Electric logistics vehicle charging station site selection and path optimization method in urban distribution

Abstract

The invention discloses an electric logistics vehicle charging station site selection and path optimization method in urban distribution. The method comprises the following steps: S1, determining a precondition hypothesis for constructing a semi-open multi-distribution center electric logistics vehicle path optimization model; s2, on the basis of considering the position of the charging station, constructing a semi-open multi-distribution-center electric logistics vehicle path optimization model; the semi-open multi-distribution-center electric logistics vehicle path optimization model comprises a charging station queuing model, a target function and constraint conditions; s3, improving an artificial bee colony algorithm; s4, substituting the specific parameters into an artificial bee colony algorithm to solve an optimal path at the moment; and S5, sensitivity analysis is carried out, different battery capacities and charging quotas are substituted into an artificial bee colony algorithm to solve an optimal path at the moment, and then comparison is carried out. According to the method, a semi-open multi-distribution-center electric logistics vehicle path optimization model is constructed, an improved artificial bee colony algorithm is adopted for solving, and an optimal distribution scheme under different battery capacities and charging limits is obtained.

Description

Technical Field

[0001] The present invention relates to a vehicle path planning method, and in particular to a method for site selection and path optimization of charging stations for electric logistics vehicles in urban distribution. Background Art

[0002] In recent years, the logistics industry, as a basic and leading industry supporting the development of the national economy, has developed steadily and positively.

[0003] In the prior art, many logistics companies use electric logistics vehicles for distribution operations. However, due to the problems of short driving range and difficulty in charging, charging equipment must be taken into consideration in distribution operations. Therefore, the path optimization research method of electric logistics vehicles (Electric Vehicle Route Problem, EVRP) has received extensive attention from scholars and companies. In previous inventions, although some studies have taken the location of charging facilities into consideration, the charging facilities studied are all fixed locations or candidate locations and their construction costs are set to the same value. In actual situations, the construction cost of charging stations will change according to the price of land. The prior art lacks a vehicle path planning research method for such scenarios. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art, provide a method for site selection and path optimization of charging stations for electric logistics vehicles in urban distribution, establish a semi-open multi-distribution center electric logistics vehicle path optimization model (MDHOEVRPTW), take the minimum total cost as the goal, consider the constraints such as queuing waiting time, vehicle load, road congestion, etc., select an improved artificial bee colony algorithm and solve the problem, and obtain the optimal distribution plan under different battery capacities and charging quotas.

[0005] The object of the present invention is achieved by: a method for site selection and path optimization of charging stations for electric logistics vehicles in urban distribution, comprising the following steps:

[0006] Step S1: Determine the premise assumptions for constructing a semi-open multi-distribution center electric logistics vehicle path optimization model;

[0007] Step S2: Based on the location of the charging station, a semi-open multi-distribution center electric logistics vehicle path optimization model is constructed; the semi-open multi-distribution center electric logistics vehicle path optimization model includes a charging station queuing model, an objective function and constraints;

[0008] Step S3: improving the artificial bee colony algorithm;

[0009] Step S4: Bring the specific parameters into the artificial bee colony algorithm to solve the optimal path at this time;

[0010] Step S5: Perform sensitivity analysis, bring different battery capacities and charging quotas into the artificial bee colony algorithm to solve the optimal path at this time, and then compare them.

[0011] As a further limitation of the present invention, the premise assumptions of the semi-open multi-distribution center electric logistics vehicle path optimization model in step S1 include:

[0012] (1) The maximum load weight and other information of the electric logistics vehicle are known, and there is only one model;

[0013] (2) The locations of distribution centers and demand points are known;

[0014] (3) Each customer can only be served once by an electric logistics vehicle;

[0015] (4) Delivery activities are limited to time windows. If electric logistics vehicles arrive earlier or later than the time required by the customer, they must pay certain opportunity costs or penalty costs;

[0016] (5) All electric logistics vehicles depart from the distribution center and return to any distribution center after completing the distribution task;

[0017] (6) The model and quantity of charging equipment at each charging station are the same, and the construction area of ​​the charging stations is the same;

[0018] (7) The distribution center’s inventory capacity and electric logistics vehicles can meet customer demand in the distribution area;

[0019] (8) The charging rate of the electric logistics vehicle remains unchanged during charging.

