Electric truck queue following robust scheduling method

Abstract

The invention relates to an electric truck queue following robust scheduling method, which comprises the following steps of: constructing a basic path planning model with network flow constraint and departure-arrival time constraint, and introducing a truck queue following technology to design a multi-truck queue scheduling constraint module; a travel energy consumption constraint, a queue energy saving constraint, a charging energy supplement constraint, a charging time constraint and an endurance anxiety constraint of the electric truck are added; the method comprises the following steps: collecting mean value and covariance matrix data of two uncertain parameters including travel time in a road network and charging service time of a charging station node, and establishing a travel time fluctuation value fuzzy set and a charging service time fluctuation value fuzzy set based on generalized moment information; a distributed robust opportunity constraint planning method is adopted to construct a robust peer-to-peer model containing two types of fuzzy sets, and the original departure-arrival time constraint and the charging time constraint in the basic model are transformed to generate a robust model. According to the invention, the power consumption in the logistics transportation process can be reduced, and the risk of pre-scheduling failure caused by uncertain factors can be effectively processed.

Description

Technical Field

[0001] The present invention relates to the technical field of electric truck logistics transportation scheduling, and in particular to a robust scheduling method for electric truck platooning and following. Background Art

[0002] With the development of an environmentally friendly economy, the concept of ecological logistics and transportation has been vigorously promoted in China and even around the world. Electric truck technology has gradually matured and its market share has continued to rise, which has also brought new challenges to traditional truck logistics and transportation scheduling. Electric truck distribution scheduling tasks need to consider the goals of high efficiency and low energy consumption of traditional logistics, while taking into account the problems of charging station replenishment and energy consumption anxiety caused by battery life, which is more complicated than traditional scheduling.

[0003] At the same time, platooning technology can indirectly reduce the demand for driving power during driving by affecting the aerodynamics of the vehicle. Its huge potential in energy conservation and emission reduction makes it a hot technology in the field of logistics and transportation. However, the dispatching technology involving platooning has developed relatively late, especially the platooning dispatching technology suitable for highway networks has not yet appeared a mature solution. On the other hand, the application of dispatching technology under the background of platooning is difficult due to high computational complexity and strong time-varying traffic. Uncertain factors such as uncertain travel time, uncertain accessibility of travel sections, and charging queue time are random and complex, resulting in the risk of failure of pre-dispatching solutions in actual transportation scenarios. Therefore, it is crucial to comprehensively study the robust dispatching method of electric trucks that can handle uncertain travel time and charging queue time under platooning. However, given the mutual influence of travel time and charging queue time uncertainties and platooning technology, there is still a blank in the logistics and transportation industry for solving such problems.

[0004] The present invention provides a robust scheduling method for electric truck platoon following, so as to reduce the failure risk of electric truck scheduling schemes in the context of platoon following technology when facing uncertain environments, make up for the defect that traditional models cannot handle the uncertainty of travel time and charging queue time, and provide a more robust, more stable and more energy-saving electric truck scheduling scheme. Summary of the invention

[0005] The purpose of the present invention is to overcome the problems existing in the prior art and to provide a robust scheduling method for electric truck platoon following. The method is based on the idea of ​​distributed robust chance constrained optimization and truck platoon following technology. By analyzing the fluctuation characteristics of travel time and charging queue time in the road network, the mean and covariance matrix data of the above-mentioned uncertain parameter fluctuation values ​​are collected and a fuzzy set based on generalized moment information is established. A distributed robust chance constrained programming method is used to construct a robust peer-to-peer model, which can reduce the power consumption in the logistics transportation process and effectively deal with the risk of pre-scheduling failure caused by uncertain factors. It has positive significance for the actual promotion of truck platoon following technology in the field of logistics and transportation.

