Optimal scheduling method and system for electric vehicle charging network design

Abstract

The invention discloses an optimal scheduling method and system for electric vehicle charging network design, and relates to the technical field of electric vehicle fleet routing. The method comprises the steps of defining a charging function and a wear cost function according to an actual battery charging process and a battery wear condition; according to the monotonic relationship between the wear cost and the charging state, constructing a mathematical model of a corresponding wear cost function, and setting constraint conditions; and solving the mathematical model of the wear cost function by adopting a three-stage efficient heuristic algorithm of the CWIGALNS to obtain an optimal scheduling scheme. According to the method, the influence of the charging process on electric vehicle dispatching is fully considered, the collaborative optimization problem of electric vehicle distribution service and charging dispatching can be systematically analyzed and solved, and the logistics operation cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle fleet routing, and in particular to an optimal scheduling method and system designed for an electric vehicle charging network. Background Art

[0002] Compared with traditional logistics vehicles, electric vehicles incur additional charging costs, such as charging time and battery wear costs. It is crucial to reasonably arrange the routes and charging plans of electric vehicles according to the power distribution service requirements and battery characteristics, delay battery degradation, and extend battery cycle life.

[0003] Most existing methods determine route scheduling or facility site selection plans under limited battery driving range. This method of adding mileage restrictions on the basis of traditional research methods usually ignores the impact of the battery charging process on electric vehicle scheduling and cannot incorporate the charging strategy into the location decision and routing problems of the charging facilities. Therefore, the present invention proposes an optimal scheduling method and system for electric vehicle charging network design. Summary of the invention

[0004] The purpose of the present invention is to provide an optimal scheduling method and system for electric vehicle charging network design, which can systematically analyze and solve the problem of coordinated optimization of electric vehicle delivery services and charging scheduling, and reduce logistics operation costs.

[0005] According to a first aspect of the present invention, in order to achieve the above-mentioned purpose, the present invention provides the following technical solution: an optimal scheduling method for electric vehicle charging network design, comprising the following steps:

[0006] According to the actual charging process of the battery and the battery wear, the charging function and the wear cost function are defined;

[0007] According to the monotonically increasing or decreasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraint conditions are set;

[0008] The three-stage efficient heuristic algorithm of CWIGALNS is used to solve the mathematical model of the wear cost function and obtain the optimal scheduling solution.

[0009] Furthermore, according to the actual charging process of the battery and the battery wear, the charging function and the wear cost function are defined as follows:

[0010] (21) The changes in current and voltage during the actual battery charging process are nonlinear, so the charging function including the constant current and constant voltage stages is nonlinear;

[0011] Assume that each charger has a specific constant current-constant voltage charging function, the charging function has b+1 breakpoints, and fits the real constant current-constant voltage concave function. Let the charging state associated with breakpoint i be S i , where i∈B, B={0,…,b};

[0012] Assuming that the charging current between consecutive breakpoints is constant, the charging function is:

[0013] T=∑ d∈{1,2,3} t d S d (1)

[0014] Where, T is the total charging time, t d S is the unit charging time of the battery charging state in interval d, d The power of the battery in the charging state interval d;

[0015] (22) Let D be the set of charging state time intervals, and let L be the length of each time interval:

[0016]

[0017] D={1,2,...,n d} (3)

[0018] In the formula, and S d They are the upper and lower limits of the charging state respectively;

[0019] Battery Price (W P ) can be expressed as:

[0020]

[0021] DOD∈{L,2L,3L,...,1}(5)

[0022] In the formula, W( S d ) is in The cost of charging and discharging per kWh, Δ q It is the energy value of each charging state interval, DOD is the depth of discharge, and ACC is the battery cycle count.

[0023] Furthermore, according to the monotonically increasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set, as follows:

[0024] (31) A mixed integer linear programming formula is proposed to minimize the operating cost as follows:

[0025]

[0026] Where, d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable, if vehicle k goes from node i to node j, then the value is 1, otherwise it is 0; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h successively at charging station j; W d The wear cost per kilowatt-hour of charging and discharging in interval d; t d is the charging time per unit of electricity in interval d; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the number of intervals d∈D where vehicle k charges in the parking lot.

