WO2020030028A1 - 配送调度 - Google Patents

配送调度 Download PDF

Info

Publication number
WO2020030028A1
WO2020030028A1 PCT/CN2019/099714 CN2019099714W WO2020030028A1 WO 2020030028 A1 WO2020030028 A1 WO 2020030028A1 CN 2019099714 W CN2019099714 W CN 2019099714W WO 2020030028 A1 WO2020030028 A1 WO 2020030028A1
Authority
WO
WIPO (PCT)
Prior art keywords
order
target
delivery
combination
index
Prior art date
Application number
PCT/CN2019/099714
Other languages
English (en)
French (fr)
Inventor
卞洁辉
张涛
郝井华
孔兵
Original Assignee
北京三快在线科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 北京三快在线科技有限公司 filed Critical 北京三快在线科技有限公司
Priority to US17/266,624 priority Critical patent/US20210312347A1/en
Publication of WO2020030028A1 publication Critical patent/WO2020030028A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G06Q10/0838—Historical data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G06Q10/047—Optimisation of routes or paths, e.g. travelling salesman problem
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N5/00—Computing arrangements using knowledge-based models
    • G06N5/01—Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G06Q10/06311—Scheduling, planning or task assignment for a person or group
    • G06Q10/063112—Skill-based matching of a person or a group to a task
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/083—Shipping
    • G06Q10/0835—Relationships between shipper or supplier and carriers
    • G06Q10/08355—Routing methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/20—Ensemble learning

