy replenishment methods (charging / battery replacement), and the feasibility of routes.

[0006] In the existing technology, Comparative Document 1 (CN117391564A) proposes a new energy logistics vehicle recharging scheduling model, dynamically selecting the charging and battery swapping method based on the SOC state, but does not solve the problems of periodic scheduling and closed-loop path planning. Comparative Document 2 (CN114819327A) uses a high-dimensional spatiotemporal network to model electric vehicle routes, introducing the dimension of battery state, but its network structure lacks a head-to-tail closed-loop design and does not incorporate a dynamic decision-making mechanism for hybrid recharging strategies. Therefore, there is an urgent need for a freight planning system that integrates a hybrid charging and battery swapping strategy, supports closed-loop scheduling, and provides efficient solutions. Summary of the Invention

[0007] (1) Technical problems solved

[0008] To solve the above problems, the present invention provides an electric vehicle freight planning system based on a hybrid charging and battery replacement strategy. The energy replenishment strategy is more flexible, the scheduling cycle can form a closed loop and can be solved efficiently.

[0009] (2) Technical solution

[0010] To achieve the above object, the present invention provides the following technical solutions:

[0011] An electric vehicle freight planning system based on a hybrid charging and battery swapping strategy, comprising:

[0012] A network construction module is used to construct a space-time convolutional network, where the time dimension is discretized into multiple time points and the space dimension includes multiple freight station nodes. The network includes charging arcs, service arcs, and convolution arcs, and the convolution arcs connect the first and last nodes of the scheduling cycle to form a closed loop path;

[0013] An optimization model module, based on the space-time convolutional network, defines a mixed integer programming model, including an objective function and constraints, wherein the objective function minimizes the total transportation cost, and the constraints include a commodity flow balance constraint, a vehicle capacity constraint, a procurement budget constraint, and a battery swap capacity constraint;

[0014] An algorithm solving module is used to solve the mixed integer programming model, using a column generation algorithm to generate a vehicle path loop and combining it with multidimensional dynamic programming to solve subproblems, where the multidimensional dynamic programming includes a power state update mechanism and a dominance rule to screen non-inferior solution paths.

[0015] Preferably, the charging arc in the space-time loop network is a horizontal arc, connecting consecutive time nodes at the same site, indicating that the electric vehicle is waiting or charging; the service arc is a diagonal arc, describing the space-time changes of the vehicle during transportation; the loop arc is a dotted arc, ensuring that the vehicle returns to the starting site, forming a closed-loop scheduling cycle.

[0016] Preferably, the parameters of the mixed integer programming model include:

[0017] Electric vehicle collection A set of discrete time points All space-time nodes Product category collection Path Ring Collection The set of all arcs The service arc set Charging Arc Collection Cloth Arc Collection

[0018] Path ring indicator (Θ s,(i,j) ), when the path ring s contains arc (i, j), the path ring indicator (Θ s,(i,j) ) is 1, when the path ring s does not contain arc (i, j), the path ring indicator (Θ s,(i,j) ) is 0;

[0019] The parameters of product k include the starting point ( k )、End point (d k ), total amount (w k );

[0020] Cost parameters include arc cost (c v,(i,j) ), commodity traffic cost (f k,(i,j) ), vehicle purchase price (b v );

[0021] Constraint parameters include the maximum vehicle load (u v ), total procurement budget (B), total battery replacement capacity (H).

[0022] Preferably, the decision variables include:

[0023] Integer variable (θ v,s ), represents the number of vehicle models v that select path ring s;

[0024] Continuous variables (y k,(i,j) ), represents the flow of goods on arc (i, j);

[0025] In the multidimensional dynamic programming, the sub-problem decision variables include:

[0026] Arc selection variable (x i,j ), when the electric vehicle selects arc (i, j), the variable value is 1, and when the electric vehicle does not select arc (i, j), the variable value is 0;

[0027] Power state variable (s i ), represents the amount of electricity of the electric vehicle at time and space point i;

[0028] Battery replacement indicator variable (q i ), where q i When it is 1, it means that the battery is replaced at the node and the battery is reset to full charge. i When it is 0, it means that no battery replacement is performed at the node.