[0027] (32) The specific constraints are as follows:

[0028] Each customer is served only once:

[0029]

[0030] The vehicle is charged at a location with charging facilities:

[0031]

[0032] Traffic flow balance at each node:

[0033]

[0034] Vehicle departure and return from the depot:

[0035]

[0036] Each EV serves only one route:

[0037]

[0038] The remaining cargo quantity before and after the electric vehicle passes a certain node:

[0039]

[0040] Range of remaining cargo in electric vehicle:

[0041]

[0042] The remaining energy of the electric vehicle at each node on the service route:

[0043]

[0044] The remaining power of the electric vehicle when it leaves the distribution center:

[0045]

[0046] At the client node, the battery charge remains unchanged:

[0047]

[0048] The vehicle's charge level before and after leaving the charging station:

[0049]

[0050] Upper and lower limits of charge capacity:

[0051]

[0052] Make sure the vehicle has enough charge to complete the assigned task:

[0053]

[0054] Ensure that the total energy charged by charging station j in all charging state intervals is equal to the total charge amount of vehicle k by charging station j:

[0055]

[0056] Formula (21)-Formula (23) ensures that sufficient power is charged in each charging state interval:

[0057]

[0058] The energy consumption cost of vehicle k charging at the charging station is:

[0059]

[0060] The charging time cost of the energy charged to vehicle k at the charging station:

[0061]

[0062] Formula (26)-Formula (27) ensures that within the charging state interval, the variable u gjhkd and u′ kd must be equivalent to a charging state interval whose upper limit is higher than the charge of vehicle k before it arrives at charging station j or leaves the charging station:

[0063]

[0064]

[0065] Binary constraints:

[0066]

[0067] Where J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o′ is another warehouse of the same type; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle's mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable. If vehicle k goes from node i to node j, the value is 1, otherwise it is 0; u′ k ' d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g.

[0068] Furthermore, according to the monotonically decreasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set, as follows:

[0069] (41) Define variables is the amount of electricity released by vehicle k on the function arg(g,h) in the charging state interval d∈D; variable S′ g ' hk is the total wear cost of vehicle k on the function arg(g,h);

[0070] If the battery of vehicle k is discharged on the function arg(g,h) over interval d, then the binary variable is equal to 1; represents the total emission of vehicle k on the function arg(g,h), the specific formula is as follows:

[0071]

[0073] (42) The above function (29) is subject to the constraints of formula (7)-formula (25) and formula (28), and is also subject to the following constraints:

[0074] Formula (31)-Formula (32) constrains the maximum acceptable interval when vehicle k leaves station j or depot o by calculating the difference between its lower limit and the current SOC:

[0075]

[0076] Ensure that the maximum discharge energy per SOC interval does not exceed L under applicable conditions:

[0077]

[0078] Ensure that for each vehicle k, the total discharge in all SOC intervals at charging station j is equal to the total discharge at charging station j:

[0079]

[0080] The total emissions of vehicle k along the route from node g to node h:

[0081]

[0082] Constraint (35) is the same as constraint (30), but only for discharge:

[0083]

[0084] The total loss cost of vehicle k discharging on the function arc(g,h):

[0085]

[0086] Where J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o′ is another warehouse of the same type; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle's mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable. If vehicle k goes from node i to node j, the value is 1, otherwise it is 0; u′ k ' d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g.

[0087] Furthermore, the three-stage efficient heuristic algorithm of CWIGALNS includes a Clarke and Wright preservation algorithm, an iterative greedy algorithm and an adaptive large neighborhood search algorithm.

[0088] Furthermore, the Clarke and Wright preservation algorithm is as follows:

[0089] (61) Create a route (oro) for each customer r;

[0090] (62) Identify the edge nodes directly connected to warehouse o;

[0091] (63) Select two edge nodes r1 and r2 in routes a1 and a2 as the saved pair and calculate the saved value and find the total demand for new routes;

[0092] (64) Select the stored pair (r1, r2) in the SPL in the order of stored values, and identify the route a1 and route a2 containing the edge nodes r1 and r2 respectively;

[0093] (65) Connect routes a1 and a2. If the total demand of the new route exceeds the vehicle's loading capacity, abandon the merging operation. Otherwise, end the merging operation.

[0094] (66) Delete the saved pair (r1, r2) from the program and repeat the process from step (64) to step (66) until the program is empty.

[0095] Furthermore, when the iterative greedy algorithm starts, all charging stations are deleted and a new charging station is selected based on the placement cost. The specific parameters are as follows:

[0096] (71) Breakpoint: Under the mileage limit, if the electric vehicle does not charge at a charging station during driving, it will not be able to reach certain nodes; the breakpoint v has the following properties: Define route k The first breakpoint on the * ;

[0097] (72) Node accessibility: In order to compare the impact of different charging stations on customer points in the route, we define the parameter q vk The accessibility of the customer points, Parameter q vk It is used to indicate whether the electric vehicle will run out of power on the way, and the minimum value of the path customer point accessibility is defined as q * ;

[0098] (73) Charging strategy: If charging time and battery wear increase with the charging state, electric vehicles need to recharge their batteries at the depot or charging station until they reach the next stop or return to the depot, and the minimum power is not less than Q min ;

[0099] (74) Placement cost: on path r k Add a charging station j∈J to the node v on the network, denoted as v, and the placement cost is

[0100]

[0101] β1+β2+β3+β4+β5=1,β1,β2,β3,β4,β5≥0 (38)

[0102]

[0103]

[0104]

[0105] In the formula, is the increase in accessibility, is the additional routing cost, is the wear cost, is the additional billable time cost, is the penalty value; q vk and q′ vk denote the reachability of v before and after insertion into site j, respectively; and They represent the minimum value of the current path reachability before and after inserting j.