[0020] As a further limitation of the present invention, the charging station queue model in step S2 includes the service intensity of the charging station, the solution when the queue system reaches a stable state, the average queue length and the average queue time;

[0021] The service intensity of the charging station is expressed as: Where g represents the service intensity of the charging station, μ represents the charging intensity of a charging pile, the number of charging piles at the charging station is m, and the law of vehicle arrival follows the Poisson distribution with an arrival rate of λ;

[0022] The number of electric logistics vehicles that need to be charged is n; the solution when the queuing system reaches a stable state is: p n represents the probability of the queue system being in different states, P 0 is the probability that the queue system is in the initial state, P 1 is the probability that there is an electric logistics vehicle in the queuing system that needs to be charged; P n+1 is the probability that there are n+1 electric logistics vehicles that need to be charged in the queuing system;

[0023]

[0024] The average queue length is L q :

[0025] The average queuing time is W q :

[0026] As a further limitation of the present invention, the objective function in step S2 is Z=min(Z 1 +Z 2 +Z 3 +Z 4 +Z 5 +Z 6 ), which consists of six objective sub-functions;

[0027] Z 1 is the investment cost of the charging station, Where S is the set of all candidate charging stations, S∈{1,…,α}; c η is the investment cost of the charging station η, z η is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0; c η is the investment cost of the charging station η, which includes land construction cost, charging pile construction cost and operation and maintenance cost; α represents the construction area of ​​the charging station; p η represents the land base price of charging station η, 10,000 yuan / m 2 ; ω represents the number of charging piles in the charging station; σ represents the unit price of the charging pile, 10,000 yuan / unit; q 0 represents the discount rate; θ represents the service life of the charging pile; the operation and maintenance cost of the charging station is linearly related to the construction cost, where τ represents the cost ratio coefficient;

[0028] Z 2 The fixed cost of using electric logistics vehicles is: Where H is the set of electric logistics vehicles H∈{1,2,…,k}, P is the set of distribution centers P∈{1,2,…,m}, and the fixed cost of using electric logistics vehicles includes the daily maintenance cost, depreciation cost and labor cost of distribution personnel. 1 represents the unit fixed usage cost of the electric logistics vehicle; is a decision variable. When distribution center h sends out electric logistics vehicle k, is 1, otherwise it is 0;

[0029] Z 3 The transportation cost of electric logistics vehicles is: Where V is the set of all vertices; b2 is the unit transportation cost of the electric logistics vehicle; d ij For vertex v i and vertex v j The distance between is a decision variable, which means that when electric logistics vehicle k moves from node i to node j, is 1, otherwise it is 0;

[0030] Z 4 Charging costs for electric logistics vehicles: Among them, S is the set of all candidate charging stations; H is the set of electric logistics vehicles; b 3 is the unit electricity price; is the charging amount of electric logistics vehicle k at charging station η; z η It is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0;

[0031] Z 5 Cost of carbon emissions: Among them, FC is the power consumption of the electric logistics vehicle when traveling 100 kilometers, unit: KWH / km, μ is the charging efficiency, and f is the carbon emission factor when generating electricity, unit: (CO 2 )kg,C tax represents carbon tax;

[0032] Z 6 The penalty cost for violating the time window is: Where C is the customer set C∈{1,2,…,n}, E i is the earliest arrival time of the goods expected by customer i, L i is the earliest arrival time of the goods expected by customer i, is the time when electric logistics vehicle k arrives at customer i; b 4 The unit penalty cost for the electric logistics vehicle to arrive at the customer point early; b 5 Unit penalty cost for electric logistics vehicles arriving late at customer points.