[0029] Preferably, the objective function is expressed as:

[0030]

[0031] Among them, c v,s is the cost of vehicle type v passing through path cycle s; the constraints include:

[0032] Traffic balance constraints:

[0033] Capacity constraints:

[0034] Procurement budget constraints:

[0035] Battery swap capacity constraints: Among them, μ v,sis the number of battery replacements for vehicle type v on path loop s;

[0036] Decision variable range constraints: y≥0.

[0037] Preferably, the column generation algorithm includes:

[0038] The main problem is a linear relaxation of the mixed integer programming model, using dual variables (α, β, ψ) to handle constraints;

[0039] The sub-problem is to solve the path loop cost for each vehicle model. The sub-problem objective function is:

[0040]

[0041] Constraints include network flow balancing constraints: And the service cycle ends and returns to the starting point;

[0042] State of charge constraints: Among them, p v,(i,j) is the change in charge of electric vehicle type v caused by arc (i, j);

[0043] Other constraints: s≥0.

[0044] Preferably, the multidimensional dynamic programming includes:

[0045] Hierarchical structure processing, passing labels layer by layer from the starting point, the label format is Used to record path information, where n is a time and space node and aw n is the cumulative number of battery replacements, ac n To reduce the cumulative cost, ap n is the cumulative power consumption (1-s n ), is the predecessor node label;

[0046] Dominance rule: For two labels of node n and If there are n intermediate nodes and or n is the terminal node and but Dominate and remove

[0047] Transfer equation update tag: ac j ,ap j ,aw j Based on the arc type and power state calculation, the equation is as follows:

[0048]

[0049] Preferably, the power status update mechanism includes:

[0050] On the service arc, the power change p v,(i,j) is positive or negative, and p v,(i,j) When it is positive, it corresponds to charging, p v,(i,j) When it is a negative value, it corresponds to discharge;

[0051] When the battery replacement indicator variable q i =1, the power of node i is s i Reset to full charge state;

[0052] In dynamic programming, the charge state is updated in real time through transfer equations to handle nonlinear constraints and optimize path feasibility.

[0053] (3) Beneficial effects

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] Improved transportation efficiency: Through a hybrid recharging strategy (dynamic selection of charging / battery swapping), charging or battery swapping can be intelligently selected based on real-time power consumption and site resources. This makes the recharging strategy more flexible and reduces waiting time for electric vehicles.

[0056] Cost-effective collaborative optimization: The closed-loop path design (circle arc) reduces the idle driving rate. Electric vehicles can return to their starting points, reducing the cost and time of idle driving and improving fleet utilization.

[0057] Breakthrough in solution efficiency: The governing rules of multi-dimensional dynamic programming reduce the amount of invalid path calculations, shortening the solution time for 10,000-node networks;

[0058] Strong scalability: The model supports the scheduling of multiple vehicle types and multiple commodities, and can be applied to scenarios such as express delivery routes and urban distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0060] Figure 1 A schematic diagram of the space-time convolutional network structure of the present invention is shown, showing the time dimension, space nodes, and arc types, specifically including the time axis (horizontal axis), space nodes (vertical axis), charging arcs (horizontal solid lines), service arcs (diagonal solid lines), and convolution arcs (closed loops with dashed lines at the beginning and end);

[0061] Figure 2The service arc modeling diagram of the present invention is shown, describing the change of power consumption during transportation;

[0062] Figure 3 The hierarchical dynamic programming structure diagram of the present invention is shown, showing the label transmission path, wherein the arrows mark the label update direction and the nodes mark the control rule screening points. DETAILED DESCRIPTION