Definitions

  • This application relates to distribution scheduling.
  • the scheduling system needs to optimize the matching of orders and delivery staff so that the orders pushed to the delivery staff are as close as possible to the situation of the delivery staff.
  • the scheduling system may generally perform order scheduling according to the matching index after the delivery person adds the target order.
  • the matching index may indicate the matching degree of the distribution path before and after the delivery order is added by the delivery person; when the matching index is greater than a threshold value, it indicates that the target order and the delivery person are relatively matched.
  • the above-mentioned delivery scheduling method ignores the influence of the deliveryman's own subjective factors on the delivery relationship. For example, if the dispatcher's willingness to accept the target order is not high, even if the matching indicators meet the requirements, the dispatcher can refuse to accept the order, which can not truly form a distribution relationship, which affects the scheduling accuracy and scheduling efficiency.
  • the present application provides a distribution scheduling method.
  • the distribution scheduling method includes: based on a combination manner of at least one target order and at least one target distribution person, planning a distribution path after the target distribution person is assigned a target order in each combination manner; and calculating the distribution in each combination manner The route's delivery efficiency index and order willingness index associated with the target order to which the target dispatcher is assigned; based on the delivery efficiency index and order willingness index of each combination method, the optimal one is selected from the at least one combination method Combined mode for distribution scheduling.
  • the present application provides a distribution scheduling device.
  • the distribution scheduling device includes a path planning unit, a calculation unit, and a scheduling unit.
  • the path planning unit is configured to plan a distribution path after the target orderer is assigned the target order based on at least one combination form formed by at least one target order and at least one target delivery person.
  • the calculation unit is configured to calculate the delivery efficiency index and the order willingness index of the delivery route in each combination manner, which are associated with the target delivery person being assigned the target order.
  • the scheduling unit is configured to select an optimal combination manner from the at least one combination manner for distribution scheduling based on the distribution efficiency index and the order willingness index of each combination manner.
  • the present application provides a computer-readable storage medium.
  • a computer program is stored in the storage medium, and the computer program is configured to execute the distribution scheduling method according to the first aspect.
  • the present application provides an electronic device.
  • the electronic device includes a processor and a memory for storing processor-executable instructions.
  • the processor is configured to execute the delivery scheduling method described in the first aspect above.
  • the embodiment of the present application provides a distribution scheduling scheme.
  • a comprehensive index for reference of the scheduling system is obtained.
  • the scheduling system determines whether to perform scheduling based on the comprehensive index. In this way, not only the objective factors such as the distribution efficiency index, but also the subjective factors such as the willingness of the delivery staff to take orders are considered.
  • the dispatcher is assigned to the order, since the delivery efficiency index and the willingness to accept order are in compliance with the requirements, the probability of the dispatcher accepting the order is effectively increased; thus the dispatching accuracy and dispatching efficiency can be effectively improved.
  • FIG. 1 is a schematic structural diagram of a distribution scheduling system according to an exemplary embodiment of the present application.
  • FIG. 2 is a flowchart illustrating a distribution scheduling method according to an exemplary embodiment of the present application
  • FIG. 3 is a hardware structural diagram of a distribution scheduling device according to an exemplary embodiment of the present application.
  • Fig. 4 is a schematic block diagram of a distribution scheduling device according to an exemplary embodiment of the present application.
  • first, second, third, etc. may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other.
  • first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information.
  • word “if” as used herein can be interpreted as “at” or "when” or "in response to determination”.
  • Fig. 1 is a schematic structural diagram of a distribution scheduling system according to an exemplary embodiment of the present application.
  • the scheduling system may include a data collection module 101, a path planning module 102, an order willingness calculation module 103, and an order allocation decision module 104.
  • the data collected by the data collection module 101 includes four types, which are order data, delivery person data, environmental data, and route data.
  • the order data may include at least one of the following: the delivery distance of the order, the delivery price, the delivery time period, the value of the item, the time for stocking (the time between the creation of the order and the time when the delivery staff can pick it up), Delivery time, order type (such as instant delivery types such as takeaway, courier), the area where the order is located, the starting position (such as the merchant's location), and the end position (such as the destination location of the order)
  • the delivery person data may include historical data of the delivery person and real-time data of the delivery person.
  • the historical data of the delivery person may include at least one of the following: historical average speed, historical average daily order acceptance volume, historical average daily rejection rate, historically delivered area, historically delivered applicants, and historically different delivery.
  • historical average speed historical average daily order acceptance volume
  • historical average daily rejection rate historically delivered area
  • historically delivered applicants historically different delivery
  • the proportion of orders received from distance orders the proportion of orders received at different delivery time periods in history, and the proportion of orders received at different delivery prices in history.
  • the real-time data of the delivery person may include at least one of the following: a delivery person level and a delivery person position.
  • the environmental data may include at least one of the following: weather of the current delivery area, the number of newly created orders within the preset duration of the current delivery area, and loader load data of the current delivery area within the preset duration, The number of free delivery staff in the current delivery area within the preset time period, and the cancellation rate of the current delivery area within the preset time period.
  • the route data may include at least one of the following: the distance between the delivery person and the starting position of each order and the time required to reach the starting position; the distance between the delivery person and the end position of each order; and The length of time required for the end position; the distance and length of the start position between orders; the distance and length of the end position between orders; the distance and length of the start position and the end position between orders.
  • the data collection module 101 can convert the collected raw data into a data format that can be used directly by the subsequent path planning module 102 and the order willingness calculation module 103.
  • data from different sources often have different data formats that cannot be used directly by the system.
  • some data is structured data (such as database data)
  • some data is unstructured data (such as office documents in various formats, XML, HTML, reports, pictures, audio and video, etc.)
  • the data collection module 101 can convert all the collected data into standardized data in a unified format, thereby facilitating the use of other modules directly.
  • the route planning module 102 is configured to plan a distribution route of a delivery person, and calculate a matching degree and an efficiency index based on the distribution route.
  • planning a distribution route requires, for example, deliveryman data, order data, environmental data, and route data collected by the data collection module 101, so that based on the location speed of the deliveryman, the start and end positions of the order, and the delivery area Data such as environment and distribution area route plan corresponding distribution route.
  • the optimal distribution path can be planned based on the path optimization algorithm, thereby calculating the optimal matching index and efficiency index.
  • the matching index indicates a degree of similarity between distribution paths before and after the target delivery person is assigned the target order
  • the efficiency index indicates how efficiently the target delivery person distributes the target order.
  • the goal of the route optimization algorithm is to minimize the delivery time required for the delivery route planned after the target delivery person is assigned the target order.
  • the delivery person j already has 5 orders to be delivered, of which 2 orders have been picked up and 3 orders have not been picked up.
  • the deliveryman j has a total of 8 destinations, that is, 3 starting positions (corresponding to the 3 uncollected orders) and 5 ending positions. Because the order of reaching the start and end positions of the order will form different delivery routes and directly affect the final delivery time, it is necessary to optimize the delivery route so that the total delivery time is the shortest.
  • the optimization algorithm needs to have at least one of the following constraints.
  • the target delivery person needs to go to the starting position of the order and then to the ending position of the order when delivering the order.
  • the complete distribution process of an order must be that the delivery person first picks up the order at the starting point of the order, and then can carry the acquired goods to the end point of the order.
  • the total number of orders after the target delivery person is assigned the target order cannot exceed the order receiving limit.
  • the order receiving upper limit may be set by the system, or may be set by the delivery person himself according to actual conditions.
  • both the current unfinished order and the target order are delivered before the latest delivery time.
  • a latest delivery time which indicates the latest delivery time that the delivery receiver may accept. If the actual delivery time exceeds this latest delivery time, Then it is a delivery timeout.
  • the delivery time is generally expected to be earlier than this latest delivery time. However, when multiple orders are delivered at the same time, due to the increase in the delivery path, the estimated delivery time of each order will also change accordingly. Scheduling When scheduling, the system must ensure that the estimated delivery time of each order in the planned distribution path does not exceed the latest delivery time.
  • the difference between the stocking time of this order and the time required by the target delivery person to reach the starting position of the order is less than the threshold.
  • the threshold In actual logistics distribution, the length of time for stocking by different distribution applicants is different, and the delivery staff's early arrival at the starting position does not mean that the goods can be picked up immediately. If the delivery applicant is still preparing the goods, the delivery staff must wait, which wastes valuable delivery time; therefore, it is necessary to ensure that the delivery staff can pick up the goods immediately or as soon as possible after reaching the starting position. For this reason, if the difference between the time required for the order to be prepared and the time required for the target delivery person to reach the starting position of the order is less than the threshold, it means that the delivery applicant can complete the preparation before or after the delivery person arrives. It is convenient for the delivery staff to complete the pickup work quickly.
