A method for solving periodic door-to-door delivery vehicle scheduling based on dynamic programming

Abstract

The present invention discloses a method for solving a periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming, which specifically includes the following steps: S1, a customer submits a periodic service demand; S2, a customer set that needs to be served within a period is constructed; S3, distance data between nodes within the period is obtained; S4, a road section travel time set between nodes is constructed; S5, a periodic vehicle path optimization model is established according to customer and vehicle information; S6, a dynamic programming algorithm is used to solve the vehicle service path within the period, and a periodic vehicle scheduling scheme is output. The present invention can not only meet the periodic door-to-door delivery demand of customers, but also help logistics companies optimize the periodic vehicle distribution path, save logistics costs, and improve service quality and service efficiency.

Description

Technical Field

[0001] The present invention specifically relates to a method for solving a periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming, and belongs to the technical field of logistics planning in transportation planning and management. Background Art

[0002] With the strong support of national policies and the continuous improvement of social and economic levels, the logistics and transportation industry has ushered in a new round of development opportunities. Logistics companies have not only expanded their business to multiple fields such as production, supply, storage, and turnover, but also covered the service scope to the national and even transnational fields, which has greatly changed the production and living styles of residents. However, the logistics industry is also facing problems such as low transportation efficiency and high transportation costs. The continuous diversification of customer demand has also put forward higher and higher requirements for logistics companies. Faced with high logistics and transportation costs, how to optimize vehicle scheduling plans and improve vehicle distribution efficiency has become a hot topic for logistics and transportation companies. In particular, retail stores and large supermarkets often put forward periodic door-to-door delivery needs, that is, companies need to formulate periodic vehicle scheduling plans according to customer needs, and arrange vehicles to pick up and deliver goods within the cycle and date selected by customers. At present, most transportation companies mainly focus on single transportation service needs. Customers cannot easily put forward periodic service needs, and companies also lack methods to formulate efficient periodic vehicle scheduling plans. This will not only cause higher logistics and transportation costs, but also reduce the service quality and efficiency of enterprises, making the vehicle scheduling process too complicated and difficult to meet customers' periodic delivery needs.

[0003] Planning a periodic door-to-door pickup and delivery vehicle dispatch plan has the following functions: (1) Providing customers with a periodic door-to-door pickup and delivery option in the logistics company's application interface, meeting customers' periodic service needs and providing customers with customized delivery services; (2) Helping logistics companies develop periodic vehicle dispatch plans, saving logistics costs, streamlining vehicle dispatch processes, and improving the company's service efficiency and service quality.

[0004] The existing technology only considers the customer's single service needs. The customer can only choose the pick-up and delivery service once in the application interface. When periodic pick-up and delivery are required, the customer needs to submit the service request repeatedly. This method of repeatedly submitting logistics orders ignores the existence of periodic service needs, which not only takes up the customer's time, but also affects the customer's service experience. In addition, the traditional vehicle scheduling algorithm formulates a vehicle scheduling plan for customers on a single day, which is complex to solve and has low vehicle scheduling efficiency, resulting in high logistics transportation costs. Solving the periodic vehicle scheduling plan based on dynamic programming can not only make up for the shortcomings of existing logistics services, meet the customer's periodic pick-up and delivery needs, and avoid customers from submitting orders repeatedly; it can also help logistics companies optimize vehicle scheduling plans and save service costs. Therefore, it is necessary to consider the customer's personalized periodic door-to-door pick-up and delivery needs, and design a method for solving the periodic door-to-door pick-up and delivery vehicle scheduling plan based on dynamic programming to improve the company's service efficiency and service quality and improve the customer's user experience. Summary of the invention

[0005] The present invention provides a method for solving a periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming, which takes into account customer customized periodic service needs, can effectively improve the efficiency of solving vehicle distribution paths, simplify the vehicle scheduling process, and reduce the operating costs of logistics companies.

[0006] In order to solve the above technical problems, the present invention adopts a method for solving a periodic door-to-door delivery vehicle scheduling plan based on dynamic programming, and the method comprises the following steps:

[0007] S1. Customers submit periodic service requirements;

[0008] S2, the set of customers that need to be served during the construction period;

[0009] S3, obtaining the distance data between each node in the cycle;

[0010] S4, constructing a set of travel time between each node;

[0011] S5. Establish a periodic vehicle routing optimization model based on customer and vehicle information;

[0012] S6. Use a dynamic programming algorithm to solve the vehicle service path within the period and output a periodic vehicle scheduling plan.

[0013] Furthermore, in S1, the specific steps for the customer to submit a periodic service demand are:

[0014] Customers submit service requirements on the logistics company's application platform, including selecting the desired service period, determining the date and location for door-to-door delivery within the period, and the quantity required for a single delivery.