[0033] As a further limitation of the present invention, the constraint conditions in step S2 specifically include:

[0034] (1) Customer needs are indivisible, and each customer can only be served once: is a decision variable, which means that when electric logistics vehicle k moves from node i to node j, is 1, otherwise it is 0;

[0035] (2) The electric logistics vehicle only departs from its distribution center once:

[0036] (3) Only the selected charging stations can provide charging services for electric logistics vehicles, M is equal to 1×10 9 : z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0;

[0037] (4) Indicates that the total weight of the goods transported by the electric logistics vehicle cannot exceed its load capacity: r i is the demand of customer point i;

[0038] (5) Indicates that the electric logistics vehicle is fully loaded when it leaves the distribution center: is the transportation volume of electric logistics vehicle k moving from node i to node j;

[0039] (6) Every customer’s needs will be met:

[0040] (7) Indicates that the electric logistics vehicle will not go from the distribution center to the charging station or another distribution center:

[0041] (8) Indicates that the process of providing services to customers is continuous: is the time when electric logistics vehicle k arrives at customer point i; is the time when the electric logistics vehicle k leaves the customer point i; t i The service time required to serve customer point i; W q Average waiting time in queues for charging electric logistics vehicles; is the charging time of electric logistics vehicle k at charging station η; S is the set of candidate charging stations S∈{1,2,…,α}; d ij is the distance between node i and node j; v 0 is the driving speed of the electric logistics vehicle; e ij is the traffic congestion coefficient between node i and node j, e ij =e ji ;

[0042] (9) represents the charging amount required for electric logistics vehicles: The charging amount of electric logistics vehicle k at charging station η; b 6 is the charging coefficient of the electric logistics vehicle; z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0;

[0043] (10) Indicates that the electric logistics vehicle is fully charged when it leaves the distribution center or charging station: Pi k is the power of electric logistics vehicle k when it leaves node i; b 7 is the battery capacity of the electric logistics vehicle;

[0044] (11) indicates that the power of the electric logistics vehicle remains unchanged before and after serving the customer: is the power of electric logistics vehicle k when it arrives at node i;

[0045] (12) indicates that the power of the electric logistics vehicle at any customer point must be greater than the minimum reserved power: b 8 Provide the lowest reserved power for electric logistics vehicles;

[0046] (13) represents the remaining power of the electric logistics vehicle when it arrives at the node: b 9 is the power consumption coefficient of the electric logistics vehicle;

[0047] (14) indicates the value range of the variable:

[0048] As a further limitation of the present invention, step S3 specifically includes:

[0049] S3-1 Initialization: Set the size of the bee population. The number of employed bees, observer bees, and scout bees is the same, all SN. Let iter be the maximum number of iterations of the algorithm. Give each bee a perception distance Visual, detal is the crowding factor, Try_number is the maximum number of foraging times, P c Denoted as the crossover probability, P m Denoted as the mutation probability, P m1 To exchange the probability of mutation, SN initial solutions are randomly generated;

[0050] S3-2 Hired bee stage: Hired bees conduct field search and generate new hired bees. If the new hired bees have better solutions than the old hired bees, the new hired bees replace the original hired bees. Then the artificial fish swarm algorithm performs swarming behavior. If the swarming is successful, it determines whether the solution at this time is better. If it is better, it replaces the current solution. Otherwise, the current solution remains unchanged. If the swarming fails, the foraging behavior is performed.

[0051] S3-3 Observer bee stage: The employed bees are selected to follow according to the roulette wheel, and the field search is carried out in the field. If a better solution is found, the new observer bee replaces the original observer bee. Otherwise, the crossover and mutation operations in the genetic algorithm are performed to determine whether a better solution is generated. If so, it is replaced, otherwise it remains unchanged.

[0052] S3-4 Scouting bee stage: If a hired bee has not updated to a better solution when the number of iterations reaches the threshold, the hired bee is transformed into a scout bee and performs a field search in its field to randomly generate a new solution;

[0053] S3-5 determines whether the number of iterations has reached the maximum. If not, continue to loop through steps S3-2 to S3-4 until the loop ends and the final result is output.