  • the path optimization algorithm may include a simulated annealing algorithm, an ant colony algorithm, a particle algorithm, and the like.
  • the order willingness calculation module 103 is configured to calculate an order willingness indicator of the dispatcher for the assigned order.
  • the order willingness indicator indicates the acceptance degree of the order by the delivery person.
  • the order willingness calculation module 103 may calculate an order willingness indicator based on a machine learning model, according to the order data, delivery person data, and environmental data obtained by the data collection module 101, and according to the matching indicators obtained by the path planning module 102. .
  • the order willingness model is obtained by training as follows: basic data and matching indicators of historical orders are used as training data, and after the historical orders are assigned to the delivery person, the delivery person accepts or rejects the label, and the machine learning algorithm is used to perform Model training.
  • the trained model is determined as the order willingness model.
  • the machine learning algorithm may include at least one of xgboost, logistic regression, random forest, decision tree, Gradient Boost Decision Tree (GBDT), and support vector machine.
  • the order allocation decision module 104 may calculate a comprehensive index according to the efficiency index and the willingness to accept orders, and then the order allocation decision module 104 determines whether to perform scheduling in a corresponding combination manner according to the comprehensive index.
  • the order allocation decision module 104 is a decision maker.
  • Fig. 2 is a flow chart of a distribution scheduling method according to an exemplary embodiment of the present application. The method may be applied to the above-mentioned scheduling system, and the method may specifically include the following steps 210 to 230.
  • Step 210 Based on at least one combination mode formed by at least one target order and at least one target delivery person, plan a distribution path after the target delivery person is assigned the target order in each combination mode.
  • the scheduling system may obtain at least one combination manner, and the at least one combination manner includes at least one target order to be allocated and at least one target delivery person.
  • a delivery person can deliver multiple orders at the same time, and the delivery person has an order limit.
  • the target delivery person refers to the aforementioned idle delivery person, and may refer to a delivery person whose number of orders delivered at the same time does not reach the order limit.
  • the scheduling system can plan the distribution path after each target distributor is assigned a target order in each combination.
  • This step 210 may be performed by the aforementioned path planning module in the scheduling system.
  • the planning the distribution path after the target orderer is assigned the target order in each combination method specifically includes: planning the optimal distribution path after the target orderer is assigned the target order in each combination method .
  • the optimal distribution path may mean that the distribution time required for the distribution path planned after the target distribution person is assigned the target order is the shortest.
  • the planning the optimal distribution path after the target dispatcher is assigned the target order in each combination mode specifically includes: planning the target dispatcher after the target order is assigned in each combination mode based on the path optimization algorithm. Excellent distribution path.
  • the goal of the route optimization algorithm is to minimize the delivery time required for the delivery route planned after the target delivery person is assigned the target order.
  • Constraints of the path optimization algorithm include at least one of the following:
  • the target delivery person needs to go to the starting position of the target order and then to the ending position of the target order when delivering the target order;
  • the total number of orders after the target delivery person is assigned the target order cannot exceed the order limit
  • both the current outstanding order and the target order will be delivered before the latest delivery time
  • the difference between the stocking time of the target order and the time required by the target delivery person to reach the starting position of the target order is less than the first threshold.
  • Step 220 Calculate the delivery efficiency index and the order willingness index of the delivery route in each combination manner that are associated with the target delivery person being assigned the target order.
  • the distribution efficiency index may include a matching index and an efficiency index.
  • the step 220 may specifically include the following steps B1 and B2.
  • Step B1 Calculate the matching index and efficiency index of the distribution path in each combination mode; wherein the matching index indicates the degree of similarity between the distribution path before and after the target delivery person is assigned the target order, and the efficiency index indicates the target distribution The efficiency of the staff to deliver the target order.
  • Step B2 Calculate the willingness index of the target delivery person corresponding to each combination method according to the matching index of each combination method; wherein the willingness index of the order indicates the acceptance degree of the target order by the target delivery person.
  • This step B1 may be performed by the aforementioned path planning module in the scheduling system.
  • the matching index may be a value between 0 and 1. The closer to 1, the higher the degree of similarity; conversely, the closer to 0, the lower the degree of similarity.
  • the efficiency index may be a value between 0 and 1. The closer it is to 1, the more efficient the target delivery person is in delivering the target order; on the other hand, the closer it is to 0, the lower the efficiency is. Generally, if the starting position or end position of the target order is close to the starting position or end position of other orders of the target delivery person, then the efficiency will be relatively high.
  • This step B2 may be performed by the aforementioned order willingness calculation module in the scheduling system.
  • step B2 may specifically include: obtaining basic data of the target order in each combination mode; and inputting the basic data and matching indicators into an order willingness model, obtaining the order willingness model and calculating The target willingness index of the target delivery person in each combination.
  • the obtaining of the basic data of the target order in each combination mode specifically includes: obtaining different types of order proportions from the historical order data of the target delivery person in each combination mode; determining the target order The order proportion of the type in the historical order data; and the determined order proportion as the basic data of the target order.
  • the different types include at least one of the following: different delivery distances, different delivery time periods, different delivery prices, and different delivery areas.
  • the order proportion of different delivery distances is obtained from the historical data of the target delivery person, and the order proportion of the target order is determined in combination with the distribution distance of the target order.
  • the order acceptance ratio can reflect the target distributor's preference for orders at this delivery distance.
  • the proportion of orders received in different delivery time periods is obtained from the historical data of the target delivery person, and the proportion of orders received in the target order is determined in combination with the distribution time period of the target order.
  • the order receiving ratio can reflect the degree of preference of the target delivery person for the order in this delivery period.
  • the order proportion of different delivery prices is obtained from the historical data of the target delivery person, and the order proportion of the target order is determined by combining the target order's delivery price.
  • the order receiving ratio can reflect the target distributor's preference for orders with this delivery price.
  • the order proportion of different delivery regions is obtained from the historical data of the target delivery person, and the order proportion of the target order is determined in combination with the distribution region of the target order.
  • the order receiving ratio can reflect the target distributor's preference for orders in this distribution area.
  • the geohash algorithm can be used to encode the historical delivery areas of the delivery staff, divide these areas into equal-shaped blocks according to latitude and longitude, and count the historical delivery times of the target delivery staff in different blocks.
  • the target block where the distribution requester's geographic location and / or the distribution receiver's geographic location are located can be determined; the historical distribution times of the target area can be obtained from the historical distribution times on the different blocks.
  • the order willingness model is obtained by training as follows: basic data and matching indicators of historical orders are used as training data, and after the historical order is assigned to a delivery person, the delivery person accepts or rejects the label. Using machine learning algorithms for model training, the trained model is determined as the order willingness model.
  • the machine learning algorithm includes at least one of xgboost, logistic regression, random forest, decision tree, GBDT, and support vector machine.
  • Step 230 Integrate the delivery efficiency index and the order willingness index of each combination mode, and select the optimal combination mode from the at least one combination mode for delivery scheduling.
  • the scheduling system may integrate the distribution efficiency index and the order willingness index of each combination mode, and select the optimal combination mode from all the combination modes for distribution scheduling. This step may be performed by the aforementioned order allocation decision module in the scheduling system.
  • the step 230 may specifically include a step A1 and a step A2.
  • Step A1 According to the delivery efficiency index and the order willingness index of each combination, calculate the comprehensive index of each combination.
  • Step A2 According to the comprehensive index of each combination method, select the best combination method from all the combination methods for distribution scheduling.
  • Steps A1 and A2 may be performed by the aforementioned order allocation decision module in the scheduling system.
  • the step A1 specifically includes: multiplying the distribution efficiency index of each combination method by an efficiency weight to obtain an efficiency value; and multiplying the order willingness index of each combination method by the willingness weight to obtain a willingness value; The efficiency value and the willingness value of each combination manner are summed to obtain a comprehensive index corresponding to each combination manner. The sum of the efficiency weight and the willingness weight is 1.
  • the distribution efficiency index may include a matching index and an efficiency index; therefore, in this embodiment, the efficiency value may specifically be an efficiency index in the distribution efficiency index multiplied by an efficiency weight, thereby obtaining an efficiency value.
  • the step A2 specifically includes: when the comprehensive indicator of the one combination method is greater than a second threshold, It is determined to perform distribution scheduling according to the combined manner.
  • the target order is one
  • the target delivery person is N
  • the combination method is N kinds
  • N is a natural number greater than 1
  • the step A2 specifically includes: from the N kinds of comprehensive indicators, The largest comprehensive index is selected, and distribution scheduling is performed according to the combination mode corresponding to the largest comprehensive index.
  • the target order is M
  • the target delivery person is N
  • the combination method is M * N
  • M and N are both natural numbers greater than 1.
  • the step A2 specifically includes: based on a decision algorithm, from Select one for each row of the M * N columns of comprehensive indicators to make the sum of the M comprehensive indicators the largest; of which, the target order of the combination method corresponding to the selected M comprehensive indicators cannot be repeated; according to the selected M indicators Comprehensive scheduling corresponding to comprehensive indicators for distribution scheduling.