[0063] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0064] See attached Figure 1 -Attached Figure 3 The embodiment of the present invention discloses an electric vehicle freight planning system based on a hybrid strategy of charging and battery replacement (see right 1), including: a network construction module for constructing a space-time convolution network, wherein the time dimension is discretized into multiple time points, the space dimension includes multiple freight station nodes, the network includes charging arcs, service arcs and convolution arcs, and the convolution arcs connect the head and tail nodes of the scheduling cycle to form a closed-loop path; an optimization model module, based on the space-time convolution network, defines a mixed integer programming model, including an objective function and constraints, the objective function minimizes the total transportation cost, and the constraints include commodity flow balance constraints, vehicle capacity constraints, procurement budget constraints and battery replacement capacity constraints; an algorithm solving module, for solving the mixed integer programming model, using a column generation algorithm to generate a vehicle path ring, and combining multi-dimensional dynamic programming to solve sub-problems, wherein the multi-dimensional dynamic programming includes a power state update mechanism and a dominance rule to screen non-inferior solution paths.

[0065] Based on the above solution, the system in this application includes three core modules:

[0066] 1. Network building module: building a spatiotemporal convolutional network ( Figure 1 ), the time dimension is discretized into a set The spatial dimension contains the collection of freight stations The network defines three types of arcs:

[0067] Charging arc (horizontal arc): connects consecutive time nodes at the same station, indicating waiting or charging;

[0068] Service arc (diagonal arc): describes the spatiotemporal changes of vehicle transportation across stations ( Figure 2 );

[0069] Roundabout arc (dashed arc): connects the first and last nodes of the scheduling cycle to form a closed loop path ( Figure 1 dotted line).

[0070] 2. Optimization model module: A mixed integer programming model is established based on the network. The objective function is to minimize the total cost (transportation cost + procurement cost). The constraints include commodity flow balance, vehicle capacity, procurement budget, and battery replacement capacity.

[0071] 3. Algorithm solution module: It uses a column generation algorithm (the main problem is the linear relaxation of the model) combined with multi-dimensional dynamic programming (sub-problem solution path loop). Dynamic programming selects non-inferior solution paths through label transfer and dominance rules.

[0072] For the spatiotemporal convolutional network structure, see Appendix Figure 1 The time dimension is represented by discrete time periods, with the vertical axis reflecting geographic location and the horizontal axis representing time progression. These arcs, connected end to end, form a complete scheduling cycle. Horizontal charging arcs (waiting arcs) in the network connect consecutive time nodes at the same station, representing the waiting time of electric trucks. Diagonal service arcs describe the spatiotemporal changes of vehicles during transportation. Dashed loop arcs connect the beginning and end of the cycle, ensuring that vehicles return to their starting station, forming a closed-loop path.

[0073] In order to construct a space-time convolution network, the following design is carried out in this embodiment (see right 2). Specifically, the charging arc in the space-time convolution network is a horizontal arc that connects consecutive time nodes at the same station, indicating that the electric vehicle is waiting or charging; the service arc is a diagonal arc that describes the space-time changes of the vehicle during transportation; the return arc is a dotted arc that ensures that the vehicle returns to the starting station, forming a closed-loop scheduling cycle.

[0074] Based on the above scheme, the construction of the spatiotemporal convolutional network in this application includes the following points:

[0075] Time discretization: The scheduling period (such as 24 hours) is divided into a set of uniform time points (e.g. one node every 15 minutes).

[0076] Spatial nodes: Covering freight sites (warehouses, distribution centers), nodes are defined as (a, t) (site a + time t).

[0077] Service Arc Modeling ( Figure 2 ): Arc (a, t0), (b, t1) represents the transportation from station a to station b. Charging is allowed during the transportation gap (for example, charging to t' after arriving at station b and then departing). Actual power consumption = theoretical power consumption - charging gain. Actual power consumption p v,(i,j) ≥0.