[0054] The present invention adopts the above technical scheme, and compared with the prior art, the beneficial effects are as follows: the present invention is based on the commonly used vehicle routing planning problem (VRP), and on this basis constructs a semi-open multi-distribution center electric logistics vehicle path optimization model (MDHOEVRPTW), which adopts an improved artificial bee colony algorithm for solving and obtains the optimal distribution plan under different battery capacities and charging amounts; the present invention starts from the perspective of combining the site selection of charging stations with the transportation routes of electric logistics vehicles, and under the current shared resources, enables enterprises to respond to customer needs more quickly, improve customer satisfaction and loyalty, and lay a solid foundation for continuous optimization and upgrading within the enterprise and sustainable development. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0056] like Figure 1 A method for selecting a charging station and optimizing a route for an electric logistics vehicle in urban distribution is shown, comprising the following steps:

[0057] Step S1: Determine the premise assumptions for constructing a semi-open multi-distribution center electric logistics vehicle path optimization model;

[0058] The premise assumptions of the semi-open multi-distribution center electric logistics vehicle route optimization model include:

[0059] (1) The maximum load weight and other information of the electric logistics vehicle are known, and there is only one model;

[0060] (2) The locations of distribution centers and demand points are known;

[0061] (3) Each customer can only be served once by an electric logistics vehicle;

[0062] (4) Delivery activities are limited to time windows. If electric logistics vehicles arrive earlier or later than the time required by the customer, they must pay certain opportunity costs or penalty costs;

[0063] (5) All electric logistics vehicles depart from the distribution center and return to any distribution center after completing the distribution task;

[0064] (6) The model and quantity of charging equipment at each charging station are the same, and the construction area of ​​the charging stations is the same;

[0065] (7) The distribution center’s inventory capacity and electric logistics vehicles can meet customer demand in the distribution area;

[0066] (8) The charging rate of the electric logistics vehicle remains unchanged during charging.

[0067] Step S2: Based on the location of the charging station, a semi-open multi-distribution center electric logistics vehicle path optimization model is constructed; the semi-open multi-distribution center electric logistics vehicle path optimization model includes a charging station queuing model, an objective function and constraints;

[0068] The charging station queue model includes the service intensity of the charging station, the solution when the queue system reaches a stable state, the average queue length and the average queue time;

[0069] The service intensity of the charging station is expressed as: Where g represents the service intensity of the charging station, μ represents the charging intensity of a charging pile, the number of charging piles at the charging station is m, and the law of vehicle arrival follows the Poisson distribution with an arrival rate of λ;

[0070] The solution when the queuing system reaches a stable state is: p n represents the probability of the queue system being in different states, P 0 is the probability that the queue system is in the initial state, P 1 is the probability that there is an electric logistics vehicle in the queuing system that needs to be charged; P n+1 is the probability that there are n+1 electric logistics vehicles that need to be charged in the queuing system;

[0071]

[0072] The average queue length is L q :

[0073] The average waiting time is W q :

[0074] The objective function is Z = min(Z 1 +Z 2 +Z 3 +Z 4 +Z 5 +Z 6 ), which consists of six objective sub-functions;

[0075] Z 1 is the investment cost of the charging station, Where S is the set of all candidate charging stations, S∈{1,…,α}; c η is the investment cost of the charging station η, z η is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0; c η is the investment cost of the charging station η, which includes land construction cost, charging pile construction cost and operation and maintenance cost; α represents the construction area of ​​the charging station; p η represents the land base price of charging station η (10,000 yuan / m 2 );ω represents the number of charging piles in the charging station;σ represents the unit price of the charging pile (10,000 yuan / unit);q 0 represents the discount rate; θ represents the service life of the charging pile; the operation and maintenance cost of the charging station is linearly related to the construction cost, where τ represents the cost ratio coefficient;

[0076] Z 2 The fixed cost of using electric logistics vehicles is: Where H is the set of electric logistics vehicles H∈{1,2,…,k}, P is the set of distribution centers P∈{1,2,…,m}, and the fixed cost of using electric logistics vehicles includes the daily maintenance cost, depreciation cost and labor cost of distribution personnel. 1 represents the unit fixed usage cost of the electric logistics vehicle; is a decision variable. When distribution center h sends out electric logistics vehicle k, is 1 if the value is true, otherwise it is 0.