[0006] In order to achieve the above technical objectives and the above technical effects, the present invention is implemented through the following technical solutions:

[0007] A robust dispatching method for electric truck platooning and following, the method comprising the following steps:

[0008] Step S1: construct a basic path planning model with network flow constraints and departure-arrival time constraints, introduce the truck platooning method to design a multi-vehicle platooning scheduling constraint module;

[0009] Step S2: adding travel energy consumption constraints, platoon energy saving constraints, charging energy replenishment constraints, charging time constraints and range anxiety constraints of electric trucks;

[0010] Step S3: Collect the mean and covariance matrix data of two uncertain parameters including the travel time in the road network and the charging service time of the charging station node, and establish the travel time fluctuation value fuzzy set and the charging service time fluctuation value fuzzy set based on generalized moment information;

[0011] Step S4: A distributed robust chance-constrained programming method is used to construct a robust peer-to-peer model containing two types of fuzzy sets. The original departure-arrival time constraints and charging time constraints in the basic model are transformed to generate a robust model. Finally, the required vehicle model information, logistics distribution demand information, network travel time and energy consumption history data are input into the model and solved.

[0012] Furthermore, in the step S1, a network flow constraint is constructed based on the balance relationship between network node inflow flow and node outflow flow and the single inlet and outlet principle, a travel time constraint is constructed based on the relationship between arrival node time, departure node time and benchmark travel time, a dependency relationship between multi-vehicle platoon variables and single-vehicle platoon variables is designed, a dependency relationship between single-vehicle platoon variables and network path selection variables is established, and a multi-vehicle platoon variable sum constraint is constructed based on the platoon length limit.

[0013] Furthermore, in step S2, a battery state variable of the electric vehicle is established with the arrival node as the index value, a path energy consumption equation of the electric truck is established based on vehicle attributes and road driving conditions, a platoon energy-saving constraint is constructed based on the platoon technology energy consumption standard, the charging replenishment equation is added to the battery state variable constraint, the charging time constraint is added to the travel time constraint, and a dangerous power state constraint is constructed based on range anxiety.

[0014] Furthermore, the step S1 specifically includes the following steps:

[0015] Step S1.1: Define the set used by the model: All nodes N in the directed graph G are regarded as nodes where vehicles can converge, and they are classified into charging nodes C and non-charging ordinary convergence nodes M according to the attribute of whether they provide charging services. Define the set table used by the model:

[0016]

[0017] Step S1.2: Define the decision variables and parameters used in the model, as follows:

[0018] Define the decision variable table used in the model:

[0019]

[0020] Define the decision parameter table used by the model:

[0021]

[0022]

[0023] Step S1.3: Define the optimization goal of the model: It is composed of the energy consumption and time cost converted from the total logistics energy consumption and the total time spent:

[0024]

[0025] Among them, F1 is the energy consumption cost converted to the total energy consumption of logistics, and F2 is the time cost converted to the total logistics time. is whether the vehicle v passes through the road segment formed by the node (i, j), is whether vehicle v forms a platoon with vehicle m when passing through the road segment formed by node (i, j), is the residence time of vehicle v at node i, is the charging time of vehicle v at charging station node i, is the earliest departure time of vehicle v, q is the charging speed, cp i is the charging unit price of charging station node i, is the state of charge (SOC) value when vehicle v arrives at node i, is the time when vehicle v arrives at node i;

[0026] Set the overall goal as:

[0027] Minimize z

[0028]

[0029] Step S1.4: Define the network flow constraint module:

[0030] This constraint limits a particular vehicle v to have at most one way out from a particular network node i;

[0031] This constraint limits a specific vehicle v to have at most one incoming flow reaching a specific network node j;

[0032] This constraint limits the inflow and outflow status of vehicle v from node i to node j to only three possible relationships depending on whether the node is a starting station, a transit station, or a terminal station: the starting station can only exit but not enter, the terminal can only enter but not exit, and the transit station has a balanced inflow and outflow;

[0033] This constraint limits the vehicle v from node i to node j and then cannot go back to node i from node j, that is, it can only flow in one direction and cannot go back;

[0034] Step S1.5: Define the travel time constraint module:

[0035] This constraint limits the departure time of vehicle v from node i must be equal to or later than the time it arrives at node i plus the residence time at node i

[0036] This constraint specifies that if vehicle v passes through segment (i, j), the time it reaches node j is must be greater than or equal to the departure time from the previous node i Add the travel time T of road section (i, j) i,j ; This constraint limits the time at which vehicle v arrives at node j. It can only exist reasonably when vehicle v reaches node j through a certain road section, otherwise it is infinite;