  • M target orders and N target delivery staff there can be M * N different combinations.
  • M * N efficiency indicators and order willingness indicators Set the efficiency index of the i-th delivery person to the j-th order as e ij and the order willingness index as w ij .
  • the M orders and N delivery staff can build a matrix of M rows and N columns, and the value of the i-th row and the j-th column in the matrix is the value of a comprehensive index, denoted as p ij .
  • p ij ⁇ * w ij + (1- ⁇ ) * e ij , where ⁇ can represent an efficiency weight, and the efficiency weight can be an artificially preset experience value; correspondingly, 1- ⁇ can represent a willingness weight.
  • the goal of the order allocation decision module is to assign an optimal delivery person to each order, so that the sum of p of all orders (M orders) is maximized.
  • the constraint here is that each order can only be assigned to one delivery person, and each delivery person has an order receiving limit. Solving the above formula is similar to the bipartite graph maximum weight perfect matching method, and decision algorithms such as KM algorithm and Hungarian algorithm can be used.
  • the embodiment of the present application provides a distribution scheduling method.
  • a comprehensive index for reference of the scheduling system is obtained;
  • the scheduling system determines whether to perform scheduling based on comprehensive indicators; in this way, not only the objective factors such as the distribution efficiency index but also the subjective factors such as the willingness of the delivery staff to take orders are considered.
  • the single willingness indicators are all in line with the requirements, so the probability of the delivery person receiving the order is effectively increased; thus the scheduling accuracy and scheduling efficiency can be effectively improved.
  • part-time delivery staff can choose to accept or reject the assigned logistics orders.
  • the delivery scheduling scheme described in this application can be applied not only to the traditional full-time delivery person mode, but also to the O2O crowdsourcing mode. Through comprehensive distribution path's distribution efficiency index and part-time delivery staff's willingness to take orders, the distribution scheduling greatly increases the probability of part-time delivery staff accepting assigned orders, thereby effectively improving scheduling accuracy and scheduling efficiency.
  • This application provides an embodiment of a delivery scheduling device.
  • This device embodiment can be applied on a server.
  • the device embodiments may be implemented by software, or by hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by reading the corresponding computer program instructions in the non-volatile memory into the memory through its processor.
  • FIG. 3 a hardware structure diagram of the distribution scheduling device of the present application is implemented in addition to the processor, memory, network interface, and non-volatile memory shown in FIG. The examples usually include other hardware according to the actual function of the distribution schedule, which will not be repeated here.
  • the distribution scheduling device may include a path planning unit 310, a calculation unit 320, and a scheduling unit 330.
  • the path planning unit 310 is configured to plan a distribution path after a target orderer is assigned a target order in each combination manner based on at least one combination form formed by at least one target order and at least one target delivery person.
  • the calculation unit 320 is configured to calculate a delivery efficiency index and an order willingness index of the delivery route in each combination manner, which are associated with the target delivery person being assigned the target order.
  • the scheduling unit 330 is configured to select an optimal combination manner from the at least one combination manner for distribution scheduling based on the distribution efficiency index and the order willingness index of each combination manner.
  • the calculation unit 320 specifically includes a first calculation sub-unit and a second calculation sub-unit.
  • the first calculation subunit is configured to calculate a matching index and an efficiency index of the distribution path in each combination mode.
  • the matching index indicates a degree of similarity between distribution paths before and after the target delivery person is assigned the target order
  • the efficiency index indicates how efficiently the target delivery person distributes the target order.
  • the second calculation sub-unit is configured to calculate, according to the matching index of each combination mode, the willingness index of the target delivery person in each combination mode.
  • the order willingness indicator indicates the acceptance degree of the target delivery person by the target order.
  • the path planning unit 310 specifically includes: an acquisition subunit and a path planning subunit.
  • the acquisition subunit is configured to acquire at least one combination manner including at least one target order to be allocated and at least one target delivery person.
  • the path planning sub-unit is configured to plan the optimal distribution path after the target orderer is assigned the target order in each combination mode.
  • the path planning subunit specifically includes: based on a path optimization algorithm, planning an optimal distribution path after the target distributor is assigned a target order in each combination mode.
  • the goal of the route optimization algorithm is to minimize the delivery time required for the delivery route planned after the target delivery person is assigned the target order.
  • the constraint conditions of the path optimization algorithm include at least one of the following:
  • the target delivery person needs to go to the starting position of the target order and then to the ending position of the target order when delivering the target order;
  • the total number of orders after the target delivery person is assigned the target order cannot exceed the order limit
  • both the current outstanding order and the target order will be delivered before the latest delivery time
  • the difference between the stocking time of the target order and the time required by the target delivery person to reach the starting position of the target order is less than the first threshold.
  • the optimization algorithm includes at least one of a simulated annealing algorithm, an ant colony algorithm, and a particle algorithm.
  • the second calculation subunit specifically includes: an acquisition subunit and a calculation subunit.
  • the acquisition subunit is configured to acquire basic data of the target order in each combination mode
  • the calculation subunit is configured to input the basic data and the matching index into the order willingness model, and obtain the order willingness index corresponding to the target delivery person calculated by the order willingness model.
  • the acquiring subunit specifically includes: a proportion acquiring subunit, a proportion determining subunit, and a data determining subunit.
  • the proportion acquiring subunit is configured to acquire different types of order proportions from the historical order receiving data of the target delivery person in each combination mode.
  • the proportion determining subunit is configured to determine an order proportion of the type of the target order in the historical order data.
  • the data determination subunit is configured to use the determined order acceptance ratio as the basic data of the target order.
  • the different types include at least one of different delivery distances, different delivery time periods, different delivery prices, and different delivery areas.
  • the order willingness model is obtained by training as follows: basic data and matching indicators of historical orders are used as training data, and after the historical order is assigned to a delivery person, the delivery person accepts or rejects the label, using Machine learning algorithms perform model training, and the trained model is determined as the willingness-to-order model.
  • the machine learning algorithm includes at least one of xgboost, logistic regression, random forest, decision tree, Gradient Boost Decision Tree (GBDT), and support vector machine.
  • the scheduling unit 330 specifically includes a first scheduling subunit and a second scheduling subunit.
  • the first scheduling sub-unit is configured to calculate a comprehensive index of each combination method according to a delivery efficiency index and an order willingness index of each combination method.
  • the second scheduling sub-unit is configured to select an optimal combination mode from the at least one combination mode for distribution scheduling according to a comprehensive index of each combination mode.
  • the first scheduling subunit specifically includes a first calculation subunit, a second calculation subunit, and a third calculation subunit.
  • the first calculation subunit is configured to multiply the distribution efficiency index of each combination method by an efficiency weight to obtain an efficiency value.
  • the second calculation sub-unit is configured to multiply the willingness index for each combination by the willingness weight to obtain the willingness value.
  • the third calculation subunit is configured to sum the efficiency value and the willingness value of each combination manner to obtain a comprehensive index corresponding to each combination manner.
  • the sum of the efficiency weight and the willingness weight is 1.
  • the second dispatching subunit specifically includes: when the comprehensive indicator is greater than a second threshold, according to the one Scheduling in combination.
  • the second scheduling subunit specifically includes: from the N kinds of comprehensive indicators , Select the largest comprehensive index, and perform distribution scheduling in accordance with the combination method corresponding to the largest comprehensive index.
  • the combination method is a natural number of M * N types, and both M and N are greater than 1, the second scheduling subunit includes selecting Subunits and scheduling subunits.
  • the selection sub-unit is configured to select one from each of the comprehensive indicators of M rows * N columns based on the decision algorithm, so that the sum of the M comprehensive indicators is the largest.
  • the target order of the combination manner corresponding to the selected M comprehensive indicators cannot be repeated.
  • the scheduling sub-unit is configured to perform distribution scheduling according to a combination manner corresponding to the selected M comprehensive indicators.
  • the decision algorithm includes at least one of a KM algorithm and a Hungarian algorithm.
  • the relevant part may refer to the description of the method embodiment.
  • the device embodiments described above are only schematic, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, may be located One place, or it can be distributed across multiple network elements. Some or all of these modules can be selected according to actual needs to achieve the purpose of the solution of this application. Those of ordinary skill in the art can understand and implement without creative efforts.
  • An embodiment of the present application provides a computer-readable storage medium, where the storage medium stores a computer program, and the computer program is configured to execute any one of the foregoing embodiments of a delivery scheduling method.
  • an embodiment of the present application further provides an electronic device.
  • the electronic device includes a processor and a memory.
  • the memory includes, for example, a memory and a non-volatile memory.
  • the memory is used to store processor-executable instructions.
  • the processor is configured to execute any one of the above embodiments of the dispatch scheduling method.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Marketing (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Game Theory and Decision Science (AREA)
  • Computational Linguistics (AREA)
  • Educational Administration (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Medical Informatics (AREA)
  • General Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本申请提供一种配送调度方法。根据一个示例，该配送调度方法包括：基于至少一个目标订单与至少一个目标配送员形成的至少一种组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径；计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标；基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。