[0077] Z 3 The transportation cost of electric logistics vehicles is: Where V is the set of all vertices; b 2 is the unit transportation cost of the electric logistics vehicle; d ij For vertex v i and vertex v j The distance between is a decision variable, which means that when the electric logistics vehicle k of the distribution center h moves from node i to node j, is 1, otherwise it is 0;

[0078] Z 4 Charging costs for electric logistics vehicles: Among them, S is the set of all candidate charging stations; H is the set of electric logistics vehicles; b 3 is the unit electricity price; is the charging amount of electric logistics vehicle k at charging station η; z η It is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0;

[0079] Z 5 Cost of carbon emissions: Among them, FC is the power consumption of the electric logistics vehicle when traveling 100 kilometers, unit: KWH / km, μ is the charging efficiency, and f is the carbon emission factor when generating electricity, unit: (CO 2 )kg,C tax represents carbon tax;

[0080] Z 6 The penalty cost for violating the time window is: Where C is the customer set C∈{1,2,…,n}, E i is the earliest arrival time of the goods expected by customer i, L i is the earliest arrival time of the goods expected by customer i, is the time when electric logistics vehicle k arrives at customer i; b 4 The unit penalty cost for the electric logistics vehicle to arrive at the customer point early; b 5 Unit penalty cost for electric logistics vehicles arriving late at customer points.

[0081] The constraints include:

[0082] (1) Customer needs are indivisible, and each customer can only be served once: is a decision variable, which means that when electric logistics vehicle k moves from node i to node j, is 1, otherwise it is 0;

[0083] (2) The electric logistics vehicle only departs from its distribution center once:

[0084] (3) Only the selected charging stations can provide charging services for electric logistics vehicles. M represents a large number, equal to 1×10 9 : z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0;

[0085] (4) Indicates that the total weight of the goods transported by the electric logistics vehicle cannot exceed its load capacity: r i is the demand of customer point i;

[0086] (5) Indicates that the electric logistics vehicle is fully loaded when it leaves the distribution center: is the transportation volume of electric logistics vehicle k moving from node i to node j;

[0087] (6) Every customer’s needs will be met:

[0088] (7) Indicates that the electric logistics vehicle will not go from the distribution center to the charging station or another distribution center:

[0089] (8) Indicates that the process of providing services to customers is continuous: is the time when electric logistics vehicle k arrives at customer point i; is the time when the electric logistics vehicle k leaves the customer point i; t i The service time required to serve customer point i; W q Average waiting time in queues for charging electric logistics vehicles; is the charging time of electric logistics vehicle k at charging station η; S is the set of candidate charging stations S∈{1,2,…,α}; d ij is the distance between node i and node j; v 0 is the driving speed of the electric logistics vehicle; e ij is the traffic congestion coefficient between node i and node j, e ij =e ji ;

[0090] (9) represents the charging amount required for electric logistics vehicles: The charging amount of electric logistics vehicle k at charging station η; b 6 is the charging coefficient of the electric logistics vehicle; z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0;

[0091] (10) Indicates that the electric logistics vehicle is fully charged when it leaves the distribution center or charging station: P i k is the power of electric logistics vehicle k when it leaves node i; b 7 is the battery capacity of the electric logistics vehicle;

[0092] (11) indicates that the power of the electric logistics vehicle remains unchanged before and after serving the customer: is the power of electric logistics vehicle k when it arrives at node i;

[0093] (12) indicates that the power of the electric logistics vehicle at any customer point must be greater than the minimum reserved power: b 8 Provide the lowest reserved power for electric logistics vehicles;

[0094] (13) represents the remaining power of the electric logistics vehicle when it arrives at the node: b 9 is the power consumption coefficient of the electric logistics vehicle;

[0095] (14) indicates the value range of the variable:

[0096] Step S3: improving the artificial bee colony algorithm;

[0097] S3-1 Initialization: Set the size of the bee population. The number of employed bees, observer bees, and scout bees is the same, all SN. Let iter be the maximum number of iterations of the algorithm. Give each bee a perception distance Visual, detal is the crowding factor, Try_number is the maximum number of foraging times, P c Denoted as the crossover probability, P m Denoted as the mutation probability, P m1 To exchange the probability of mutation, SN initial solutions are randomly generated;

[0098] S3-2 Hired bee stage: Hired bees conduct field search and generate new hired bees. If the new hired bees have better solutions than the old hired bees, the new hired bees replace the original hired bees. Then the artificial fish swarm algorithm performs swarming behavior. If the swarming is successful, it determines whether the solution at this time is better. If it is better, it replaces the current solution. Otherwise, the current solution remains unchanged. If the swarming fails, the foraging behavior is performed.

[0099] S3-3 Observer bee stage: The employed bees are selected to follow according to the roulette wheel, and the field search is carried out in the field. If a better solution is found, the new observer bee replaces the original observer bee. Otherwise, the crossover and mutation operations in the genetic algorithm are performed to determine whether a better solution is generated. If so, it is replaced, otherwise it remains unchanged.

[0100] S3-4 Scouting bee stage: If a hired bee has not updated to a better solution when the number of iterations reaches the threshold, the hired bee is transformed into a scout bee and performs a field search in its field to randomly generate a new solution;

[0101] S3-5 determines whether the number of iterations has reached the maximum. If not, continue to loop through steps S3-2 to S3-4 until the loop ends and the final result is output.

[0102] Step S4: Bring the specific parameters into the artificial bee colony algorithm to solve the optimal path at this time;

[0103] Step S5: Perform sensitivity analysis, bring different battery capacities and charging quotas into the artificial bee colony algorithm to solve the optimal path at this time, and then compare them.

[0104] Based on a series of factors such as the different corporate customer points, charging stations, distribution center points, the load capacity of electric logistics vehicles, the unit fixed cost of electric logistics vehicles, the unit transportation cost of electric logistics vehicles, etc., an improved artificial bee colony algorithm is used to generate the optimal path, at which point the total cost is minimized.

[0105] The present invention is not limited to the above-mentioned embodiments. On the basis of the technical solution disclosed in the present invention, technicians in this field can make some substitutions and deformations to some technical features therein according to the disclosed technical content without creative labor, and these substitutions and deformations are all within the protection scope of the present invention.

Claims

1. A method for site selection and route optimization of electric logistics vehicle charging stations in urban distribution, characterized in that: The following steps are involved: Step S1: Determine the premise assumptions for constructing a semi-open multi-distribution center electric logistics vehicle path optimization model; Step S2: Based on the location of the charging station, a semi-open multi-distribution center electric logistics vehicle path optimization model is constructed; the semi-open multi-distribution center electric logistics vehicle path optimization model includes a charging station queuing model, an objective function and constraints; Step S3: improving the artificial bee colony algorithm; Step S4: Bring the specific parameters into the artificial bee colony algorithm to solve the optimal path at this time; Step S5: Perform sensitivity analysis, bring different battery capacities and charging quotas into the artificial bee colony algorithm to solve the optimal path at this time, and then compare them.

2. According to claim 1, a method for site selection and route optimization of charging stations for electric logistics vehicles in urban distribution, characterized in that: The premise assumptions of the semi-open multi-distribution center electric logistics vehicle path optimization model in step S1 include: (1) The maximum load weight and other information of the electric logistics vehicle are known, and there is only one model; (2) The locations of distribution centers and demand points are known; (3) Each customer can only be served once by an electric logistics vehicle; (4) Delivery activities are limited to time windows. If electric logistics vehicles arrive earlier or later than the time required by the customer, they must pay certain opportunity costs or penalty costs; (5) All electric logistics vehicles depart from the distribution center and return to any distribution center after completing the distribution task; (6) The model and quantity of charging equipment at each charging station are the same, and the construction area of ​​the charging stations is the same; (7) The distribution center’s inventory capacity and electric logistics vehicles can meet customer demand in the distribution area; (8) The charging rate of the electric logistics vehicle remains unchanged during charging.