[0106] Furthermore, in the ALNS algorithm, a delete operator is used to delete n from the current route. * customer nodes and repeatedly use the insertion operator to reinsert them into the route, * Let it be a random value in [μ1|I|,μ2|I|], where μ1,μ2∈(0,1):

[0107] (81) Assign s0 to the current solution s and the optimal solution s * ;

[0108] (82) If the number of iterations is less than a given number, the insertion and removal operators are selected and s is assigned to the neighborhood solution s′, otherwise, the search ends;

[0109] (83) Use the removal and insertion operators to process s′;

[0110] (84) If the acceptance criteria are met at this time, the value s′ is assigned to s;

[0111] (85) If z(s) <z(s * ), then assign s to s * ;

[0112] (86) Update the score and weight of each operator and return to step (82);

[0113] The adaptive search mechanism used by the ALNS algorithm is as follows:

[0114] At the beginning of each stage, the operator weights are adjusted and the operators are selected according to the preset rules to achieve adaptive update of the search process, as follows:

[0115] Let the weight of operator i in stage j be w ij , then the probability of the operator being selected is p ij =w ij / ∑ h∈H w hj , where H is the operator set; the jth stage requires updating the operator weights according to the following rules after each search stage:

[0116] Define the number of times operator I is used in stage j, and the scores are ∈ ij and π ij ; If operator i is used in stage j, its weight in the next stage is (1-θ)w ij +θπ ij / ∈ ij , otherwise the weight remains unchanged, where θ is the reaction coefficient, set to 0.3;

[0117] If the neighborhood solution s′ is better than the current solution s, the neighborhood solution is retained; otherwise, the probability that the neighborhood solution is retained is e -(z(s′)-z(s)) / T , where the initial temperature T0 = 10000, T n =cT n-1 , c=0.995.

[0118] According to a second aspect of the present invention, the present invention provides an optimal scheduling system for electric vehicle charging network design, which is used to implement the above-mentioned optimal scheduling method for electric vehicle charging network design, including:

[0119] Function definition module, used to define charging function and wear cost function according to the actual charging process of the battery and the battery wear;

[0120] A construction module is used to construct a mathematical model of a corresponding wear cost function according to a monotonically increasing or decreasing relationship between the wear cost and the charging state, and to set constraint conditions;

[0121] The solution output module is used to solve the mathematical model of the wear cost function using the three-stage efficient heuristic algorithm of CWIGALNS to obtain the optimal scheduling solution.

[0122] The present invention has at least the following beneficial effects:

[0123] 1. The present invention fully considers the impact of the charging process on electric vehicle scheduling, can systematically analyze the collaborative optimization problem of electric vehicle distribution service and charging scheduling, establish an overall service decision-making plan based on nonlinear charging time function and battery wear characteristics, and reduce logistics operation costs.

[0124] 2. The present invention proposes a nonlinear charging function, so that the battery charging rate is no longer a constant value, but a function of the current state of the battery.

[0125] 3. The present invention uses a three-stage efficient heuristic algorithm which is more efficient and has stronger stability than the traditional CPLEX algorithm.

[0126] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0127] Figure 1 A schematic diagram of the flow of the scheduling method described in an embodiment of the present invention. DETAILED DESCRIPTION

[0128] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0129] See also Figure 1 The present invention provides a technical solution: an optimal scheduling method for electric vehicle charging network design, comprising the following steps:

[0130] S1. According to the actual battery charging process and battery wear, the charging function and wear cost function are defined as follows:

[0131] S11. The changes of current and voltage in the actual battery charging process are nonlinear, so the charging function including constant current (CC) and constant voltage (CV) stages is nonlinear;

[0132] As time goes by, the state of charge (SOC) is piecewise linearly approximated. It is assumed that each charger has a specific constant current-constant voltage charging function with b+1 breakpoints, and the real constant current-constant voltage concave function is fitted. Let the charging state associated with breakpoint i be S i , where i∈B, B={0,…,b};

[0133] Assuming that the charging current between consecutive breakpoints is constant, the charging function is:

[0134] T=∑ d∈{1,2,3} t d S d (1)

[0135] Where, T is the total charging time, t dS is the unit charging time of the battery charging state in interval d, d The power of the battery in the charging state interval d;

[0136] S12. In actual applications, a battery will go through multiple cycles in its life cycle, and the cycle life and depth of discharge (DOD) and initial achievable cycle count (ACC) of each cycle are different, which will cause different degrees of wear on the battery;

[0137] Let D be the set of charging state time intervals, and let L be the length of each time interval:

[0138]

[0139] D={1,2,...,n d} (3)

[0140] In the formula, and S d They are the upper and lower limits of the charging state respectively;