[0078] For the service arc model, see the Figure 2:This figure describes the service arc situation starting from the time-space point, where a and b in the brackets represent the station respectively, and t0 and t1 represent the time point. Each node has dual time and space attributes, and the time required for the service shall not exceed the preset upper limit. If the actual time is less than the upper limit, the electric truck can be charged during transportation. For example, arcs ((a, t0), (a, t')) and ((b, t'), (b, t1)) represent charging at the starting and ending terminals of the service. The driver can adjust the time points t' and t', but the premise is that there is enough battery power to perform the service at the departure time t'. In addition, the time difference t'-t' should be equal to the actual time required for the service. For example, at node (a, t0), if the battery power is sufficient for the actual transportation service, the driver can depart at time t0 = t', and then charge when arriving at b. Therefore, the battery power consumption of the service is equal to the actual battery power consumption minus the charging rate. We assume that the power consumption of each service arc is positive

[0079] In practice, the transportation cycle can be one day or several days, depending on operational needs. Each truck needs to return to the starting station at the end of the cycle to form a path loop. The mathematical model constructed in this study includes integer variables, which are used to determine the size of the mixed fleet and the path loop selection of each vehicle in the space-time network; continuous variables are used to describe the flow of goods and the remaining power of each vehicle. These variables comprehensively characterize the decision-making process of the mixed fleet from procurement to operation; in order to establish a mixed integer programming model, the following design is carried out in this embodiment (see rights 3-5). Specifically, the parameters of the mixed integer programming model include: electric vehicle set A set of discrete time points All space-time nodes Product category collection Path Ring Collection The set of all arcs The service arc set Charging Arc Collection Cloth Arc Collection Path ring indicator (Θ s,(i,j) ), when the path ring s contains arc (i, j), the path ring indicator (Θ s,(i,j) ) is 1, when the path ring s does not contain arc (i, j), the path ring indicator (Θ s,(i,j) ) is 0; the parameters of commodity k include the starting point ( k )、End point (d k ), total amount (w k ); cost parameters include arc cost (c v,(i,j) ), commodity traffic cost (f k,(i,j) ), vehicle purchase price (b v ); Constraint parameters include the maximum vehicle load (u v), total purchase budget (B), total battery replacement capacity (H). Decision variables include: integer variables (θ v,s ), represents the number of models v that choose path loop s; continuous variable (y k,(i,j) ), which represents the flow of goods on arc (i, j); in multidimensional dynamic programming, the sub-problem decision variables include: arc selection variables (x i,j ), when the electric vehicle selects arc (i, j), the variable value is 1, when the electric vehicle does not select arc (i, j), the variable value is 0; the power state variable (s i ), represents the amount of electricity of the electric vehicle at the time and space point i; the battery replacement indicator variable (q i ), where q i When it is 1, it means that the battery is replaced at the node and the battery is reset to full charge. i When it is 0, it means that no battery replacement is performed at the node. The objective function is expressed as: Among them, c v,s is the cost of vehicle type v passing through path ring s; the constraints include: flow balance constraint: Capacity constraints: Procurement budget constraints: Battery swap capacity constraints: (Extended to the whole network battery swap resource coordination), where μ v,s is the number of battery replacements for vehicle type v on path loop s; the range constraints of decision variables are: y≥0.

[0080] The multi-dimensional dynamic programming algorithm is designed as follows in this embodiment (see Sections 6-8). Specifically, the column generation algorithm includes: the main problem is a linear relaxation form of the mixed integer programming model, and the constraints are processed using dual variables (α, β, ψ); the subproblem is to solve the path cycle cost for each vehicle model, and the subproblem objective function is: Constraints include network flow balancing constraints: And the service cycle ends and returns to the origin; power state constraints: Among them, p v,(i,j) is the change in the amount of electricity of electric vehicle type v caused by arc (i, j); other constraints: s≥0. Multidimensional dynamic programming includes: hierarchical structure processing, passing labels layer by layer from the starting point, the label format is Used to record path information, where n is a time and space node and aw n is the cumulative number of battery replacements, ac n To reduce the cumulative cost, ap n is the cumulative power consumption (1-sn ), is the predecessor node label; dominance rule: for the two labels of node n and If there are n intermediate nodes and or n is the terminal node and but Dominate and remove Transfer equation update tag: ac j ,ap j ,aw j Based on the arc type and power state calculation, the equation is as follows:

[0081]

[0082] The power status update mechanism includes: on the service arc, the power change p v,(i,j) is positive or negative, and p v,(i,j) When it is positive, it corresponds to charging, p v,(i,j) When it is negative, it corresponds to discharge; when the battery replacement indicator variable q i =1, the power of node i is s i Reset to full charge state, this mechanism is coupled with label transfer; in dynamic programming, the power state is updated in real time through the transfer equation to handle nonlinear constraints and optimize path feasibility.

[0083] For hierarchical dynamic programming structures, see Appendix Figure 3 :For space-time nodes The label format is defined as Used to record path information. In this tag, Refers to the label of the previous node passed in. ac n and ap n Mark the cumulative reduction cost and power consumption (1-s n ),aw n Indicates the cumulative number of battery swaps to point n. In the process of the labeling algorithm, each node may generate multiple labels. These labels need to be compared according to the dominance rules. Each node only retains the labels that are not dominated.

[0084] For the transfer equation: First we define the label of the starting point Then start from the second layer, point by point, and follow the ac j to ap j Then to aw jMark the nodes in the order they are marked until the endpoint is reached. Since there are multiple starting points, each one generates a different path cycle. Therefore, each computation of the subproblem generates a path cycle for each service station, resulting in multiple path cycles. If the generated paths and path cycles reduce the cost, they are added to the linearly relaxed main problem. Otherwise, the generated paths and path cycles are used to solve the main problem using integer programming.

[0085] In actual application, the system in this application can be implemented through the following steps:

[0086] 1. System initialization:

[0087] Input parameters: freight station coordinates, vehicle model list (load u v , purchase price b v ), commodity demand w k , time discrete granularity.

[0088] 2. Network construction steps:

[0089] Generate space-time node set sites

[0090] Adding a charging arc

[0091] Adding a Service Arc Calculate p v,(i,j) ;

[0092] Add Clothback Arc

[0093] 3. Algorithm solution example:

[0094] Column generation iteration:

[0095] Step 1: Solve the master problem (relaxed model) and obtain the dual variables (α, β, ψ);

[0096] Step 2: For each vehicle type v, dynamically program to solve the minimum cost path cycle;

[0097] Step 3: If the ring cost is less than 0, add it to And update the main question.

[0098] Dynamic programming example:

[0099] Tags i Passed to node j.

[0100] 4. Draw conclusions:

[0101] By adopting the system in this application, the empty driving rate and transportation costs can be reduced, the solution time (10,000 nodes) can be reduced, and the accuracy of battery swapping decisions and the utilization rate of battery swapping stations can be improved (this conclusion is drawn from the overall content of the specific implementation method section).

[0102] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0103] In addition, it should be noted that the use of terms such as "first" and "second" to limit components is only for the convenience of distinguishing the corresponding components. Unless otherwise stated, the above terms have no special meaning and therefore cannot be understood as limiting the scope of protection of this application.

[0104] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. An electric vehicle freight planning system based on a hybrid charging and battery swapping strategy, characterized in that: include: A network construction module is used to construct a space-time convolutional network, where the time dimension is discretized into multiple time points and the space dimension includes multiple freight station nodes. The network includes charging arcs, service arcs, and convolution arcs, and the convolution arcs connect the first and last nodes of the scheduling cycle to form a closed loop path; An optimization model module, based on the space-time convolutional network, defines a mixed integer programming model, including an objective function and constraints, wherein the objective function minimizes the total transportation cost, and the constraints include a commodity flow balance constraint, a vehicle capacity constraint, a procurement budget constraint, and a battery swap capacity constraint; An algorithm solving module is used to solve the mixed integer programming model, using a column generation algorithm to generate a vehicle path loop and combining it with multidimensional dynamic programming to solve subproblems, where the multidimensional dynamic programming includes a power state update mechanism and a dominance rule to screen non-inferior solution paths.

2. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 1 is characterized in that: The charging arc in the space-time loop network is a horizontal arc that connects consecutive time nodes at the same site, indicating that the electric vehicle is waiting or charging; the service arc is a diagonal arc that describes the space-time changes of the vehicle during transportation; the loop arc is a dotted arc that ensures that the vehicle returns to the starting site, forming a closed-loop scheduling cycle.

3. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 1 is characterized in that: The parameters of the mixed integer programming model include: Electric vehicle collection A set of discrete time points All space-time nodes Product category collection Path Ring Collection The set of all arcs The service arc set Charging Arc Collection Cloth Arc Collection Path ring indicator (Θ s,(i,j) ), when the path ring s contains arc (i, j), the path ring indicator (Θ s,(i,j) ) is 1, when the path ring s does not contain arc (i, j), the path ring indicator (Θ s,(i,j) ) is 0; The parameters of product k include the starting point ( k )、End point (d k ), total amount (w k ); Cost parameters include arc cost (c v,(i,j) ), commodity traffic cost (f k,(i,j) ), vehicle purchase price (b v ); Constraint parameters include the maximum vehicle load (u v ), total procurement budget (B), total battery replacement capacity (H).

4. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 3 is characterized in that: The decision variables include: Integer variable (θ v,s ), represents the number of vehicle models v that select path ring s; Continuous variables (y k,(i,j) ), represents the flow of goods on arc (i, j); In the multidimensional dynamic programming, the sub-problem decision variables include: Arc selection variable (x i,j ), when the electric vehicle selects arc (i, j), the variable value is 1, and when the electric vehicle does not select arc (i, j), the variable value is 0; Power state variable (s i ), represents the amount of electricity of the electric vehicle at time and space point i; Battery replacement indicator variable (q i ), where q i When it is 1, it means that the battery is replaced at the node and the battery is reset to full charge. i When it is 0, it means that no battery replacement is performed at the node.

5. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 3 is characterized in that: The objective function is expressed as: Among them, c v,s is the cost of vehicle type v passing through path cycle s; the constraints include: Traffic balance constraints: Capacity constraints: Procurement budget constraints: Battery swap capacity constraints: Among them, μ v,s is the number of battery replacements for vehicle type v on path loop s; Decision variable range constraints: y≥0.

6. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 1 is characterized in that: The column generation algorithm includes: The main problem is a linear relaxation of the mixed integer programming model, using dual variables (α, β, ψ) to handle constraints; The sub-problem is to solve the path loop cost for each vehicle model. The sub-problem objective function is: Constraints include network flow balancing constraints: And the service cycle ends and returns to the starting point; State of charge constraints: Among them, p v,(i,j) is the change in charge of electric vehicle type v caused by arc (i, j); Other constraints: s≥0.

7. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 6 is characterized in that: The multidimensional dynamic programming includes: Hierarchical structure processing, passing labels layer by layer from the starting point, the label format is Used to record path information, where n is a time and space node and aw n is the cumulative number of battery replacements, ac n To reduce the cumulative cost, ap n is the cumulative power consumption (1-s n ), is the predecessor node label; Dominance rule: For two labels of node n and If there are n intermediate nodes and or n is the terminal node and but Dominate and remove Transfer equation update tag: ac j ,ap j ,aw j Based on the arc type and power state calculation, the equation is as follows:

8. The electric vehicle freight planning system based on a hybrid charging and battery swapping strategy according to claim 6 is characterized in that: The power status update mechanism includes: On the service arc, the power change p v,(i,j) is positive or negative, and p v,(i,j) When it is positive, it corresponds to charging, p v,(i,j) When it is a negative value, it corresponds to discharge; When the battery replacement indicator variable q i =1, the power of node i is s i Reset to full charge state; In dynamic programming, the charge state is updated in real time through transfer equations to handle nonlinear constraints and optimize path feasibility.

Images