Description

配送调度
相关申请的交叉引用
本申请要求于2018年8月8日提交的、申请号为2018108991706、发明名称为“一种配送调度方法及装置”的中国专利申请的优先权，该申请的全文以引用的方式并入本文中。
技术领域
本申请涉及配送调度。
背景技术
在相关技术中，为了提升物流配送效率，调度系统需要对订单和配送员进行优化匹配，使得推送给配送员的订单尽量符合配送员顺路的情况。具体地，调度系统一般可以根据配送员新增目标订单后的匹配指标进行订单调度。所述匹配指标可以表示配送员新增目标订单前后的配送路径的匹配程度；当匹配指标大于阈值时，说明目标订单和目标配送员较为匹配。
然而上述配送调度方式忽视了配送员自身主观因素对配送关系的影响。举例说明，如果配送员对目标订单的接单意愿不高，即使匹配指标符合要求，配送员也可以拒绝接单，从而无法真正形成配送关系，影响了调度准确性和调度效率。
发明内容
第一方面，本申请提供一种配送调度方法。所述配送调度方法包括：基于至少一个目标订单与至少一个目标配送员的组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径；计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标；基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
第二方面，本申请提供一种配送调度装置。所述配送调度装置包括：路径规划单元、计算单元和调度单元。路径规划单元被配置为基于至少一个目标订单与至少一个目标配送员形成的至少一个组合方式，规划出每种组合方式下目标配送员被分配目标订单后的 配送路径。计算单元被配置为计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标。调度单元被配置为基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
第三方面，本申请提供一种计算机可读存储介质。所述存储介质存储有计算机程序，所述计算机程序用于执行上述第一方面所述的配送调度方法。
第四方面，本申请提供一种电子设备。该电子设备包括：处理器和用于存储处理器可执行指令的存储器。所述处理器被配置为执行上述第一方面所述的配送调度方法。
本申请实施例，提供了一种配送调度方案，通过计算目标配送员对被分配目标订单的接单意愿指标，再将该接单意愿指标与配送效率指标相结合得到供调度系统参考的综合指标；调度系统基于该综合指标来确定是否进行调度。如此不仅考虑了配送效率指标这样的客观因素也考虑了配送员接单意愿这样的主观因素。当配送员被分配到订单后，由于配送效率指标和接单意愿指标均符合要求，因此有效增加了配送员接受订单的几率；从而可以有效提高调度准确性和调度效率。
附图说明
图1是本申请一示例性实施例示出的一种配送调度系统的结构示意图；
图2是本申请一示例性实施例示出的一种配送调度方法的流程图；
图3是本申请一示例性实施例示出的一种配送调度装置的硬件结构图；
图4是本申请一示例性实施例示出的一种配送调度装置的模块示意图。
具体实施方式
这里将详细地对示例性实施例进行说明，其示例表示在附图中。下面的描述涉及附图时，除非另有表示，不同附图中的相同数字表示相同或相似的要素。以下示例性实施例中所描述的实施方式并不代表与本申请相一致的所有实施方式。相反，它们仅是与如所附权利要求书中所详述的、本申请的一些方面相一致的装置和方法的例子。
在本申请使用的术语是仅仅出于描述特定实施例的目的，而非旨在限制本申请。在本申请和所附权利要求书中所使用的单数形式的“一种”、“所述”和“该”也旨在包括多数形式，除非上下文清楚地表示其他含义。还应当理解，本文中使用的术语“和/或”是指并 包含一个或多个相关联的列出项目的任何或所有可能组合。
应当理解，尽管在本申请可能采用术语第一、第二、第三等来描述各种信息，但这些信息不应限于这些术语。这些术语仅用来将同一类型的信息彼此区分开。例如，在不脱离本申请范围的情况下，第一信息也可以被称为第二信息，类似地，第二信息也可以被称为第一信息。取决于语境，如在此所使用的词语“如果”可以被解释成为“在……时”或“当……时”或“响应于确定”。
图1是本申请一示例性实施例示出的一种配送调度系统的架构示意图。该调度系统可以包括：数据收集模块101，路径规划模块102，接单意愿计算模块103和订单分配决策模块104。
在一实施例中，所述数据收集模块101收集的数据包括4类，分别为订单数据、配送员数据、环境数据和路径数据。
在一实施例中，所述订单数据可以包括如下至少一种：订单的配送距离、配送价格、配送时间段、物品价值、备货时长(订单创建至配送员能取货之间的时长)、最晚送达时刻、订单类型(如外卖、快递等即时配送类型)、订单所处区域、起点位置(如商家位置)、终点位置(如下单的目的地位置)等。
在一实施例中，所述配送员数据可以包括配送员历史数据和配送员实时数据。
其中，所述配送员历史数据可以包括如下至少一种：历史平均速度、历史平均日接单量、历史平均日拒单率、历史配送过的区域、历史配送过的配送申请方、历史不同配送距离订单的接单比例，历史不同配送时间段订单的接单比例，历史不同配送价格下的订单的接单比例。
其中，所述配送员实时数据可以包括如下至少一种：配送员等级、配送员位置。
在一实施例中，所述环境数据可以包括如下至少一种：当前配送区域的天气、当前配送区域在预设时长内新创建的订单数量、当前配送区域在预设时长内配送员负载数据，当前配送区域在预设时长内空闲的配送员数量、当前配送区域在预设时长内分配订单的取消率。
在一实施例中，所述路径数据可以包括如下至少一种：配送员与每个订单的起点位置的距离以及达到起点位置所需的时长；配送员与每个订单的终点位置的距离以及达到终点位置所需的时长；订单间起点位置的距离和时长；订单间终点位置的距离和时长； 订单间起点位子和终点位置的距离和时长。
所述数据收集模块101可以将上述收集到的原始数据转换为后续路径规划模块102、接单意愿计算模块103能够直接使用的数据格式。一般的，不同来源的数据往往存在数据格式不同无法直接被系统使用的问题，例如有些数据是结构化数据(如数据库数据)，有些数据是非结构化数据(如各种格式的办公文档、XML、HTML、报表、图片、音视频等)，这里数据收集模块101可以将所有收集到的数据转换为统一格式的标准化数据，从而方便其它模块直接使用。
在一实施例中，所述路径规划模块102用于规划配送员的配送路径，并且基于配送路径计算匹配程度和效率指标。如图1所示，规划配送路径需要用到所述数据收集模块101收集到的例如配送员数据、订单数据、环境数据、路径数据，从而基于配送员位置速度，订单的起终点位置，配送区域环境、配送区域路径等数据规划相应配送路径。进一步的，可以基于路径优化算法规划最优的配送路径，从而计算出最优的匹配指标和效率指标。其中，所述匹配指标表示目标配送员被分配目标订单前后配送路径之间的相似程度，所述效率指标表示目标配送员配送目标订单的效率高低程度。
其中，所述路径优化算法的目标为目标配送员被分配目标订单后规划的配送路径所需配送时长最短。
举例说明，假设获取到某个物流订单i，以及某个配送员j；所述配送员j已经有5个待配送的订单，其中有2个订单已取件，3个订单未取件；此时该配送员j共有8个目的地，即3个起点位置(对应那3个未取件订单)和5个终点位置。由于达到订单的起点位置和终点位置的顺序不同会形成不同的配送路径，并直接影响最终的配送时长，因此需要优化配送路径，使得总的配送时长最短。
需要说明的是，为了适应物流配送场景的业务逻辑限制，所述优化算法需要具有以下至少一种的约束条件。
1：目标配送员在配送订单时需先前往该订单的起点位置，再前往该订单的终点位置。在实际物流配送中，一个订单的完整配送过程必然是，配送员先前往订单的起点位置取货，然后才可以携带取得的货物前往订单的终点位置。
2：目标配送员被分配目标订单后的订单总数不能超过接单上限。在实际物流配送中，每个配送员可以配送的订单数量是有上限的。如果一个配送员同时接了过多的订单，那么就无法保证每个订单的时效性。过多订单往往意味着必然有部分订单会存在配送超时 的问题，因此可以设置配送员的接单上限。配送员被分配目标订单后的订单总数不能超过接单上限。所述接单上限可以是系统设置的，也可以是配送员自己根据实际情况设置的。
3：目标配送员被分配目标订单后，当前未完成订单和目标订单都在最晚送达时刻前送达。在实际物流配送中，每个订单在创建后，都会对应有一个最晚送达时刻，表示配送接收方最晚可能接受的送达时刻，如果实际送达时刻超过了这个最晚送达时刻，那么就属于配送超时。在配送单个订单是，一般预计送达时刻会早于这个最晚送达时刻，然而在同时配送多个订单时，由于配送路径会增加，每个订单的预计送达时刻也会相应变化，调度系统在进行调度时，必须保证规划的配送路径中每个订单的预计送达时刻都不超过最晚送达时刻。
4：该订单的备货时长与目标配送员前往订单的起点位置所需时长之差小于阈值。在实际物流配送中，不同配送申请方的备货时长都是不同的，配送员过早到达起点位置并不意味着可以马上取货。如果配送申请方还在备货，那么配送员必须等待，这样就浪费了宝贵的配送时间；因此，需要保证配送员达到起点位置后，可以马上或者尽快取货。为此，该订单的备货时长以及目标配送员前往订单的起点位置所需时长之间的差小于阈值时，说明配送申请方可以在配送员达到之前或者达到之后的短时间内就可以完成备货，方便配送员快速完成取货工作。
本申请中，所述路径优化算法可以包括模拟退火算法、蚁群算法、粒子算法等。
在一实施例中，所述接单意愿计算模块103用于计算配送员对被分配的订单的接单意愿指标。其中，所述接单意愿指标表示所述配送员对所述订单的接受程度。具体地，所述接单意愿计算模块103可以基于机器学习模型，根据数据收集模块101获取的订单数据、配送员数据和环境数据，以及根据路径规划模块102获取的匹配指标计算出接单意愿指标。
所述接单意愿模型通过如下方式训练得出：以历史订单的基础数据和匹配指标为训练数据，以所述历史订单被分配给配送员后配送员接受或者拒绝为标签，采用机器学习算法进行模型训练，将训练得到的模型确定为接单意愿模型。
所述机器学习算法可以包括xgboost、逻辑回归、随机森林、决策树、梯度提升决策树(Gradient Boost Decision Tree，GBDT)和支持向量机中的至少一种。
在一实施例中，所述订单分配决策模块104可以根据所述效率指标和接单意愿指标 计算出综合指标，然后订单分配决策模块104根据该综合指标确定是否按照对应的组合方式进行调度。在一示例中，订单分配决策模块104是决策器。
图2是本申请一示例性实施例示出的一种配送调度方法流程图。所述方法可以应用在上述的调度系统中，该方法具体可以包括如下步骤210～230。
步骤210：基于至少一个目标订单与至少一个目标配送员形成的至少一个组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径。
在一示例中，调度系统可以获取至少一种组合方式，该至少一种组合方式包括至少一个待分配的目标订单与至少一个目标配送员。如前所述，一个配送员可以同时配送多个订单，并且配送员有一个接单上限。而目标配送员是指前述的空闲的配送员，可以是指同时配送的订单数量未到达接单上限的配送员。
接着，调度系统可以规划出每种组合方式下每个目标配送员被分配目标订单后的配送路径。该步骤210可以由调度系统中前述的路径规划模块执行。
在一实施例中，所述规划出每种组合方式下目标配送员被分配目标订单后的配送路径，具体包括：规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
在一实施例中，所述最优的配送路径可以是指目标配送员被分配目标订单后规划的配送路径所需配送时长最短。
进一步的，所述规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径，具体包括：基于路径优化算法，规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
所述路径优化算法的目标为目标配送员被分配目标订单后规划的配送路径所需配送时长最短。
所述路径优化算法的约束条件包括以下至少一种：
目标配送员在配送目标订单时需先前往该目标订单的起点位置，再前往该目标订单的终点位置；
目标配送员被分配目标订单后的订单总数不能超过接单上限；
目标配送员被分配目标订单后，当前未完成订单和目标订单都在最晚送达时刻前送达；
目标订单的备货时长与目标配送员前往该目标订单的起点位置所需时长之差小于第一阈值。
步骤220：计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标。
在一实施例中，所述配送效率指标可以包括匹配指标和效率指标。
在一实施例中，所述步骤220具体可以包括以下步骤B1和B2。
步骤B1：计算每种组合方式下所述配送路径的匹配指标和效率指标；其中，所述匹配指标表示目标配送员被分配目标订单前后配送路径之间的相似程度，所述效率指标表示目标配送员配送目标订单的效率高低程度。
步骤B2：根据每种组合方式的匹配指标，计算各组合方式对应的目标配送员的接单意愿指标；其中，所述接单意愿指标表示目标配送员对目标订单的接受程度。
该步骤B1可以由调度系统中前述的路径规划模块执行。
在一示例中，所述匹配指标可以为0至1之间的数值。越接近1就表示相似程度越高；反之，越接近0表示相似程度越低。
在一示例中，所述效率指标可以为0至1之间的数值。越接近1表示目标配送员配送目标订单的效率越高；反之，越接近0表示目标配送员配送目标订单的效率越低。通常，如果目标订单的起点位置或者终点位置与目标配送员的其它订单的起点位置或者或终点位置较为接近，那么效率会比较高。