[0037] This constraint limits the length of time that vehicle v stays at node i. It can only exist reasonably when vehicle v needs to start from node i and pass through a certain road section, otherwise it is infinite;

[0038] This constraint limits the length of time that the driver of vehicle v stays during the entire process, that is, the length of rest time Must be greater than or equal to the total driving time T longer than the mandatory rest period;

[0039] Step S1.6: Construct queue scheduling module:

[0040] This constraint means that if vehicle v follows vehicle m in a platoon on road segment (i, j), the two vehicles start at the same time from node i;

[0041] This constraint means that if vehicle v and vehicle m fail to pass through the road segment (i, j) at the same time, then the two cannot drive in a platoon;

[0042] This constraint means that both vehicle v and vehicle m cannot be the leader vehicle in the platoon at the same time;

[0043] This constraint means that the prerequisite for vehicle v to form a platoon with any vehicle m in the road segment (i, j) is to pass through the road segment (i, j);

[0044] This constraint means that for a platoon led by vehicle m, the number of vehicles following it cannot exceed the maximum length minus one.

[0045] Furthermore, the step S2 specifically includes the following steps:

[0046] Step S2.1: Define the electric truck travel energy consumption constraints:

[0047] This constraint limits the soc value of vehicle v to node i only when it arrives at node i. It makes sense, and at this time node i cannot be included in the departure node Ov set of vehicle v;

[0048] This constraint limits the soc value of vehicle v to only when it departs from node i. It makes sense, and at this time node i cannot be included in the arrival node Dv set of vehicle v;

[0049] This constraint limits the soc value of the arrival time of vehicle v from node i to node j. Must be less than or equal to the soc value at the departure time Subtract energy consumption Constrained The purpose is to make this constraint meaningful only when vehicle v passes through the road section (i, j); Step S2.2: Define the energy-saving constraint of the electric truck platoon:

[0050]

[0051] This constraint represents the energy consumption coefficient of vehicle v queuing on road section (i, j) Depends on its queue status, when it is not queued Indicates that vehicle v does not have any vehicle m that can serve as its leader or following vehicle. At this time, the energy consumption coefficient =ω3=0, that is, the single driving state fails to obtain the energy saving benefits of platooning; when it is platooning and acting as the leading vehicle Indicates that vehicle v has one or more vehicles m that can be its following vehicles. At this time, the energy consumption coefficient =ω2=10%, that is, the platoon driving state as the leading vehicle can obtain 10% platoon energy saving benefits; when it is in the platoon and as a following vehicle Indicates that vehicle v has vehicle m that can be its leader vehicle. At this time, the energy consumption coefficient =ω1=15%, that is, the platoon driving state as the lead vehicle can obtain 15% platoon energy saving benefits;

[0052] Step S2.3: Define the electric truck charging constraints:

[0053] This constraint limits the vehicle to be unable to charge when it is in the non-charging station node N\C set, and the soc value when leaving the node Equal to the soc value when reaching the node

[0054] This constraint limits the vehicle to be charged when it is in the charging station node C set, and the soc value when leaving the node Equal to the soc value when reaching the node Plus charging to supplement the soc value

[0055] This constraint limits the amount of electricity replenished by vehicle v at node i to less than or equal to the maximum soc value, i.e., 1. This constraint is valid only if vehicle v can depart from node i.

[0056] Step S2.4: Define the electric truck charging time constraints:

[0057] This constraint represents the time that the vehicle can charge at node i. Plus charging queue time cw imust be less than or equal to the final total stay time at this node i This constraint is valid under the condition that vehicle v can depart from node i;

[0058] Step S2.5: Define the electric truck range anxiety constraint:

[0059] This constraint represents the soc value when vehicle v arrives at node i. To be greater than or equal to the danger soc anxiety value soc min .