[0141] Battery Price (W P ) can be expressed as:

[0142]

[0143] DOD∈{L,2L,3L,...,1}(5)

[0144] In the formula, W( S d ) is in The cost of charging and discharging per kWh, Δ q is the energy value for each charging state interval, DOD is the depth of discharge, and ACC is the battery cycle count;

[0145] S2. According to the monotonically increasing or decreasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set;

[0146] S21. According to the monotonically increasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set, as follows:

[0147] (S21.1) The solution derived from the model always automatically fills the lower SOC interval, and the lowest acceptable interval is no longer bounded by the length L of the interval, but by the difference between its upper limit and the SOC of the vehicle arriving at the current station (or destination);

[0148] In a non-decreasing wear function with respect to SOC, all the charged energy will subsequently be released under any optimal solution of the Electric Vehicle Routing Problem with Nonlinear Charging and Battery Wear (EVRP-NCBW), and any preliminary charge in the vehicle battery will be fully discharged.

[0149] Therefore, a mixed integer linear programming formula is proposed to minimize the operating cost (mainly including: construction cost, route cost, battery wear cost and charging time cost), as follows:

[0150]

[0151] Where, d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable, if vehicle k goes from node i to node j, then the value is 1, otherwise it is 0; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h successively at charging station j; W d The wear cost per kilowatt-hour of charging and discharging in interval d; t d is the charging time per unit of electricity in interval d; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the number of intervals d∈D where vehicle k charges in the parking lot.

[0152] (S21.2) The constraints are as follows:

[0153] Each customer is served only once:

[0154]

[0155] The vehicle is charged at a location with charging facilities:

[0156]

[0157] Traffic flow balance at each node:

[0158]

[0159] Vehicle departure and return from the depot:

[0160]

[0161] Each electric vehicle (EV) serves only one route:

[0162]

[0163] The remaining cargo quantity before and after the electric vehicle passes a certain node:

[0164]

[0165] Range of remaining cargo in electric vehicle:

[0166]

[0167] The remaining energy of the electric vehicle at each node on the service route:

[0168]

[0169] The remaining power of the electric vehicle when it leaves the distribution center:

[0170]

[0171] At the client node, the battery charge remains unchanged:

[0172]

[0173] The vehicle's charge level before and after leaving the charging station:

[0174]

[0175] Upper and lower limits of charge capacity:

[0176]

[0177] Make sure the vehicle has enough charge to complete the assigned task:

[0178]

[0179] Ensure that the total energy charged by charging station j in all charging state intervals is equal to the total charge amount of vehicle k by charging station j:

[0180]

[0181] Formula (21)-Formula (23) ensures that sufficient power is charged in each charging state interval:

[0182]

[0183] The energy consumption cost of vehicle k charging at the charging station is:

[0184]

[0185] The charging time cost of the energy charged to vehicle k at the charging station:

[0186]

[0187] Formula (26)-Formula (27) ensures that within the charging state interval, the variable u gjhkd and u′ kd must be equivalent to a charging state interval whose upper limit is higher than the charge of vehicle k before it arrives at charging station j or leaves the charging station:

[0188]

[0189] Binary constraints:

[0190]

[0191] Where J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o′ is another warehouse of the same type; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle's mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable. If vehicle k goes from node i to node j, the value is 1, otherwise it is 0; u′ k ' d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g.

[0192] S22. According to the monotonically decreasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set, as follows:

[0193] (S22.1) The wear cost function decreases monotonically with respect to SOC, indicating that battery degradation will be more serious when cycling at a lower SOC. In this case, in order to achieve (or maintain) a more ideal SOC interval, part of the energy will be charged in the optimization scheme of the electric freight vehicle distribution network model with nonlinear charging and battery wear model (ELRP-NCBW). That is, some energy may be charged without ever having a chance to discharge. Therefore, the cycle degradation of charge and discharge must be calculated separately, so that the charge and discharge energy of each SOC interval in the degradation model can be monitored separately;

[0194] Defining variables is the amount of electricity released by vehicle k on the function arg(g,h) in the charging state interval d∈D; variable S′ g ' hk is the total wear cost of vehicle k on the function arg(g,h);

[0195] If the battery of vehicle k is discharged on the function arg(g,h) over interval d, then the binary variable is equal to 1; represents the total emission of vehicle k on the function arg(g,h), the specific formula is as follows:

[0196]

[0197] (S22.2) The above function (29) is subject to the following constraints while being constrained by formula (7)-formula (25) and formula (28):

[0198] Formula (31)-Formula (32) constrains the maximum acceptable interval when vehicle k leaves station j or depot o by calculating the difference between its lower limit and the current SOC:

[0199]

[0200] Ensure that the maximum discharge energy per SOC interval does not exceed L under applicable conditions:

[0201]

[0202] Ensure that for each vehicle k, the total discharge in all intervals at charging station j is equal to the total discharge at charging station j:

[0203]

[0204] The total emissions of vehicle k along the route from node g to node h:

[0205]

[0206] Constraint (35) is the same as constraint (30), but only for discharge:

[0207]

[0208] The total loss cost of vehicle k discharging on the function arc(g,h):

[0209]

[0210] In the formula, J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o′ is another warehouse of the same type; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle's mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable. If vehicle k goes from node i to node j, the value is 1, otherwise it is 0; u′ k ' d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g;

[0211] S3. The three-stage efficient heuristic algorithm of CWIGALNS is used to solve the mathematical model of the wear cost function and obtain the optimal scheduling solution;

[0212] The three-stage efficient heuristic algorithms of CWIGALNS are Clarke and Wright preservation algorithm (CW), iterative greedy algorithm (IG) and adaptive large neighborhood search (ALNS), as follows:

[0213] S31. Clarke and Wright preservation algorithm is as follows:

[0214] (S31.1) Create a route for each customer;

[0215] (S31.2) identifying edge nodes directly connected to the warehouse;

[0216] (S31.3) Select two edge nodes and in the routes and as a preservation pair, calculate the preservation value and find the total demand of the new route;

[0217] (S31.4) selecting a saved pair in the SPL in order of saved values, identifying a route and a route including edge nodes and respectively;

[0218] (S31.5) connecting the routes and, if the total demand of the new route exceeds the loading capacity of the vehicle, abandoning the current merging operation, otherwise, ending the current merging operation;

[0219] (S31.6) Delete the saved pair from the program, and repeat the process from step (S31.4) to step (S31.6) until the program is empty;

[0220] S32. At the beginning of the iterative greedy algorithm, all charging stations will be deleted and reselected according to the placement cost. The specific parameters are as follows:

[0221] (S32.1) Breakpoint: Under the mileage limit, if the electric vehicle does not charge at a charging station during driving, it will not be able to reach certain nodes; the breakpoint v has the following properties: Define route k The first breakpoint on the * ;

[0222] (S32.2) Node accessibility: In order to compare the impact of different charging stations on customer points in the route, define the parameter q vk The accessibility of the customer points, Parameter q vk It is used to indicate whether the electric vehicle will run out of power on the way, and the minimum value of the path customer point accessibility is defined as q * ;

[0223] (S32.3) Charging strategy: If charging time and battery wear increase with charging status, electric vehicles need to recharge their batteries at the depot or charging station until they reach the next stop or return to the depot, and the minimum power is not less than Q min ; For scenarios with monotonically decreasing or general wear cost functions, the MIP solver Cplex or Gurobi can be called using JAVA language to directly calculate the charge amount of the electric vehicle;

[0224] (S32.4) Placement cost: on path r k Add a charging station j∈J to the node v on the network, denoted as v, and the placement cost is

[0225]

[0226] β1+β2+β3+β4+β5=1,β1,β2,β3,β4,β5≥0 (38)

[0227]

[0228]

[0229]

[0230] In the formula, is the increase in accessibility, is the additional routing cost, is the wear cost,

[0231] is the additional billable time cost, is the penalty value; q vk and q′ vk denote the reachability of v before and after insertion into site j, respectively; and Respectively represent the minimum value of the current path reachability before and after inserting j

[0232] In the S33.ALNS algorithm, the delete operator is used to delete n from the current route. * customer nodes and repeatedly use the insertion operator to reinsert them into the route, *Let it be a random value in [μ1|I|,μ2|I|], where μ1,μ2∈(0,1):

[0233] (S33.1) Assign s0 to the current solution s and the optimal solution s * ;

[0234] (S33.2) If the number of iterations is less than a given number, the insertion and removal operators are selected and s is assigned to the neighborhood solution s′, otherwise, the search ends;

[0235] (S33.3) Use the removal and insertion operators to process s′;

[0236] (S33.4) If the acceptance criteria are met at this time, the value s′ is assigned to s;

[0237] (S33.5) If z(s) <z(s * ), then assign s to s * ;

[0238] (S33.6) Update the score and weight of each operator and return to step (S33.2);

[0239] The adaptive search mechanism used by the ALNS algorithm is as follows:

[0240] At the beginning of each stage, the operator weights are adjusted and the operators are selected according to the preset rules to achieve adaptive update of the search process, as follows:

[0241] Let the weight of operator i in stage j be w ij , then the probability of the operator being selected is p ij =w ij / ∑ h∈H w hj , where H is the operator set; the jth stage requires updating the operator weights according to the following rules after each search stage:

[0242] Define the number of times operator I is used in stage j, and the scores are ∈ ij and π ij ; If operator i is used in stage j, its weight in the next stage is (1-θ)w ij +θπ ij / ∈ ij , otherwise the weight remains unchanged, where θ is the reaction coefficient, set to 0.3;

[0243] If the neighborhood solution s′ is better than the current solution s, the neighborhood solution is retained; otherwise, the probability that the neighborhood solution is retained is e -(z(s′)-z(s)) / T , where the initial temperature T0 = 10000, T n =cT n-1 , c=0.995.