该步骤B2可以由调度系统中前述的接单意愿计算模块执行。
在一实施例中，步骤B2具体可以包括：获取每种组合方式下目标订单的基础数据；以及，将所述基础数据和匹配指标输入到接单意愿模型，获取所述接单意愿模型计算出的每种组合方式下的目标配送员的接单意愿指标。
在一实施例中，所述获取每种组合方式下目标订单的基础数据，具体包括：从每种组合方式下的目标配送员的历史接单数据中获取不同类型的接单比例；确定目标订单所属类型在历史接单数据中的接单比例；以及将所确定的接单比例作为所述目标订单的基础数据。
在一实施例中，所述不同类型包括如下至少一种：不同配送距离，不同配送时间段，不同配送价格和不同配送区域。
例如，从目标配送员的历史数据中获取不同配送距离的接单比例，结合目标订单的配送距离，确定该目标订单的接单比例。该接单比例可以反映出目标配送员对这个配送 距离的订单的喜好程度。
再例如，从目标配送员的历史数据中获取不同配送时间段的接单比例，结合目标订单的配送时间段，确定该目标订单的接单比例。该接单比例可以反映出目标配送员对这个配送时间段的订单的喜好程度。
再例如，从目标配送员的历史数据中获取不同配送价格的接单比例，结合目标订单的配送价格，确定该目标订单的接单比例。该接单比例可以反映出目标配送员对这个配送价格的订单的喜好程度。
再例如，从目标配送员的历史数据中获取不同配送区域的接单比例，结合目标订单的配送区域，确定该目标订单的接单比例。该接单比例可以反映出目标配送员对这个配送区域的订单的喜好程度。值得一提的是，可以采用geohash算法，对配送员历史配送过的区域进行编码，按照经纬度将这些区域划分为等状区块，统计目标配送员在不同区块上历史配送次数；同样地，根据geohash算法可以确定目标订单的配送申请方地理位置和/或配送接收方地理位置所在的目标区块；从上述统计出的不同区块上历史配送次数获取目标区域的历史配送次数。
在一实施例中，所述接单意愿模型通过如下方式训练得出：以历史订单的基础数据和匹配指标为训练数据，以所述历史订单被分配给配送员后配送员接受或者拒绝为标签，采用机器学习算法进行模型训练，将训练得到的模型确定为接单意愿模型。
所述机器学习算法包括xgboost、逻辑回归、随机森林、决策树、GBDT、支持向量机中的至少一种。
步骤230：综合每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
该实施例中，调度系统可以综合每种组合方式的配送效率指标和接单意愿指标，从所有组合方式中选取最优的组合方式进行配送调度。该步骤可以由调度系统中前述的订单分配决策模块执行。
在一实施例中，所述步骤230，具体可以包括步骤A1和步骤A2。
步骤A1：根据每种组合方式的配送效率指标和接单意愿指标，计算每种组合方式的综合指标。
步骤A2：根据每种组合方式的综合指标，从所有组合方式中选取最优的组合方式进 行配送调度。
步骤A1和A2可以由调度系统中前述的订单分配决策模块执行。
在一实施例中，所述步骤A1具体包括：将每种组合方式的配送效率指标乘以效率权重，得到效率值；将每种组合方式的接单意愿指标乘以意愿权重，得到意愿值；将每种组合方式的所述效率值和意愿值求和，得到每种组合方式对应的综合指标。其中，所述效率权重和意愿权重之和为1。
如前所述，配送效率指标可以包括匹配指标和效率指标；因此，该实施例中，效率值具体可以是配送效率指标中的效率指标乘以效率权重，从而得到效率值。
在一实施例中，在目标订单为1个、目标配送员为1个，组合方式为1种时；所述步骤A2具体包括：在所述1种组合方式的综合指标大于第二阈值时，确定按照所述组合方式进行配送调度。
在一实施例中，在目标订单为1个、目标配送员为N个，组合方式为N种，N为大于1的自然数时；所述步骤A2具体包括：从所述N种综合指标中，选取最大的综合指标，并按照所述最大的综合指标对应的组合方式进行配送调度。
在一实施例中，在目标订单为M个、目标配送员为N个，组合方式为M*N种，M和N均大于1的自然数时；所述步骤A2具体包括：基于决策算法，从M行*N列的综合指标中每行选取1个，使得M个综合指标之和最大；其中，所述选取的M个综合指标对应的组合方式的目标订单不能重复；按照所选取的M个综合指标对应的组合方式进行配送调度。
举例说明，假设有M个目标订单，N个目标配送员，则相应的可以有M*N种不同的组合方式。同样也可以有M*N个效率指标和接单意愿指标，设定第i个配送员对第j个订单的效率指标为e ij，接单意愿指标为w ij。那么可以这M个订单和N个配送员可以构建一个M行N列的矩阵，该矩阵中第i行第j列的值即为一个综合指标的数值，记作p ij。
本申请中，p ij＝λ*w ij+(1-λ)*e ij，其中λ可以表示效率权重，该效率权重可以是人为预先设置的经验值；相应地1-λ可以表示意愿权重。订单分配决策模块的目标是给每个订单分配一个最合适的配送员，使得所有订单(M个订单)的p之和最大。这里的约束条件为每个订单只能分配给一个配送员，每个配送员设有接单上限。求解上述公式类 似二分图最大权完美匹配方式，可以采用例如KM算法、匈牙利算法等决策算法。
本申请实施例提供了一种配送调度方法，通过计算目标配送员对被分配目标订单的接单意愿指标，再将该接单意愿指标与配送效率指标相结合得到供调度系统参考的综合指标；调度系统基于综合指标来确定是否进行调度；如此不仅考虑了配送效率指标这样的客观因素也考虑了配送员接单意愿这样的主观因素，当配送员被分配到订单后，由于配送效率指标和接单意愿指标均符合要求，因此有效增加了配送员接受订单的几率；从而可以有效提高调度准确性和调度效率。
随着物流配送业务的不断增长，现有物流配送资源越来越无法满足即使配送需求。例如，专业的配送员团队人数有限，而配送需求量却与日俱增，有限的配送员远远无法满足日常配送需求，从而导致配送订单的积压和延误。在全职配送员无法快速增长的情况下，通过调动社会闲散劳动力参与物流配送业务的物流配送新模式随之产生。例如O2O(Online To Offline)众包模式。与传统的基于全职配送员进行物流配送不同的是，由于这些兼职配送员通常只在顺路情况下才会接单，而在不顺路的情况下往往不愿意接单。因此，在O2O众包模式中，兼职配送员可以选择接受或者拒绝被分配的物流订单。本申请所述的配送调度方案不仅可以适用于传统的全职配送员模式，同样也可以适用于所述O2O众包模式。通过综合配送路径的配送效率指标以及兼职配送员的接单意愿进行配送调度，使得兼职配送员接受被分配订单的几率大大增加，从而有效提升调度准确性和调度效率。
本申请提供了配送调度装置的实施例。该装置实施例可以应用在服务器上。装置实施例可以通过软件实现，也可以通过硬件或者软硬件结合的方式实现。以软件实现为例，作为一个逻辑意义上的装置，是通过其所在处理器将非易失性存储器中对应的计算机程序指令读取到内存中运行形成的。从硬件层面而言，如图3所示，为本申请配送调度装置所在的一种硬件结构图，除了图3所示的处理器、内存、网络接口、以及非易失性存储器之外，实施例中通常根据该配送调度的实际功能，还可以包括其他硬件，对此不再赘述。
请参考图4，在一种软件实施方式中，该配送调度装置可以包括：路径规划单元310、计算单元320和调度单元330。
路径规划单元310被配置为，基于至少一个目标订单与至少一个目标配送员形成的至少一种组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径。
计算单元320被配置为，计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标。
调度单元330被配置为，基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
可选的，计算单元320具体包括：第一计算子单元和第二计算子单元。
第一计算子单元被配置为，计算每种组合方式下所述配送路径的匹配指标和效率指标。其中，所述匹配指标表示目标配送员被分配目标订单前后配送路径之间的相似程度，所述效率指标表示目标配送员配送目标订单的效率高低程度。
第二计算子单元被配置为，根据每种组合方式的匹配指标，计算每种组合方式下目标配送员的接单意愿指标。其中，所述接单意愿指标表示目标配送员对目标订单的接受程度。
可选的，所述路径规划单元310具体包括：获取子单元和路径规划子单元。
获取子单元被配置为，获取包括至少一个待分配的目标订单与至少一个目标配送员的至少一个组合方式。
路径规划子单元被配置为规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
可选的，所述路径规划子单元具体包括：基于路径优化算法，规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
可选的，所述路径优化算法的目标为目标配送员被分配目标订单后规划的配送路径所需配送时长最短。
可选的，所述路径优化算法的约束条件包括以下至少一种：
目标配送员在配送目标订单时需先前往该目标订单的起点位置，再前往该目标订单的终点位置；
目标配送员被分配目标订单后的订单总数不能超过接单上限；
目标配送员被分配目标订单后，当前未完成订单和目标订单都在最晚送达时刻前送达；
目标订单的备货时长与目标配送员前往该目标订单的起点位置所需时长与之差小于第一阈值。
可选的，所述优化算法包括模拟退火算法、蚁群算法和粒子算法中的至少一种。
可选的，所述第二计算子单元具体包括：获取子单元和计算子单元。
获取子单元被配置为获取每种组合方式下目标订单的基础数据；
计算子单元被配置为，将所述基础数据和匹配指标输入到接单意愿模型，获取所述接单意愿模型计算出的对应目标配送员的接单意愿指标。
可选的，所述获取子单元具体包括：比例获取子单元、比例确定子单元和数据确定子单元。
比例获取子单元被配置为，从每种组合方式下的目标配送员的历史接单数据中获取不同类型的接单比例。
比例确定子单元被配置为，确定目标订单所属类型在历史接单数据中的接单比例。
数据确定子单元被配置为，将所确定的接单比例作为所述目标订单的基础数据。
可选的，所述不同类型包括：不同配送距离，不同配送时间段，不同配送价格和不同配送区域中的至少一种。
可选的，所述接单意愿模型通过如下方式训练得出：以历史订单的基础数据和匹配指标为训练数据，以所述历史订单被分配给配送员后配送员接受或者拒绝为标签，采用机器学习算法进行模型训练，将训练得到的模型确定为接单意愿模型。
可选的，所述机器学习算法包括xgboost、逻辑回归、随机森林、决策树、梯度提升决策树(Gradient Boost Decision Tree，GBDT)和支持向量机中的至少一种。
可选的，所述调度单元330具体包括：第一调度子单元和第二调度子单元。
第一调度子单元被配置为，根据每种组合方式的配送效率指标和接单意愿指标，计算每种组合方式的综合指标。
第二调度子单元被配置为，根据每种组合方式的综合指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
可选的，所述第一调度子单元具体包括：第一计算子单元、第二计算子单元和第三计算子单元。
第一计算子单元被配置为，将每种组合方式的配送效率指标乘以效率权重，得 到效率值。
第二计算子单元被配置为，将每种组合方式的接单意愿指标乘以意愿权重，得到意愿值。
第三计算子单元被配置为，将每种组合方式的所述效率值和意愿值求和，得到每种组合方式对应的综合指标。其中，所述效率权重和意愿权重之和为1。
可选的，在目标订单为1个、目标配送员为1个、组合方式为1种时，所述第二调度子单元具体包括：在所述综合指标大于第二阈值时，按照所述1种组合方式进行调度。
可选的，在目标订单为1个、目标配送员为N个、组合方式为N种，N为大于1的自然数时，所述第二调度子单元具体包括：从所述N种综合指标中，选取最大的综合指标，并按照所述最大的综合指标对应的组合方式进行配送调度。
可选的，在待分配的目标订单为M个、空闲的目标配送员为N个、组合方式为M*N种、M和N均大于1的自然数时，所述第二调度子单元包括选取子单元和调度子单元。
选取子单元被配置为，基于决策算法，从M行*N列的综合指标中每行选取1个，使得M个综合指标之和最大。其中，所述选取的M个综合指标对应的组合方式的目标订单不能重复。
调度子单元被配置为按照所选取的M个综合指标对应的组合方式进行配送调度。
可选的，所述决策算法包括KM算法和匈牙利算法中的至少一种。
上述装置中各个单元的功能和作用的实现过程具体详见上述方法中对应步骤的实现过程，在此不再赘述。
对于装置实施例而言，由于其基本对应于方法实施例，所以相关之处参见方法实施例的部分说明即可。以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本申请方案的目的。本领域普通技术人员在不付出创造性劳动的情况下，即可以理解并实施。
本申请实施例提供一种计算机可读存储介质，所述存储介质存储有计算机程序， 所述计算机程序用于执行上述任一配送调度方法实施例。
本申请实施例还提供一种电子设备，如图3所示，该电子设备包括处理器和存储器。该存储器例如包括内存和非易失性存储器。存储器用于存储处理器可执行指令。处理器被配置为执行上述任一配送调度方法实施例。
本申请中的各个实施例均采用递进的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于电子设备实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
以上所述仅为本申请的较佳实施例而已，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。