[0060] Furthermore, the step S3 specifically includes the following steps:

[0061] Step S3.1: Obtain the travel time history data for each road segment (i, j) and the charging queue time history data of each charging station node i The sample size can be selected based on the modeling accuracy;

[0062] Step S3.2: Based on the historical travel time data and charging queue time historical data Calculate the mean travel time of road section (i, j) respectively and the mean charging queue time of node i Let the uncertainty parameter be equal to the sum of the nominal value plus the fluctuation value:

[0063] Step S3.3: Calculate the covariance matrix ∑ of travel time fluctuations T and the covariance matrix of the charging queue time fluctuation value ∑ CW :

[0064] Step S3.4: Construct the travel time fuzzy set and charging service time fuzzy set based on the first-order moment and second-order moment information:

[0065]

[0066] Furthermore, the step S4 specifically includes the following steps:

[0067] Step S4.1: Construct the opportunity constraint form involving travel time uncertainty in step S1.6:

[0068] This constraint is equivalent to the original constraint in step S1.6, represents the probability that this constraint holds for the road segment (i, j), and derives this constraint into a tractable form:

[0069]

[0070] Step S4.2: Construct the opportunity constraint form involving the uncertainty of charging queue time in step S2.4:

[0071] make The original constraints are simplified to:

[0072] Constructing Chance Constraints:

[0073] This constraint is equivalent to the original constraint in step S2.4. represents the probability that this constraint holds for the road segment (i, j), and derives this constraint into a tractable form:

[0074]

[0075] Step S4.3: Use the robust peer constraints in steps S4.1 and S4.2 to replace the original departure-arrival time constraints and charging queue time constraints in steps S1.6 and S2.4 to construct a robust peer model;

[0076] Step S4.4: Input the required vehicle model information, logistics distribution demand information, network travel time and road segment energy consumption data into the model and use the business planner to solve it.

[0077] The beneficial effects of the present invention are:

[0078] The present invention analyzes the fluctuation characteristics of travel time and charging queue time in the road network, collects the mean and covariance matrix data of the fluctuation values ​​of the above uncertain parameters and establishes a fuzzy set based on generalized moment information, and adopts a distributed robust chance-constrained programming method to construct a robust peer-to-peer model. It can reduce the power consumption in the logistics transportation process and effectively deal with the risk of pre-scheduling failure due to uncertain factors. It has positive significance for the actual promotion of truck platooning following technology in the field of logistics and transportation. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is a framework diagram of the present invention;

[0080] Figure 2 This is a road network structure diagram of a preferred example of the present invention;

[0081] Figure 3 It is a flow chart of the overall idea of ​​the present invention. DETAILED DESCRIPTION

[0082] The present invention will be described in detail below with reference to the accompanying drawings and in combination with embodiments.

[0083] A robust scheduling method for electric truck platooning is shown in the following figure. Figure 1As shown in FIG. 1 , in this embodiment, a typical road network structure including 7 nodes and 24 paths is used as an example, and 3 electric trucks with different transportation tasks are added. Figure 2 , according to the actual logistics transportation scheduling rules, the scenario parameters are set as shown in Tables 1, 2, 3, and 4, and historical data including uncertain travel time and charging queue time are generated. The specific implementation method of the method proposed in the present invention is as follows:

[0084] Define the collection used by the model:

[0085]

[0086] Where G represents a directed graph, N represents a node set {A, B, C, D, E, F, G}, and converts them into hot-single encodings {0, 1, 2, 3, 4, 5, 6} in sequence, and E represents a road segment set in the directed graph. See Table 1 for details, which is a road segment information parameter table:

[0087]

[0088] , OD represents the starting and ending points of vehicle v, and V represents the set of all vehicles. See Table 2 for details, which is the truck information parameter table:

[0089]

[0090]

[0091] , C represents the set of all charging nodes, and M represents the set of all non-charging nodes. The present invention regards all nodes N in the directed graph G as nodes where vehicles can converge, where they can be classified into charging nodes C and non-charging ordinary convergence nodes M according to the attribute of whether to provide charging services. See Table 3 for details, which is a node information parameter table:

[0092]

[0093] Define the decision variables used in the model;

[0094]