3. According to a method for site selection and route optimization of electric logistics vehicle charging stations in urban distribution in claim 1, it is characterized in that: The charging station queue model in step S2 includes the service intensity of the charging station, the solution when the queue system reaches a stable state, the average queue length and the average queue time; The service intensity of the charging station is expressed as: Where g represents the service intensity of the charging station, μ represents the charging intensity of a charging pile, the number of charging piles at the charging station is m, and the law of vehicle arrival follows the Poisson distribution with an arrival rate of λ; The number of electric logistics vehicles that need to be charged is n; the solution when the queuing system reaches a stable state is: p n represents the probability of the queue system being in different states, P0 is the probability that the queue system is in the initial state, P1 is the probability that there is an electric logistics vehicle in the queuing system that needs to be charged; P n+1 is the probability that there are n+1 electric logistics vehicles that need to be charged in the queuing system; The average queue length is L q : The average queuing time is W q :

4. According to a method for site selection and route optimization of electric logistics vehicle charging stations in urban distribution in claim 1, it is characterized in that: The objective function in step S2 is Z=min(Z1+Z2+Z3+Z4+Z5+Z6), which is composed of six objective sub-functions; Z1 is the investment cost of the charging station, Where S is the set of all candidate charging stations, S∈{1,…,α}; c η is the investment cost of the charging station η, z η is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0; c η is the investment cost of the charging station η, which includes land construction cost, charging pile construction cost and operation and maintenance cost; α represents the construction area of ​​the charging station; p η represents the land base price of charging station η, 10,000 yuan / m 2 ; ω represents the number of charging piles in the charging station; σ represents the unit price of the charging pile, 10,000 yuan / unit; q0 represents the discount rate; θ represents the service life of the charging pile; the operation and maintenance cost of the charging station shows a linear relationship with the construction cost, where τ represents the cost ratio coefficient; Z2 is the fixed cost of using the electric logistics vehicle: Where H is the set of electric logistics vehicles H∈{1, 2, …, k}, P is the set of distribution centers P∈{1, 2, …, m}, the fixed use cost of electric logistics vehicles includes the daily maintenance cost of the vehicle, depreciation cost and labor cost of the distribution personnel, and b1 represents the unit fixed use cost of the electric logistics vehicle; is a decision variable. When distribution center h sends out electric logistics vehicle k, is 1, otherwise it is 0; Z3 is the transportation cost of the electric logistics vehicle: Where V is the set of all vertices; b2 is the unit transportation cost of the electric logistics vehicle; d ij For vertex v i and vertex v j The distance between is a decision variable, which means that when electric logistics vehicle k moves from node i to node j, is 1, otherwise it is 0; The charging cost of Z4 electric logistics vehicle is: Among them, S is the set of all alternative charging stations; H is the set of electric logistics vehicles; b3 is the unit electricity price; is the charging amount of electric logistics vehicle k at charging station η; z η It is a 0-1 variable, which is 1 when the charging station is selected, otherwise it is 0; Z5 is the cost of carbon emissions: Among them, FC is the power consumption of the electric logistics vehicle when traveling 100 kilometers, unit: KWH / km, μ is the charging efficiency, f is the carbon emission factor when generating electricity, unit: (CO2)kg, C tax represents carbon tax; Z6 is the penalty cost for violating the time window: Where C is the customer set C∈{1,2,…,n}, E i is the earliest arrival time of the goods expected by customer i, L i is the earliest arrival time of the goods expected by customer i, is the time when electric logistics vehicle k arrives at customer i; b4 is the unit penalty cost of the electric logistics vehicle arriving at the customer point early; b5 is the unit penalty cost of the electric logistics vehicle arriving at the customer point late.