Claims (21)

  1. 一种配送调度方法，所述方法包括：
    基于至少一个目标订单与至少一个目标配送员形成的至少一种组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径；
    计算每种组合方式下所述配送路径的、与所述目标配送员被分配目标订单关联的配送效率指标和接单意愿指标；
    基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
  2. 根据权利要求1所述的方法，
    所述配送效率指标包括匹配指标和效率指标，
    所述计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联的配送效率指标和接单意愿指标，包括：
    计算每种组合方式下所述配送路径的匹配指标和效率指标；其中，所述匹配指标表示目标配送员被分配目标订单前后配送路径之间的相似程度，所述效率指标表示目标配送员配送目标订单的效率高低程度；
    根据每种组合方式的匹配指标，计算每种组合方式下目标配送员的接单意愿指标；其中，所述接单意愿指标表示目标配送员对目标订单的接受程度。
  3. 根据权利要求1所述的方法，所述规划出每种组合方式下目标配送员被分配目标订单后的配送路径包括：
    规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
  4. 根据权利要求3所述的方法，所述规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径包括：
    基于路径优化算法，规划出每种组合方式下目标配送员被分配目标订单后最优的配送路径。
  5. 根据权利要求4所述的方法，所述路径优化算法的目标为目标配送员被分配目标订单后规划的配送路径所需配送时长最短。
  6. 根据权利要求4所述的方法，所述路径优化算法的约束条件包括以下至少一种：
    目标配送员在配送目标订单时需先前往所述目标订单的起点位置，再前往所述目标订单的终点位置；
    目标配送员被分配目标订单后的订单总数不能超过接单上限；
    目标配送员被分配目标订单后，当前未完成订单和目标订单都在最晚送达时刻前送 达；
    所述目标订单的备货时长与所述目标配送员前往所述目标订单的起点位置所需时长之差小于第一阈值。
  7. 根据权利要求4至6中任一项所述的方法，所述优化算法包括模拟退火算法、蚁群算法和粒子算法中的至少一种。
  8. 根据权利要求2所述的方法，所述根据每种组合方式的匹配指标，计算每种组合方式下目标配送员的接单意愿指标包括：
    获取每种组合方式下目标订单的基础数据；
    将所述基础数据和所述匹配指标输入到接单意愿模型，获取所述接单意愿模型计算出的每种组合方式下目标配送员的接单意愿指标。
  9. 根据权利要求8所述的方法，所述获取每种组合方式下目标订单的基础数据包括：
    从每种组合方式下的目标配送员的历史接单数据中获取不同类型的接单比例；
    确定目标订单所属类型在历史接单数据中的接单比例；
    将所确定的接单比例作为所述目标订单的基础数据。
  10. 根据权利要求9所述的方法，所述不同类型包括：
    不同配送距离，不同配送时间段，不同配送价格和不同配送区域中的至少一种。
  11. 根据权利要求8所述的方法，所述接单意愿模型通过如下方式训练得出：
    以历史订单的基础数据和匹配指标为训练数据，以所述历史订单被分配给配送员后配送员接受或者拒绝为标签，采用机器学习算法进行模型训练，将训练得到的模型确定为接单意愿模型。
  12. 根据权利要求11所述的方法，所述机器学习算法包括xgboost、逻辑回归、随机森林、决策树、梯度提升决策树和支持向量机中的至少一种。
  13. 根据权利要求1所述的方法，所述基于所述每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度，包括：
    根据每种组合方式的配送效率指标和接单意愿指标，计算每种组合方式的综合指标；
    根据每种组合方式的所述综合指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
  14. 根据权利要求13所述的方法，所述根据每种组合方式的配送效率指标和接单意愿指标，计算每种组合方式的综合指标，包括：
    将每种组合方式的配送效率指标乘以效率权重，得到效率值；
    将每种组合方式的接单意愿指标乘以意愿权重，得到意愿值；
    将每种组合方式的所述效率值和意愿值求和，得到每种组合方式对应的综合指标；其中，所述效率权重和所述意愿权重之和为1。
  15. 根据权利要求13所述的方法，在目标订单为1个、目标配送员为1个，组合方式为1种时；
    所述根据每种组合方式的综合指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度，包括：
    在所述1种组合方式的综合指标大于第二阈值时，按照所述组合方式进行配送调度。
  16. 根据权利要求13所述的方法，在目标订单为1个、目标配送员为N个，组合方式为N种，N为大于1的自然数时；
    所述根据每种组合方式的综合指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度，包括：
    从所述N种综合指标中，选取最大的综合指标，并按照所述最大的综合指标对应的组合方式进行配送调度。
  17. 根据权利要求13所述的方法，在目标订单为M个、目标配送员为N个，组合方式为M*N种，M和N均大于1的自然数时；
    所述根据每种组合方式的综合指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度，包括：
    基于决策算法，从M行*N列的综合指标中每行选取1个，使得M个综合指标之和最大；其中，所述选取的M个综合指标对应的组合方式的目标订单不能重复；
    按照所选取的M个综合指标对应的组合方式进行配送调度。
  18. 根据权利要求17所述的方法，所述决策算法包括KM算法和匈牙利算法中的至少一种。
  19. 一种配送调度装置，所述装置包括：
    路径规划单元，基于至少一个目标订单与至少一个目标配送员形成的至少一种组合方式，规划出每种组合方式下目标配送员被分配目标订单后的配送路径；
    计算单元，计算每种组合方式下所述配送路径的、与目标配送员被分配目标订单关联之间的配送效率指标和接单意愿指标；
    调度单元，基于每种组合方式的配送效率指标和接单意愿指标，从所述至少一种组合方式中选取最优的组合方式进行配送调度。
  20. 一种计算机可读存储介质，所述存储介质存储有计算机程序，所述计算机程序 用于执行上述权利要求1-18中任一项所述的方法。
  21. 一种电子设备，包括：
    处理器；和
    用于存储所述处理器可执行指令的存储器；
    所述处理器被配置为实现上述权利要求1至18中任一项所述的方法。
PCT/CN2019/099714 2018-08-08 2019-08-08 配送调度 WO2020030028A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/266,624 US20210312347A1 (en) 2018-08-08 2019-08-08 Dispatching distribution

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810899170.6 2018-08-08
CN201810899170.6A CN109214551B (zh) 2018-08-08 2018-08-08 一种配送调度方法及装置

Publications (1)

Publication Number Publication Date
WO2020030028A1 true WO2020030028A1 (zh) 2020-02-13

Family

ID=64988511

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/099714 WO2020030028A1 (zh) 2018-08-08 2019-08-08 配送调度

Country Status (3)

Country Link
US (1) US20210312347A1 (zh)
CN (1) CN109214551B (zh)
WO (1) WO2020030028A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112951004A (zh) * 2021-03-09 2021-06-11 东部机场集团有限公司 一种多目标多时段航班保障资源动态优化分配方法
CN113112203A (zh) * 2021-04-14 2021-07-13 杭州拼便宜网络科技有限公司 一种基于混合蚁群算法多配送中心车辆路径安排系统
US20220292450A1 (en) * 2021-03-12 2022-09-15 Nec Corporation Delivery support device, delivery support method, and storage medium
EP4030368A4 (en) * 2020-02-21 2023-08-30 Beijing Jingdong Zhenshi Information Technology Co., Ltd. ROUTE DETERMINATION METHOD AND DEVICE FOR COLD CHAIN DISTRIBUTION, SERVER AND STORAGE MEDIUM