[0015] Furthermore, in S2, the specific steps of constructing the customer set are:

[0016] S21. Classify customers according to the length of service period, and classify customers with the same service period into one category;

[0017] S22. For each service cycle, the customers who need to be served every day are determined according to the date on which the customers allow door-to-door delivery, and a set of customers who need to be served every day in the cycle is constructed.

[0018] Furthermore, in S3, the specific steps for obtaining the distance data between nodes are:

[0019] S31, obtaining service location information through customer orders, and obtaining distance data between the vehicle start and end points and the customer nodes through local data of the distribution center or network map data, including one or more of the departure node, arrival node, departure time, arrival time, travel time and travel distance;

[0020] S32, node missing value processing, for nodes that lack distance or time data, select adjacent reachable nodes within a 500-meter service range centered on the node, and use the data of the adjacent nodes to supplement the missing values.

[0021] Furthermore, in S4, the specific steps of constructing the road segment travel time set are:

[0022] S41, judging whether the starting and ending points of the vehicle and the customer nodes are reachable based on the vehicle information, customer information, traffic network and traffic control information; if the nodes are not reachable, setting the travel time of the road section between the nodes to infinity;

[0023] S42: If the nodes are reachable, obtain the road section travel time corresponding to different departure times according to the inter-node driving distance and driving time data obtained in S3, and construct a road section travel time set.

[0024] Furthermore, in S5, the specific steps of establishing the vehicle path optimization model within the cycle are:

[0025] S51, by proportion The travel time t of segment (i, j) i，j Converted to vehicle transportation cost c i，j ,Right now

[0026] S52. Select binary variables is the decision variable, =1 means vehicle v passes through road section (i, j) on day d, =0 means other cases; set the objective function Even if the total transportation cost of all vehicles in the cycle is minimized;

[0027] S53, Ca represents the vehicle capacity of each vehicle, q p represents the service demand of customer p, s represents the starting point of each vehicle, and e represents the end point of each vehicle. The following constraints are set:

[0028]

[0029]

[0030]

[0031]

[0032]

[0033]

[0034] Constraints (1) and (2) indicate that each vehicle starts from the starting point and returns to the end point every day; constraint (3) indicates the flow balance constraint of the intermediate node; constraint (4) indicates the vehicle capacity constraint of each vehicle every day; constraint (5) indicates that each customer can only be served by one vehicle once every day; constraint (6) indicates the 0-1 constraint of the decision variable.

[0035] Furthermore, in S6, the specific steps of dynamically programming to solve the periodic vehicle scheduling solution are:

[0036] S61, initialize the labels L(i) of all nodes and the forward nodes, the label L(s) of the starting point is 0, and the labels L(i) of the remaining nodes are +∞;

[0037] S62, solve the shortest path problem for each car on each day in turn; starting from the starting point s, traverse each node i, when the subsequent node j satisfies L(i)+c i，j <L(j) and Ca-the customer demand of node j>0, update the label of node j to L(j)=L(i)+c i，j , the forward node of node j is updated to node i; continue to update the next node label until the label update of all nodes is completed;

[0038] S63, according to the node numbers, trace back from the end point e and output the optimal service path for each vehicle on each day;

[0039] S64. Prepare a periodic vehicle dispatch plan based on the optimal service route for each vehicle every day.

[0040] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0041] 1. Allow customers to submit periodic service requirements once on the application interface and flexibly select the date and location of periodic door-to-door delivery. This can avoid customers from submitting delivery orders repeatedly, provide customers with customized delivery services, simplify customer usage processes, and improve customers' logistics service experience;

[0042] 2. Combined with the periodic distribution needs of customers, a periodic vehicle routing optimization model is established to simulate the actual periodic logistics distribution problems and simplify the scheduling work of logistics companies;

[0043] 3. Designing a dynamic programming algorithm to solve the periodic vehicle scheduling problem can effectively improve the efficiency of solving the vehicle distribution path. The formulated periodic vehicle scheduling plan can be used repeatedly to avoid repetitive vehicle scheduling processes. In addition, formulating an efficient vehicle scheduling plan can also effectively utilize limited resources such as warehouses and vehicles, reduce the operating costs of logistics companies, make up for the shortcomings of the existing vehicle scheduling model, and provide a certain reference for companies to improve service quality and service efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flow chart of the method of the present invention;

[0045] Figure 2 It is a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0046] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0047] S1. Customers submit periodic service requirements;

[0048] Customers submit periodic door-to-door pickup and delivery orders on the logistics company's application platform (APP, website, etc.). The order information includes the customer's desired service cycle (2 days, 4 days, 6 days, etc.), the date and location of vehicle pickup and delivery within the period, and the demand for a single pickup and delivery. Figure 2 It shows a delivery scenario with a cycle of 2 days in a six-node transportation network, where node A is the parking lot.