[0095] in, Indicates whether vehicle v passes through the road segment formed by nodes (i, j), Indicates whether vehicle v forms a platoon with vehicle m when passing through the road segment formed by node (i, j), represents the residence time of vehicle v at node i, represents the charging time of vehicle v at charging station node i, when node i belongs to the charging station node set C, Represents the charging decision variable, which is a 0-1 type Boolean variable;

[0096] Define the parameters used by the model:

[0097] L,R,q,cp i ,soc min ,M,T i,j ,E i,j , cw i

[0098] in, represents the earliest departure time of vehicle v, represents the latest arrival time of vehicle v, L represents the maximum length limit of the queue, R represents the mandatory rest time that the driver should take every T hours of driving, q represents the charging speed, cp i represents the charging unit price of charging station node i, which is uniformly selected as 400 in this preferred case, indicating the charging cost per hour; soc min It represents the dangerous state of charge of the electric vehicle, that is, the dangerous soc value. If it is lower than this value, anxiety will occur. M represents an infinite integer. T i,j represents the uncertain travel time required to pass the road segment (i, j), E i,j represents the energy consumption required to pass the road section (i, j), represents the platoon energy consumption coefficient of vehicle v passing through road section (i, j), represents the state of charge (soc) value when vehicle v arrives at node i, represents the state of charge (soc) value of vehicle v when it departs from node i, represents the time when vehicle v departs from node i, represents the time when vehicle v arrives at node i, cw i It represents the uncertain charging queue time of the vehicle at charging station node i. The specific parameters are shown in Tables 1, 2, 3, and 4. The optimization objective of the defined model is composed of the energy consumption cost and time cost converted from the total logistics energy consumption and the total time spent. The total logistics energy consumption is converted into energy consumption cost:

[0099]

[0100] The total logistics time is converted into time cost, where tp is the hourly wage. The configuration is shown in Table 1:

[0101]

[0102] The overall optimization goal is set as follows:

[0103] Minimize z

[0104]

[0105] Define the network flow constraint module:

[0106]

[0107]

[0108] Define the travel time constraint module:

[0109]

[0110] Add a constraint involving uncertain travel times:

[0111]

[0112] This constraint specifies that if vehicle v passes through segment (i, j), the time it reaches node j is must be greater than or equal to the departure time from the previous node i Add the travel time T of road section (i, j) i,j

[0113] Build a queue scheduling module:

[0114]

[0115] Define the energy consumption constraints for the electric truck trip:

[0116]

[0117] Define the energy conservation constraints for electric truck platooning:

[0118]

[0119] This constraint represents the energy consumption coefficient of vehicle v queuing on road section (i, j) Depends on its queue status, when it is not queued Indicates that vehicle v does not have any vehicle m that can serve as its leader or following vehicle. At this time, the energy consumption coefficient =ω3=0, that is, the single driving state fails to obtain the energy saving benefits of platooning; when it is platooning and acting as the leading vehicle Indicates that vehicle v has one or more vehicles m that can be its following vehicles. At this time, the energy consumption coefficient =ω2=10%, that is, the platoon driving state as the leading vehicle can obtain 10% platoon energy saving benefits; when it is in the platoon and as a following vehicle Indicates that vehicle v has vehicle m that can be its leader vehicle. At this time, the energy consumption coefficient Equal to ω1 = 15%, that is, the platoon driving state as the head vehicle can obtain 15% of the platoon energy saving benefits. Define the electric truck charging and energy replenishment constraints:

[0120]

[0121]

[0122] Define the charging time constraint for electric trucks, which involves the uncertain charging queue time:

[0123]

[0124] Defining the electric truck range anxiety constraints:

[0125]

[0126] This constraint represents the soc value when vehicle v arrives at node i. To be greater than or equal to the danger soc anxiety value soc min , here we can take 0.2 to ensure that the vehicle v can retain at least 20% of its power to reach the nearby charging station.