[0244] In summary, the present invention first describes the actual charging process and battery wear; then, according to the wear cost and the monotonicity of the state of charge (SOC) under different conditions, a corresponding comprehensive mathematical model is constructed; finally, a three-stage efficient heuristic algorithm named CWIGALNS is proposed to solve the model, which can quickly and systematically analyze the collaborative optimization problem of electric vehicle delivery service and charging scheduling, establish an overall service decision-making plan based on nonlinear charging time function and battery degradation characteristics, and reduce logistics operation costs.

[0245] Embodiment 2:

[0246] This embodiment provides an optimal scheduling system for electric vehicle charging network design, which is used to implement the optimal scheduling method for electric vehicle charging network design described in the first embodiment, including:

[0247] Function definition module, used to define charging function and wear cost function according to the actual charging process of the battery and the battery wear;

[0248] A construction module is used to construct a mathematical model of a corresponding wear cost function according to a monotonically increasing or decreasing relationship between the wear cost and the charging state, and to set constraint conditions;

[0249] The solution output module is used to solve the mathematical model of the wear cost function using the three-stage efficient heuristic algorithm of CWIGALNS to obtain the optimal scheduling solution.

[0250] Specifically, the above-mentioned function definition module, construction module and solution output module can be embedded in a computer processing system. The computer calls the above-mentioned modules to complete the task of coordinated optimization of electric vehicle distribution service and charging scheduling based on the above-mentioned optimal scheduling method for electric vehicle charging network design; the above-mentioned function definition module, construction module and solution output module can perform operations according to the specific steps given in the above-mentioned optimal scheduling method for electric vehicle charging network design.

[0251] It should be noted that it should be understood that the division of the various modules of the above system is only the division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated, and these modules can all be implemented in the form of software calling through processing elements; they can also be all implemented in the form of hardware; some modules can also be implemented in the form of software calling through processing elements, and some modules can be implemented in the form of hardware. For example, the definition module can be a separately established processing element, or it can be integrated in a chip of the above device for implementation. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a processing element of the above device. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the hardware integrated logic circuit in the processor element or the instructions in the form of software.

[0252] For example, the above modules may be one or more integrated circuits configured to implement the above methods, such as one or more application specific integrated circuits (ASIC), or one or more digital singnal processors (DSP), or one or more field programmable gate arrays (FPGA). For another example, when a module is implemented in the form of a processing element scheduling program code, the processing element may be a general-purpose processor, such as a central processing unit (CPU) or other processors that can call program code. For another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0253] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0254] For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances. When an element is referred to as being "assembled on", "installed on", "fixed on" or "set on" another element, it can be directly on the other element or there can also be a centered element. When an element is considered to be "connected" to another element, it can be directly connected to the other element or there may be a centered element at the same time. The terms "vertical", "horizontal", "up", "down", "left", "right" and similar expressions used herein are only for illustrative purposes and are not intended to be the only implementation method.

[0255] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

[0256] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

Claims

1. An optimal scheduling method for electric vehicle charging network design, characterized in that: The following steps are involved: According to the actual charging process and wear of the battery, the charging function and the wear cost function are defined; According to the monotonically increasing or decreasing relationship between the wear cost and the charging state, a mathematical model of the corresponding wear cost function is constructed, and constraint conditions are set; The three-stage efficient heuristic algorithm of CWIGALNS is used to solve the mathematical model of the wear cost function and obtain the optimal scheduling solution.

2. The optimal scheduling method for electric vehicle charging network design according to claim 1 is characterized in that: According to the actual battery charging process and battery wear, the charging function and wear cost function are defined as follows: (21) The changes in current and voltage during the actual battery charging process are nonlinear, so the charging function including the constant current and constant voltage stages is nonlinear; (22) Assume that each charger has a specific constant current-constant voltage charging function with b+1 breakpoints, and fit the real constant current-constant voltage concave function. Let the charging state associated with breakpoint i be S i , where i∈B, B={0,…,b}; Assuming that the charging current between consecutive breakpoints is constant, the charging function is: T=∑ d∈{1,2,3} t d S d (1) Where, T is the total charging time, t d S is the unit charging time of the battery charging state in interval d, d The battery charge state is in interval d; Let D be the set of charging state time intervals, and let L be the length of each time interval: D={1,2,...,n d } (3) In the formula, and S d They are the upper and lower limits of the charging state respectively; Battery Price (W P ) can be expressed as: IN P =2·ACC(ADD)·∑ d∈D: S d ≥1-ADD(IN( S d )·Δ q ) (4) DOD∈{L,2L,3L,...,1} (5) In the formula, W( S d ) is in The cost of charging and discharging per kWh, Δ q It is the energy value of each charging state interval, DOD is the depth of discharge, and ACC is the battery cycle count.