[0049] S2, the set of customers that need to be served every day during the construction cycle;

[0050] Logistics companies need to develop periodic vehicle dispatch plans based on the length of the cycle. Customers who choose different service cycles need to be served in different categories. In addition, within a single service cycle, different customers have different service frequency requirements and do not need to be served every day. Therefore, it is necessary to develop a set of customers who need to be served every day within the cycle based on the order information submitted by the customer. The following steps are included:

[0051] S21. Based on customer orders, customers are classified according to the length of service period, and customers with the same service period are classified into one category;

[0052] S22. For each service cycle, the customers who need to be served every day are determined according to the date on which the customers allow door-to-door delivery, and a set of customers who need to be served every day in the cycle is constructed.

[0053] like Figure 2 In the six-node transportation network shown, according to the customer order information, customer B, customer E and customer F need to be served on the first day, and customer B, customer C, customer D and customer F need to be served on the second day. That is, the set of customers that need to be served each day is: {(Day1: B, E, F); (Day2: B, C, D, F)}.

[0054] S3, obtain the distance data between nodes;

[0055] Before planning the vehicle delivery route, it is necessary to obtain the distance data between nodes, including the topological relationship between nodes, vehicle travel distance, travel time, etc. The specific steps include:

[0056] S31. Obtain the door-to-door delivery location desired by the customer based on the order submitted by the customer; obtain the distance data between the vehicle's starting and ending points and the customer's node through the local data of the distribution center or the network map data, including the departure node, arrival node, departure time, arrival time, driving time, driving distance, etc. Figure 2 The number on each road segment corresponds to the travel time of that road segment.

[0057] S32, node missing value processing. For nodes that lack distance or time data, select adjacent reachable nodes within a 500-meter service range centered on the node, and use the data of the adjacent nodes to supplement the missing values.

[0058] S4, constructing the travel time set of the road sections between nodes;

[0059] According to the travel distance and travel time data between nodes, the travel time of the road segments between nodes is determined, including the following steps:

[0060] S41. According to the customer order information and traffic control information, it is determined whether the starting and ending points of the vehicle and the customer nodes are reachable, that is, whether the customer can be served; if the nodes are not reachable, the travel time of the road section between the nodes is set to infinity, indicating that the customer cannot be served;

[0061] S42: If the nodes are reachable, construct a set of road section travel times corresponding to different departure times based on the distance and time data between each road section.

[0062] S5. Establish a periodic vehicle routing optimization model;

[0063] The optimization model of the periodic vehicle routing problem is established based on the customer's order information, the service capabilities of the logistics company, the transportation network and control information, which mainly includes the following steps:

[0064] S51. According to a certain proportion The travel time t of segment (i, j) i，j Converted to vehicle transportation cost c i，j ,Right now

[0065] S52. Select binary variables is the decision variable, =1 means vehicle v passes through road section (i, j) on day d, =0 means other cases. Set the objective function Even if the total transportation cost of all vehicles in the cycle is minimized;

[0066] S53, Ca represents the vehicle capacity of each vehicle, q p represents the service demand of customer p, s represents the starting point of each vehicle, and e represents the end point of each vehicle. The following constraints are set:

[0067]

[0068]

[0069]

[0070]

[0071]

[0072]

[0073] Constraints (1) and (2) indicate that each vehicle starts from the starting point and returns to the end point every day; constraint (3) indicates the flow balance constraint of the intermediate nodes to ensure that a complete vehicle service path is found; constraint (4) indicates the vehicle capacity constraint of each vehicle every day, that is, the total service volume of each vehicle every day cannot exceed the total vehicle capacity; constraint (5) indicates that each customer must and can only be served once by one vehicle every day; constraint (6) indicates the 0-1 constraint of the decision variable.

[0074] S6, solving and outputting a periodic vehicle scheduling plan;

[0075] According to the established optimization model of periodic vehicle routing problem, a dynamic programming algorithm is designed to solve it, which mainly includes the following steps:

[0076] S61, algorithm initialization; initialize the labels L(i) of all nodes and forward nodes, where the label L(s) of the starting point is 0, and the labels L(i) of the remaining nodes are +∞.

[0077] S62, node label update; solve the shortest path problem for each car on each day in turn; starting from the starting point s, traverse each node i, when the subsequent node j satisfies L(i)+ci ，j <L(j) and Ca-the customer demand of node j>0, update the label of node j to L(j)=L(i)+c i，j , the forward node of node j is updated to node i. Continue to update the next node label until all node labels are updated.

[0078] S63, finding the optimal path; according to the node numbers, backtracking from the end point e, to obtain the optimal service path for each vehicle every day.