[0127] Establish travel time fuzzy set and charging service time fuzzy set based on generalized moment information:

[0128] Get the travel time history data for each road segment (i, j) and the charging queue time history data of each charging station node i In the preferred case of the present invention, 10 is selected as the sample size:

[0129] Based on historical travel time data and charging queue time historical data Calculate the mean travel time of road section (i, j) respectively and the mean charging queue time of node i See Table 4 for details, which is a table of configuration parameters for the scheduling task scenario:

[0130]

[0131] , let the uncertainty parameter be equal to the sum of the mean and the fluctuation:

[0132]

[0133]

[0134] Calculate the covariance matrix Σ of travel time fluctuation values T and the covariance matrix Σ of the charging queue time fluctuation value CW :

[0135]

[0136] In the preferred case of the present invention, both the uncertain travel time and the uncertain charging queue time are designed as one-dimensional vectors, so the calculation method of the covariance matrix is ​​the variance calculation method. Taking the uncertain travel time as an example, the calculation method is as follows:

[0137]

[0138] Construct the travel time fuzzy set and charging service time fuzzy set based on the first-order moment and second-order moment information:

[0139]

[0140] The opportunity constraint involving travel time uncertainty in step S1.6 is constructed as:

[0141]

[0142] The form of the opportunity constraint involving uncertainty in charging queue time in step S2.4 is constructed as:

[0143]

[0144] The original departure-arrival time constraint and charging queue time constraint in steps S1.6 and S2.4 are replaced by the robust equivalent constraints in steps S4.1 and S4.2, and the probability is set to 0.95.

[0145] The required vehicle model information, logistics distribution demand information, network travel time and road section energy consumption data are input into the model and solved using a business planner. In the preferred case of the present invention, Python+Gurobi planner is used for solving.

[0146] The solution shows that the model considering uncertainty factors is effective when the historical data of travel time and charging queue time are as shown in Table 5, where Table 5 is the configuration parameter table of the scheduling task scenario:

[0147]

[0148] , it takes 2.63 seconds to obtain the optimal cost of 7101.2. Compared with the random programming method, the fuzzy set boundary of the uncertainty parameter in the preferred case of the present invention is smaller than the probability distribution boundary. To simplify the comparative test, when using a single historical data as the travel time and queue time parameters of the deterministic model in the random programming, a total of 100 scenarios are selected, which takes a total of 289.1 seconds. In this case, the method can improve the computational efficiency by 91% and meet the scheduling feasibility under various uncertainty parameter configurations.

[0149] The present invention uses the idea of ​​distributed robust chance-constrained optimization and truck platoon-following technology. By analyzing the fluctuation characteristics of travel time and charging queue time in the road network, a fuzzy set based on generalized moment information is established, and a robust chance-constrained peer-to-peer model is constructed. It has the advantages of reducing the power consumption in the logistics transportation process and the risk of pre-scheduling failure due to uncertain factors. At the same time, the present invention can handle the problem of electric truck platoon-following scheduling within the network, and has the advantages of less computing resource consumption, low processing difficulty, and a wide range of application scenarios.

[0150] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A robust dispatching method for electric truck platooning, characterized in that: The method comprises the following steps: Step S1: construct a basic path planning model with network flow constraints and departure-arrival time constraints, introduce the truck platooning method to design a multi-vehicle platooning scheduling constraint module; Step S2: adding travel energy consumption constraints, platoon energy saving constraints, charging energy replenishment constraints, charging time constraints and range anxiety constraints of electric trucks; Step S3: Collect the mean and covariance matrix data of two uncertain parameters including the travel time in the road network and the charging service time of the charging station node, and establish the travel time fluctuation value fuzzy set and the charging service time fluctuation value fuzzy set based on generalized moment information; Step S4: A distributed robust chance-constrained programming method is used to construct a robust peer-to-peer model containing two types of fuzzy sets. The original departure-arrival time constraints and charging time constraints in the basic model are transformed to generate a robust model. Finally, the required vehicle model information, logistics distribution demand information, network travel time and energy consumption history data are input into the model and solved.

2. The robust dispatching method for electric truck platooning according to claim 1 is characterized in that: In the step S1, a network flow constraint is constructed based on the balance relationship between the inflow flow of a network node and the outflow flow of a node and the single inlet and outlet principle; a travel time constraint is constructed based on the relationship between the arrival node time, the departure node time and the benchmark travel time; a dependency relationship between a multi-vehicle platoon variable and a single-vehicle platoon variable is designed; a dependency relationship between a single-vehicle platoon variable and a network path selection variable is established; and a multi-vehicle platoon variable sum constraint is constructed based on a platoon length limit.