3. The optimal scheduling method for electric vehicle charging network design according to claim 2 is characterized in that: According to the monotonically increasing relationship between wear cost and charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set as follows: A mixed integer linear programming formula is proposed to minimize the running cost as follows: Where, d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable, if vehicle k goes from node i to node j, then the value is 1, otherwise it is 0; S ′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h successively at charging station j; W d The wear cost per kilowatt-hour of charging and discharging in interval d; t d is the charging time per unit of electricity in interval d; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the number of intervals d∈D where vehicle k charges in the parking lot. (32) The specific constraints are as follows: Each customer is served only once: The vehicle is charged at a location with charging facilities: Traffic flow balance at each node: Vehicle departure and return from the depot: Each electric vehicle serves at most one route: The remaining cargo quantity before and after the electric vehicle passes a certain node: Range of remaining cargo in electric vehicle: The remaining energy of the electric vehicle at each node on the service route: The remaining power of the electric vehicle when it leaves the distribution center: At the client node, the battery charge remains unchanged: The vehicle's charge level before and after leaving the charging station: Upper and lower limits of charge capacity: Make sure the vehicle has enough charge to complete the assigned task: Ensure that the total energy charged by charging station j in all charging state intervals is equal to the total charge amount of charging station j to vehicle k: Formula (21)-Formula (23) ensures that sufficient power is charged in each charging state interval: The energy consumption cost of vehicle k charging at the charging station is: The charging time cost of the energy charged to vehicle k at the charging station: Formula (26)-Formula (27) ensures that within the charging state interval, the variable u gjhkd and u′ kd must be equivalent to a charging state interval whose upper limit is higher than the charge of vehicle k before it arrives at charging station j or leaves the charging station: Binary constraints: Where J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o ′ is the same warehouse; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable, which takes the value 1 if vehicle k goes from node i to node j, otherwise it takes the value 0; u ′ k ′ d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S ′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g.

4. The optimal scheduling method for electric vehicle charging network design according to claim 1 is characterized in that: According to the monotonically decreasing relationship between wear cost and charging state, a mathematical model of the corresponding wear cost function is constructed, and constraints are set as follows: (41) Define variables is the amount of electricity released by vehicle k on the function arg(g,h) in the charging state interval d∈D; variable S ′ g ′ hk is the total wear cost of vehicle k on the function arg(g,h); If the battery of vehicle k is discharged on the function arg(g,h) over interval d, then the binary variable is equal to 1; represents the total emission of vehicle k on the function arg(g,h), the specific formula is as follows: (42) The above function (29) is subject to the constraints of formula (7)-formula (25) and formula (28), and is also subject to the following constraints: Formula (31)-Formula (32) constrains the maximum acceptable interval when vehicle k leaves station j or depot o by calculating the difference between its lower limit and the current SOC: Ensure that the maximum discharge energy per SOC interval does not exceed L under applicable conditions: Ensure that for each vehicle k, the total discharge in all SOC intervals at charging station j is equal to the total discharge at charging station j: The total emissions of vehicle k along the route from node g to node h: Constraint (35) is the same as constraint (30), but only for discharge: The total loss cost of vehicle k discharging on the function arc(g,h): Where J is the set of candidate stations; I is the set of customers; K is the set of vehicles; D is the set of charging state intervals; o is a single warehouse; o ′ is the same warehouse; V is the vertex set; α is the battery wear cost coefficient; ∈ is the vehicle segment charging time cost coefficient, γ is the path cost per unit distance, Q is the vehicle mileage, and the remaining power must be maintained at a specific minimum and maximum value Q min and Q max Between; d gh is the distance from node g to node h; f j is the unit charging station construction cost of node j; y j is a binary variable, which takes the value 1 if the site is located at node j, and 0 otherwise; ijk is a binary variable, which takes the value 1 if vehicle k goes from node i to node j, otherwise it takes the value 0; u ′ k ′ d is a binary variable, which takes the value of 1 if vehicle k is charged in the vehicle segment using the time interval d, and 0 otherwise; u gjhkd is a binary variable that takes the value 1 if vehicle k is charged at charging station j using interval d and passes through nodes g, j, and h consecutively, and 0 otherwise; g gjhk The total amount of money charged by vehicle k at charging station j when it passes through nodes g, j and h successively; c ghk is the remaining load when vehicle k leaves node g and arrives at node h; S gjhkd The number of battery charges for vehicle k at charging station j and passing through nodes g, j and h successively; S ′ gjhk is the total wear cost of vehicle k that passes through nodes g, j, and h consecutively at charging station j; is the number of intervals d∈D where vehicle k is charged in the parking lot; T gjhk is the charging time cost of vehicle k at charging station j when it passes through nodes g, j, and h continuously; is the maximum distance allowed by the remaining power when vehicle k leaves node g to h; is the maximum distance allowed by the remaining power when vehicle k reaches node h from g.