[0079] S64. Output the solution result; compile a periodic vehicle dispatching plan based on the optimal service path of each vehicle every day. The logistics company arranges vehicles to provide door-to-door pickup and delivery services based on the periodic vehicle dispatching plan.

[0080] Figure 2 The paper shows the periodic vehicle dispatching scheme obtained by the dynamic programming algorithm in a 2-day delivery scenario, that is, the optimal service path of the vehicle on the first day is A→B→F→E→A, and the optimal service path of the vehicle on the second day is A→B→F→D→C→A. Logistics companies can provide periodic services to customers according to the optimal vehicle delivery path.

[0081] The above embodiments are only preferred implementation modes of the present invention. It should be pointed out that ordinary technicians in this technical field can make several improvements and equivalent substitutions without departing from the principles of the present invention. These technical solutions after improvements and equivalent substitutions to the claims of the present invention all fall within the protection scope of the present invention.

Claims

1. A method for solving a periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming, characterized in that: The method comprises the following steps: S1. Customers submit periodic service requirements; S2, the set of customers that need to be served during the construction period; S3, obtaining the distance data between each node in the cycle; S4, constructing a set of travel time between the sections of each node; S5. Establish a periodic vehicle routing optimization model based on customer and vehicle information. The specific steps are as follows: S51, by proportion The travel time t of road segment (i, j) i,j Converted to vehicle transportation cost c i,j ,Right now S52. Select binary variables is the decision variable, =1 means vehicle v passes through road section (i, j) on day d, =0 means other cases; set the objective function Even if the total transportation cost of all vehicles in the cycle is minimized; S53, Ca represents the vehicle capacity of each vehicle, q p represents the service demand of customer p, s represents the starting point of each vehicle, and e represents the end point of each vehicle. The following constraints are set: Constraints (1) and (2) indicate that each vehicle starts from the starting point and returns to the end point every day; constraint (3) indicates the flow balance constraint of the intermediate node; constraint (4) indicates the vehicle capacity constraint of each vehicle every day; constraint (5) indicates that each customer can only be served by one vehicle once every day; constraint (6) indicates the 0-1 constraint of the decision variable; S6. Use a dynamic programming algorithm to solve the vehicle service path within the period and output a periodic vehicle scheduling plan.

2. The method for solving the periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming according to claim 1 is characterized in that: In S1, the specific steps for customers to submit periodic service requirements are: Customers submit service requirements on the logistics company's application platform, including selecting the desired service period, determining the date and location for door-to-door delivery within the period, and the quantity required for a single delivery.

3. The method for solving the periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming according to claim 1, characterized in that: In S2, the specific steps for constructing the customer set are: S21. Classify customers according to the length of service period, and classify customers with the same service period into one category; S22. For each service cycle, the customers who need to be served every day are determined according to the date on which the customers allow door-to-door delivery, and a set of customers who need to be served every day in the cycle is constructed.

4. The method for solving the periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming according to claim 1, characterized in that: In S3, the specific steps to obtain the distance data between nodes are: S31, obtaining service location information through customer orders, and obtaining distance data between the vehicle start and end points and the customer nodes through local data of the distribution center or network map data, including one or more of the departure node, arrival node, departure time, arrival time, travel time and travel distance; S32, node missing value processing, for nodes that lack distance or time data, select adjacent reachable nodes within a 500-meter service range centered on the node, and use the data of the adjacent nodes to supplement the missing values.

5. The method for solving the periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming according to claim 1, characterized in that: In S4, the specific steps of constructing the road segment travel time set are: S41, judging whether the starting and ending points of the vehicle and the customer nodes are reachable based on the vehicle information, customer information, traffic network and traffic control information; if the nodes are not reachable, setting the travel time of the road section between the nodes to infinity; S42: If the nodes are reachable, obtain the road section travel time corresponding to different departure times according to the inter-node driving distance and driving time data obtained in S3, and construct a road section travel time set.

6. The method for solving the periodic door-to-door delivery vehicle scheduling scheme based on dynamic programming according to claim 1, characterized in that: In S6, the specific steps of dynamically programming to solve the periodic vehicle scheduling solution are: S61, initialize the labels L(i) of all nodes and the forward nodes, the label L(s) of the starting point is 0, and the labels L(i) of the remaining nodes are +∞; S62, solve the shortest path problem for each car on each day in turn; starting from the starting point s, traverse each node i, when the subsequent node j satisfies L(i)+c i，j <L(j) and Ca-the customer demand of node j>0, update the label of node j to L(j)=L(i)+c i，j , the forward node of node j is updated to node i; continue to update the next node label until the label update of all nodes is completed; S63, according to the node numbers, trace back from the end point e and output the optimal service path for each vehicle on each day; S64. Prepare a periodic vehicle dispatch plan based on the optimal service route for each vehicle every day.

Images