3. The robust dispatching method for electric truck platooning according to claim 2 is characterized in that: In the step S2, a battery state variable of the electric vehicle is established with the arrival node as the index value, a path energy consumption equation of the electric truck is established according to the vehicle attributes and road driving conditions, a platoon energy-saving constraint is constructed based on the platoon technology energy consumption standard, the charging replenishment equation is added to the battery state variable constraint, the charging time constraint is added to the travel time constraint, and a dangerous power state constraint is constructed according to the range anxiety.

4. The robust dispatching method for electric truck platooning according to claim 3 is characterized in that: The step S1 specifically includes the following steps: Step S1.1: Define the set used by the model: All nodes N in the directed graph G are regarded as nodes where vehicles can converge, and they are classified into charging nodes C and non-charging ordinary convergence nodes M according to the attribute of whether they provide charging services; Step S1.2: Define the decision variables and parameters used in the model; Step S1.3: Define the optimization goal of the model: It is composed of the energy consumption and time cost converted from the total logistics energy consumption and the total time spent: Among them, F1 is the energy consumption cost converted to the total energy consumption of logistics, and F2 is the time cost converted to the total logistics time. is whether the vehicle v passes through the road segment formed by nodes (i, j), is whether vehicle v forms a platoon with vehicle m when passing through the road segment formed by node (i, j), is the residence time of vehicle v at node i, is the charging time of vehicle v at charging station node i, is the earliest departure time of vehicle v, q is the charging speed, cp i is the charging unit price of charging station node i, is the state of charge (SOC) value when vehicle v arrives at node i, is the time when vehicle v arrives at node i; Set the overall goal as: Minimize z Step S1.4: Define the network flow constraint module: This constraint limits a particular vehicle v to have at most one way out from a particular network node i; This constraint limits a specific vehicle v to have at most one incoming flow reaching a specific network node j; This constraint limits the inflow and outflow status of vehicle v from node i to node j to only three possible relationships depending on whether the node is a starting station, a transit station, or a terminal station: the starting station can only exit but not enter, the terminal can only enter but not exit, and the transit station has a balanced inflow and outflow; This constraint limits the vehicle v from node i to node j and then cannot go back to node i from node j, that is, it can only flow in one direction and cannot go back; Step S1.5: Define the travel time constraint module: This constraint limits the departure time of vehicle v from node i must be equal to or later than the time it arrives at node i plus the residence time at node i This constraint specifies that if vehicle v passes through segment (i, j), the time it reaches node j is must be greater than or equal to the departure time from the previous node i Add the travel time T of road section (i, j) i，j ; This constraint limits the time at which vehicle v arrives at node j. It can only exist reasonably when vehicle v reaches node j through a certain road section, otherwise it is infinite; This constraint limits the length of time that vehicle v stays at node i. It can only exist reasonably when vehicle v needs to start from node i and pass through a certain road section, otherwise it is infinite; This constraint limits the length of time that the driver of vehicle v stays during the entire process, that is, the length of rest time Must be greater than or equal to the total driving time T longer than the mandatory rest period; Step S1.6: Construct queue scheduling module: This constraint means that if vehicle v follows vehicle m in a platoon on road segment (i, j), the two vehicles start at the same time from node i; This constraint means that if vehicle v and vehicle m fail to pass through road segment (i, j) at the same time, then the two cannot drive in a platoon; This constraint means that both vehicle v and vehicle m cannot be the leader vehicle in the platoon at the same time; This constraint means that the prerequisite for vehicle v to form a platoon with any vehicle m in the road segment (i, j) is to pass through the road segment (i, j); This constraint means that for a platoon led by vehicle m, the number of vehicles following it cannot exceed the maximum length minus one.