Families Citing this family (30)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109214551B (zh) * 2018-08-08 2022-08-26 北京三快在线科技有限公司 一种配送调度方法及装置
US20200118071A1 (en) * 2018-10-13 2020-04-16 Walmart Apollo, Llc Delivery prediction generation system
RU2739873C2 (ru) * 2019-02-07 2020-12-29 Акционерное общество "Лаборатория Касперского" Способ поиска пользователей, соответствующих требованиям
CN109934537A (zh) * 2019-03-12 2019-06-25 北京同城必应科技有限公司 订单分配方法、装置、服务器和存储介质
CN111815212A (zh) * 2019-04-11 2020-10-23 北京三快在线科技有限公司 路径规划方法、装置及存储介质
CN110097288B (zh) * 2019-05-08 2023-11-10 哈尔滨工业大学（威海） 一种基于图搜索的城市众包配送任务分配方法及装置
CN110231044B (zh) * 2019-06-10 2020-09-04 北京三快在线科技有限公司 一种路径规划方法及装置
CN110288232A (zh) * 2019-06-25 2019-09-27 北京三快在线科技有限公司 订单调度方法及装置
CN112541716A (zh) * 2019-09-20 2021-03-23 北京三快在线科技有限公司 选择待配送任务节点的方法、装置、存储介质和电子设备
CN110689254A (zh) * 2019-09-23 2020-01-14 拉扎斯网络科技（上海）有限公司 数据处理方法、装置、电子设备和计算机可读存储介质
CN110751433A (zh) * 2019-09-25 2020-02-04 北京三快在线科技有限公司 订单的分配方法和装置，电子设备及存储介质
CN112836914A (zh) * 2019-11-25 2021-05-25 北京三快在线科技有限公司 订单结构的调整方法、装置、存储介质及电子设备
CN112966887B (zh) * 2019-12-13 2024-05-28 多点(深圳)数字科技有限公司 生成分配信息的方法、装置、电子设备和介质
CN113222487B (zh) * 2020-01-21 2023-04-18 北京三快在线科技有限公司 调度路径生成方法，装置，存储介质及电子设备
CN111553526B (zh) * 2020-04-24 2023-08-25 新石器慧通（北京）科技有限公司 一种物品配送方法及装置
CN113673736A (zh) * 2020-05-13 2021-11-19 北京三快在线科技有限公司 运单分配方法、装置、存储介质和电子设备
CN111815229B (zh) * 2020-06-12 2024-01-23 北京顺达同行科技有限公司 订单信息的处理方法、装置、电子设备及存储介质
CN114202132A (zh) * 2020-09-02 2022-03-18 北京三快在线科技有限公司 订单分配方法、装置、存储介质及电子设备
CN112668924A (zh) * 2021-01-05 2021-04-16 拉扎斯网络科技（上海）有限公司 物流运力的调配方法、装置及系统
CN113095553A (zh) * 2021-03-29 2021-07-09 北京沃东天骏信息技术有限公司 调度方法、装置、电子设备及存储介质
CN114066256A (zh) * 2021-11-17 2022-02-18 北京同城必应科技有限公司 一种在供需调度下支撑运营体系运行策略的解决方案
CN114254902B (zh) * 2021-12-13 2024-04-02 四川启睿克科技有限公司 一种多产线人员调度方法
CN114997767A (zh) * 2022-02-14 2022-09-02 山东开创云计算有限公司 一种多订单同向配送规划系统
CN114626913A (zh) * 2022-03-03 2022-06-14 南京领行科技股份有限公司 一种订单处理方法、装置及电子设备
CN114581020B (zh) * 2022-04-28 2022-08-09 深圳市运无忧网络科技有限公司 一种智能物流管理方法和系统
CN114819755A (zh) * 2022-06-24 2022-07-29 浙江口碑网络技术有限公司 订单调度方法、装置、电子设备和存储介质
CN115034727B (zh) * 2022-08-06 2022-12-02 浙江口碑网络技术有限公司 一种运单处理方法、装置及电子设备
CN116228088A (zh) * 2023-03-01 2023-06-06 深圳市中源盛科技有限公司 基于事件与人员联合匹配的车辆调度方法、装置及设备
CN117252322B (zh) * 2023-11-10 2024-02-02 青岛冠成软件有限公司 一种物流供应链管理方法
CN117455200B (zh) * 2023-12-22 2024-03-29 烟台大学 众包环境下的多阶段任务分配方法、系统、设备及介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070194912A1 (en) * 2006-01-10 2007-08-23 Lg Chem, Ltd. Method for optimal multi-vehicle dispatch and system for the same
CN107133697A (zh) * 2017-05-03 2017-09-05 百度在线网络技术（北京）有限公司 预估司机接单意愿的方法、装置、设备及存储介质
CN107220789A (zh) * 2017-05-15 2017-09-29 浙江仟和网络科技有限公司 一种物流配送调度方法及系统
CN108229864A (zh) * 2018-03-05 2018-06-29 北京三快在线科技有限公司 订单的分配方法、装置及电子设备
CN109214551A (zh) * 2018-08-08 2019-01-15 北京三快在线科技有限公司 一种配送调度方法及装置

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2016119749A1 (zh) * 2015-01-29 2016-08-04 北京嘀嘀无限科技发展有限公司 一种订单分配系统及方法
CN105117777A (zh) * 2015-07-28 2015-12-02 北京嘀嘀无限科技发展有限公司 一种订单分配的方法及装置
CN107092997A (zh) * 2016-07-29 2017-08-25 北京小度信息科技有限公司 一种物流调度方法及装置
CN107748923A (zh) * 2016-08-29 2018-03-02 北京三快在线科技有限公司 订单处理方法、装置及服务器
CN106530188B (zh) * 2016-09-30 2021-06-11 百度在线网络技术（北京）有限公司 在线叫车服务平台中司机的接单概率评价方法和装置
CN107392405B (zh) * 2017-01-26 2018-06-01 北京小度信息科技有限公司 数据处理方法、装置及设备
CN107230014B (zh) * 2017-05-15 2020-11-03 浙江仟和网络科技有限公司 一种末端即时物流的智能调度系统
CN107844930B (zh) * 2017-06-26 2020-09-18 北京星选科技有限公司 订单分配方法及装置
CN107357852A (zh) * 2017-06-28 2017-11-17 镇江五八到家供应链管理服务有限公司 一种货运司机对订单意愿的判断方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070194912A1 (en) * 2006-01-10 2007-08-23 Lg Chem, Ltd. Method for optimal multi-vehicle dispatch and system for the same
CN107133697A (zh) * 2017-05-03 2017-09-05 百度在线网络技术（北京）有限公司 预估司机接单意愿的方法、装置、设备及存储介质
CN107220789A (zh) * 2017-05-15 2017-09-29 浙江仟和网络科技有限公司 一种物流配送调度方法及系统
CN108229864A (zh) * 2018-03-05 2018-06-29 北京三快在线科技有限公司 订单的分配方法、装置及电子设备
CN109214551A (zh) * 2018-08-08 2019-01-15 北京三快在线科技有限公司 一种配送调度方法及装置

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP4030368A4 (en) * 2020-02-21 2023-08-30 Beijing Jingdong Zhenshi Information Technology Co., Ltd. ROUTE DETERMINATION METHOD AND DEVICE FOR COLD CHAIN DISTRIBUTION, SERVER AND STORAGE MEDIUM
CN112951004A (zh) * 2021-03-09 2021-06-11 东部机场集团有限公司 一种多目标多时段航班保障资源动态优化分配方法
CN112951004B (zh) * 2021-03-09 2022-04-08 东部机场集团有限公司 一种多目标多时段航班保障资源动态优化分配方法
US20220292450A1 (en) * 2021-03-12 2022-09-15 Nec Corporation Delivery support device, delivery support method, and storage medium
CN113112203A (zh) * 2021-04-14 2021-07-13 杭州拼便宜网络科技有限公司 一种基于混合蚁群算法多配送中心车辆路径安排系统

Also Published As

Publication number Publication date
US20210312347A1 (en) 2021-10-07
CN109214551A (zh) 2019-01-15
CN109214551B (zh) 2022-08-26

Similar Documents

Publication Publication Date Title
WO2020030028A1 (zh) 配送调度
CN107844915B (zh) 一种基于话务预测的呼叫中心的自动排班方法
FI111196B (fi) Käytettävyysprosessori ja -menetelmä
CN110046749B (zh) 一种基于实时路况的电商包裹与同城o2o包裹共同配送系统
CN104471573B (zh) 更新高速缓存的数据库查询结果
US20220004985A1 (en) Method and system for selection of a path for deliveries
CN109165895B (zh) 一种基于配送业务的定价方法及装置
WO2013160721A1 (en) Database system using batch-oriented computation
CN102197395A (zh) 存储器端存储请求管理
CN110544068B (zh) 基于快递柜的派件路径生成方法、装置、设备和存储介质
CN110782064A (zh) 一种集车辆调度优化与任务分配于一体的可视化方法及系统
Chen et al. Courier dispatch in on-demand delivery
CN113764087B (zh) 医院物资配送方法、装置、服务器及存储介质
CN109345091A (zh) 基于蚁群算法的整车物流调度方法及装置、存储介质、终端
US11625673B2 (en) Methods and systems for path selection using vehicle route guidance
KR102411671B1 (ko) 기상정보를 활용한 외식기업의 매출 추정 시스템 및 매출 추정액의 산출 방법
Dai et al. O2O on-demand delivery optimization with mixed driver forces
CN113592440A (zh) 一种基于大数据的智慧物流取件分析系统及方法
CN109635069A (zh) 一种基于信息熵的地理空间数据自组织方法
CN116090942A (zh) 基于物联网的多场景机器人配送方法及系统
CN109583634A (zh) 一种基于现代投资组合理论的外卖配送路径选择方法
CN109492008A (zh) 一种基于HBase的网络大数据模型设计方法及系统
US20150317588A1 (en) Predicting resources required by a customer and their cost
Lamsal et al. Continuous time scheduling for sugarcane harvest logistics in Louisiana
CN107274135A (zh) 基于协同信息共享的烟叶原料统筹方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19846972

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19846972

Country of ref document: EP

Kind code of ref document: A1