5. According to a method for site selection and route optimization of electric logistics vehicle charging stations in urban distribution in claim 1, it is characterized in that: The constraints in step S2 specifically include: (1) Customer needs are indivisible, and each customer can only be served once: is a decision variable, which means that when electric logistics vehicle k moves from node i to node j, is 1, otherwise it is 0; (2) The electric logistics vehicle only departs from its distribution center once: (3) Only the selected charging stations can provide charging services for electric logistics vehicles, M is equal to 1×10 9 : z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0; (4) Indicates that the total weight of the goods transported by the electric logistics vehicle cannot exceed its load capacity: r i is the demand of customer point i; (5) Indicates that the electric logistics vehicle is fully loaded when it leaves the distribution center: is the transportation volume of electric logistics vehicle k moving from node i to node j; (6) Every customer’s needs will be met: (7) Indicates that the electric logistics vehicle will not go from the distribution center to the charging station or another distribution center: (8) Indicates that the process of providing services to customers is continuous: is the time when electric logistics vehicle k arrives at customer point i; is the time when the electric logistics vehicle k leaves the customer point i; t i The service time required to serve customer point i; W q Average waiting time in queues for charging electric logistics vehicles; is the charging time of electric logistics vehicle k at charging station η; S is the set of candidate charging stations S∈{1,2,…,α}; d ij is the distance between node i and node j; v0 is the driving speed of the electric logistics vehicle; e ij is the traffic congestion coefficient between node i and node j, e ij =e ji ; (9) represents the charging amount required for electric logistics vehicles: The charging amount of electric logistics vehicle k at charging station η; b6 is the charging coefficient of the electric logistics vehicle; z η is the decision variable, which is 1 when the charging station η is selected, otherwise it is 0; (10) Indicates that the electric logistics vehicle is fully charged when it leaves the distribution center or charging station: P i k is the power of the electric logistics vehicle k when it leaves node i; b7 is the battery capacity of the electric logistics vehicle; (11) indicates that the power of the electric logistics vehicle remains unchanged before and after serving the customer: is the power of electric logistics vehicle k when it arrives at node i; (12) indicates that the power of the electric logistics vehicle at any customer point must be greater than the minimum reserved power: b8 is the minimum reserved power of the electric logistics vehicle; (13) represents the remaining power of the electric logistics vehicle when it arrives at the node: b9 is the power consumption coefficient of the electric logistics vehicle; (14) indicates the value range of the variable:

6. A method for site selection and route optimization of electric logistics vehicle charging stations in urban distribution according to claim 1, characterized in that: The step S3 specifically includes: S3-1 Initialization: Set the size of the bee population. The number of employed bees, observer bees, and scout bees is the same, all SN. Let iter be the maximum number of iterations of the algorithm. Give each bee a perception distance Visual, detal is the crowding factor, Try_number is the maximum number of foraging times, P c Denoted as the crossover probability, P m Denoted as the mutation probability, P m1 To exchange the probability of mutation, SN initial solutions are randomly generated; S3-2 Hired bee stage: Hired bees conduct field search and generate new hired bees. If the new hired bees have better solutions than the old hired bees, the new hired bees replace the original hired bees. Then the artificial fish swarm algorithm performs swarming behavior. If the swarming is successful, it determines whether the solution at this time is better. If it is better, it replaces the current solution. Otherwise, the current solution remains unchanged. If the swarming fails, the foraging behavior is performed. S3-3 Observer bee stage: The employed bees are selected to follow according to the roulette wheel, and the field search is carried out in the field. If a better solution is found, the new observer bee replaces the original observer bee. Otherwise, the crossover and mutation operations in the genetic algorithm are performed to determine whether a better solution is generated. If so, it is replaced, otherwise it remains unchanged. S3-4 Scouting bee stage: If a hired bee has not updated to a better solution when the number of iterations reaches the threshold, the hired bee is transformed into a scout bee and performs a field search in its field to randomly generate a new solution; S3-5 determines whether the number of iterations has reached the maximum. If not, continue to loop through steps S3-2 to S3-4 until the loop ends and the final result is output.

Images