5. The robust dispatching method for electric truck platooning according to claim 4 is characterized in that: The step S2 specifically includes the following steps: Step S2.1: Define the electric truck travel energy consumption constraints: This constraint limits the soc value of vehicle v to node i only when it arrives at node i. It makes sense, and at this time node i cannot be included in the departure node Ov set of vehicle v; This constraint limits the soc value of vehicle v to only when it departs from node i. It makes sense, and at this time node i cannot be included in the arrival node Dv set of vehicle v; This constraint limits the soc value of the arrival time of vehicle v from node i to node j. Must be less than or equal to the soc value at the departure time Subtract energy consumption Constrained The purpose is to make this constraint meaningful only when vehicle v passes through the road section (i, j); Step S2.2: Define the energy-saving constraint of the electric truck platoon: ， This constraint represents the energy consumption coefficient of vehicle v queuing on road section (i, j) Depends on its queue status, when it is not queued Indicates that vehicle v does not have any vehicle m that can serve as its leader or following vehicle. At this time, the energy consumption coefficient =ω3=0, that is, the single driving state fails to obtain the energy saving benefits of platooning; when it is platooning and acting as the leading vehicle Indicates that vehicle v has one or more vehicles m that can be its following vehicles. At this time, the energy consumption coefficient =ω2=10%, that is, the platoon driving state as the leading vehicle can obtain 10% platoon energy saving benefits; when it is in the platoon and as a following vehicle Indicates that vehicle v has vehicle m that can be its leader vehicle. At this time, the energy consumption coefficient =ω1=15%, that is, the platoon driving state as the lead vehicle can obtain 15% platoon energy saving benefits; Step S2.3: Define the electric truck charging constraints: This constraint limits the vehicle to be unable to charge when it is in the non-charging station node N\C set, and the soc value when leaving the node Equal to the soc value when reaching the node This constraint limits the vehicle to be charged when it is in the charging station node C set, and the soc value when leaving the node Equal to the soc value when reaching the node Plus charging to supplement the soc value This constraint limits the amount of electricity replenished by vehicle v at node i to less than or equal to the maximum soc value, i.e., 1. This constraint is valid only if vehicle v can depart from node i. Step S2.4: Define the electric truck charging time constraints: This constraint represents the time that the vehicle can charge at node i. Plus charging queue time cw i must be less than or equal to the final total stay time at this node i This constraint is valid under the condition that vehicle v can depart from node i; Step S2.5: Define the electric truck range anxiety constraint: This constraint represents the soc value when vehicle v arrives at node i. To be greater than or equal to the danger soc anxiety value soc min .

6. The robust dispatching method for electric truck platooning according to claim 5, characterized in that: The step S3 specifically comprises the following steps: Step S3.1: Obtain the travel time history data for each road segment (i, j) and the charging queue time history data of each charging station node i The sample size can be selected based on the modeling accuracy; Step S3.2: Based on the historical travel time data and charging queue time historical data Calculate the mean travel time of road section (i, j) respectively and the mean charging queue time of node i Let the uncertainty parameter be equal to the sum of the nominal value plus the fluctuation value: Step S3.3: Calculate the covariance matrix ∑ of travel time fluctuations T and the covariance matrix ∑ of the charging queue time fluctuation value CW : Step S3.4: Construct the travel time fuzzy set and charging service time fuzzy set based on the first-order moment and second-order moment information:

7. The robust dispatching method for electric truck platooning according to claim 6, characterized in that: The step S4 specifically comprises the following steps: Step S4.1: Construct the opportunity constraint form involving travel time uncertainty in step S1.6: This constraint is equivalent to the original constraint in step S1.6, represents the probability that this constraint holds for the road segment (i, j), and derives this constraint into a tractable form: Step S4.2: Construct the opportunity constraint form involving the uncertainty of charging queue time in step S2.4: make The original constraints are simplified to: Constructing Chance Constraints: This constraint is equivalent to the original constraint in step S2.

4. represents the probability that this constraint holds for the road segment (i, j), and derives this constraint into a tractable form: Step S4.3: Use the robust peer constraints in steps S4.1 and S4.2 to replace the original departure-arrival time constraints and charging queue time constraints in steps S1.6 and S2.4 to construct a robust peer model; Step S4.4: Input the required vehicle model information, logistics distribution demand information, network travel time and road segment energy consumption data into the model and use the business planner to solve it.

Images