5. The optimal scheduling method for electric vehicle charging network design according to claim 4 is characterized in that: The three-stage efficient heuristic algorithm of CWIGALNS includes Clarke and Wright preservation algorithm, iterative greedy algorithm and adaptive large neighborhood search algorithm.

6. The optimal scheduling method for electric vehicle charging network design according to claim 5 is characterized in that: The Clarke and Wright preservation algorithm is as follows: (61) Create a route (oro) for each customer r; (62) Identify the edge nodes directly connected to warehouse o; (63) Select two edge nodes r1 and r2 in routes a1 and a2 as the saved pair and calculate the saved value and find the total demand for new routes; (64) Select the stored pair (r1, r2) in the SPL in the order of stored values, and identify the route a1 and route a2 containing the edge nodes r1 and r2 respectively; (65) Connect routes a1 and a2. If the total demand of the new route exceeds the vehicle's loading capacity, abandon the merging operation. Otherwise, end the merging operation. (66) Delete the saved pair (r1, r2) from the program and repeat the process from step (64) to step (66) until the program is empty.

7. The optimal scheduling method for electric vehicle charging network design according to claim 5 is characterized in that: At the beginning of the iterative greedy algorithm, all charging stations are deleted and reselected according to the placement cost. The specific parameters are as follows: (71) Breakpoint: Under the mileage limit, if the electric vehicle does not charge at a charging station during driving, it will not be able to reach certain nodes; the breakpoint v has the following properties: Define route k The first breakpoint on the * ; (72) Node accessibility: In order to compare the impact of different charging stations on customer points in the route, we define the parameter q vk The accessibility of the customer points, j,v∈r k , parameter q vk It is used to indicate whether the electric vehicle will run out of power on the way, and the minimum value of the path customer point accessibility is defined as q * ; (73) Charging strategy: If charging time and battery wear increase with the charging state, electric vehicles need to recharge their batteries at the depot or charging station until they reach the next stop or return to the depot, and the minimum power is not less than Q min ; (74) Placement cost: on path r k Add a charging station j∈J to the node v on the network, denoted as v, and the placement cost is β1+β2+β3+β4+β5＝1,β1,β2,β3,β4,β5≥0 (38) In the formula, is the increase in accessibility, is the additional routing cost, is the wear cost, is the additional billable time cost, is the penalty value; q vk and q′ vk denote the reachability of v before and after insertion into site j, respectively; and They represent the minimum value of the current path reachability before and after inserting j.

8. The optimal scheduling method for electric vehicle charging network design according to claim 5 is characterized in that: In the ALNS algorithm, the delete operator is used to delete n from the current route. * customer nodes and repeatedly use the insertion operator to reinsert them into the route, * Let it be a random value in [μ1|I|,μ2|I|], where μ1,μ2∈(0,1): (81) Assign s0 to the current solution s and the optimal solution s * ; (82) If the number of iterations is less than a given number, the insertion and removal operators are selected and s is assigned to the neighborhood solution s′, otherwise, the search ends; (83) Use the removal and insertion operators to process s′; (84) If the acceptance criteria are met at this time, the value s′ is assigned to s; (85) If z(s) <z(s * ), then assign s to s * ; (86) Update the score and weight of each operator and return to step (82); The adaptive search mechanism used by the ALNS algorithm is as follows: At the beginning of each stage, the operator weights are adjusted and the operators are selected according to the preset rules to achieve adaptive update of the search process, as follows: Let the weight of operator i in stage j be w ij , then the probability of the operator being selected is p ij =w ij / ∑ h∈H w hj , where H is the operator set; the jth stage requires updating the operator weights according to the following rules after each search stage: Define the number of times operator I is used in stage j, and the scores are ∈ ij and π ij ; If operator i is used in stage j, its weight in the next stage is (1-θ)w ij +θπ ij / ∈ ij , otherwise the weight remains unchanged, where θ is the reaction coefficient, set to 0.3; If the neighborhood solution s′ is better than the current solution s, the neighborhood solution is retained; otherwise, the probability that the neighborhood solution is retained is e -(z(s′)-z(s)) / T , where the initial temperature T0 = 10000, T n =cT n-1 , c=0.

995.

9. An optimal scheduling system for electric vehicle charging network design, used to implement the optimal scheduling method for electric vehicle charging network design as described in any one of claims 1 to 8, characterized in that: include: Function definition module, used to define charging function and wear cost function according to the actual charging process of the battery and the battery wear; A construction module is used to construct a mathematical model of a corresponding wear cost function according to a monotonically increasing or decreasing relationship between the wear cost and the charging state, and to set constraint conditions; The solution output module is used to solve the mathematical model of the wear cost function using the three-stage efficient heuristic algorithm of CWIGALNS to obtain the optimal scheduling solution.

Images