KR20220130945A - Apparatus, method and computer program for executing assignment of logistics vehicles - Google Patents

Abstract

An apparatus that dispatches logistics vehicles includes: an order information generation unit that generates virtual order information based on past transportation data; a storage unit for storing the virtual order information and vehicle information of the logistics vehicle in a buffer; a simulation unit for simulating the dispatch time of the logistics vehicle by including a decision unit for determining whether to dispatch the logistics vehicle based on the virtual order information and the vehicle information, and by including an updating unit for distributing the logistics vehicle through a preset automatic dispatch model based on whether the logistics vehicle is dispatched or not and updating vehicle information stored in the buffer based on a result of the dispatch of the logistics vehicle; a dispatch time determination unit for determining the dispatch time of the logistics vehicle with respect to real-time order information based on the result of the simulation; and a dispatch instruction unit for instructing dispatch of the logistics vehicle based on the dispatch time of the logistics vehicle. The order information generation unit, the storage unit, and the simulation unit function in an environment of a reinforcement learning model. The decision unit functions as an agent of the reinforcement learning model. The present invention simulates the dispatch time of a logistics vehicle to update vehicle information stored in a buffer.

Description

Apparatus, method and computer program for distributing logistics vehicles

The present invention relates to an apparatus, method and computer program for distributing a logistics vehicle.

Logistics refers to the act of moving goods or services from producers to consumers or a specific region, or inventory planning, delivery planning, procurement planning, etc. for this purpose. Logistics service creates clusters for transportation points in districts/dong units of administrative districts for ease of transportation. Here, the person in charge of the logistics company directly set the number and density of delivery points in the cluster, so it was dependent on the manpower-based know-how.

Recently, a logistics transportation plan is established in real time by an algorithm, and in this regard, Korea Patent Publication No. 2010-0062848, which is a prior art, discloses a logistics operation method and a system thereof.

Logistics transport drivers are instructed to dispatch the ordered goods based on the logistic transport plan. However, since the logistics and transport drivers have different preferences with respect to the item of the article, the delivery location, and the like, satisfaction with the dispatch route through which they will transport the article may vary. In consideration of this, automatic dispatch using AI (Artificial Intelligence) or heuristic automatic dispatch system has been used recently. Have.

Determines whether to dispatch a logistics vehicle based on virtual order information and vehicle information, dispatches a logistics vehicle through a preset automatic dispatch model based on whether the logistics vehicle is dispatched, and stores in a buffer based on the distribution result of the logistics vehicle An object of the present invention is to provide an apparatus, a method, and a computer program for simulating the dispatch time of a logistics vehicle to update stored vehicle information.

An object of the present invention is to provide an apparatus, a method, and a computer program for determining a dispatch time of a logistics vehicle with respect to real-time order information based on a simulation result and instructing the dispatch of a logistics vehicle based on the dispatch time of the logistics vehicle.

It is intended to provide an apparatus, method, and computer program that enables the GBP (Generator, Buffer, Processor) model to function as an environment for the reinforcement learning model and as an agent for the reinforcement learning model.

However, the technical problems to be achieved by the present embodiment are not limited to the technical problems described above, and other technical problems may exist.

As a means for achieving the above-described technical problem, an embodiment of the present invention is an order information generating unit that generates virtual order information based on past transportation data, and buffers the virtual order information and vehicle information of the distribution vehicle. The logistics vehicle through a storage unit to store in, a determination unit for determining whether to dispatch the distribution vehicle based on the virtual order information and the vehicle information, and a preset automatic distribution model based on whether the distribution vehicle is dispatched A simulation unit simulating the dispatch time of the distribution vehicle, including an update unit for distributing the vehicle, and updating the vehicle information stored in the buffer based on the distribution result of the distribution vehicle, and the real-time order information based on the result of the simulation and a dispatch time determining unit for determining the dispatch time of the distribution vehicle and a dispatch instruction unit for instructing the dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle, wherein the order information generation unit, the storage unit, and the simulation unit are reinforced It may provide a vehicle dispatching device that functions as an environment of the learning model, and the determining unit functions as an agent of the reinforcement learning model.

Another embodiment of the present invention includes the steps of generating virtual order information based on past transportation data by an order information generating unit, and storing the virtual order information and vehicle information of a logistics vehicle in a buffer by a storage unit , determines whether the distribution vehicle is dispatched based on the virtual order information and the vehicle information by the determining unit of the simulation unit, and a preset automatic dispatch model based on whether the distribution vehicle is dispatched by the update unit of the simulation unit distributing the logistics vehicle through and simulating the dispatch timing of the logistics vehicle to update the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle, based on the result of the simulation for real-time order information Determining the dispatch time of the distribution vehicle and instructing the distribution of the distribution vehicle based on the dispatch time of the distribution vehicle, wherein the order information generation unit, the storage unit, and the simulation unit of the reinforcement learning model It functions as an environment, and the determining unit may provide a method for performing dispatch that functions as an agent of the reinforcement learning model.

Another embodiment of the present invention, when the computer program is executed by the computing device, generates virtual order information based on past transportation data by the order information generating unit, and the virtual order information and logistics by the storage unit Storing the vehicle information of the vehicle in a buffer, determining whether to dispatch the distribution vehicle based on the virtual order information and the vehicle information by the determination unit of the simulation unit, and dispatching the distribution vehicle by the update unit of the simulation unit Dispatching the distribution vehicle through a preset automatic distribution model based on whether or not, simulates the dispatch time of the distribution vehicle to update the vehicle information stored in the buffer based on the dispatch result of the distribution vehicle, and a dispatch time determination unit determines the dispatch time of the distribution vehicle with respect to the real-time order information based on the result of the simulation, and instructs the dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle by the dispatch instructing unit, and generates the order information The unit, the storage unit, and the simulation unit may provide a computer program stored in a medium including a sequence of instructions to function as an environment of the reinforcement learning model, and the determination unit to function as an agent of the reinforcement learning model.

The above-described problem solving means are merely exemplary, and should not be construed as limiting the present invention. In addition to the exemplary embodiments described above, there may be additional embodiments described in the drawings and detailed description.

According to any one of the above-described problem solving means of the present invention, virtual order information and vehicle information of a logistics vehicle are generated and stored in a buffer, and a logistics vehicle based on a reinforcement learning model using the order information and vehicle information stored in the buffer It is possible to provide an apparatus, a method and a computer program for learning to determine whether to dispatch or not.

1 is a block diagram of an apparatus for performing a vehicle dispatch according to an embodiment of the present invention.
2 is an exemplary diagram for explaining a process of generating virtual order information according to an embodiment of the present invention.
3 is an exemplary view for explaining a process of storing virtual order information and vehicle information of a logistics vehicle in a buffer according to an embodiment of the present invention.
4 is an exemplary view for explaining a process of determining whether to dispatch a logistics vehicle based on order information and vehicle information according to an embodiment of the present invention.
5 is an exemplary view for explaining a process of simulating a dispatch time of a logistics vehicle according to an embodiment of the present invention.
6 is a flowchart of a method for distributing a logistics vehicle in the distributing apparatus according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. However, the present invention may be embodied in several different forms and is not limited to the embodiments described herein. And in order to clearly explain the present invention in the drawings, parts irrelevant to the description are omitted, and similar reference numerals are attached to similar parts throughout the specification.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . Also, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless otherwise stated, and one or more other features However, it is to be understood that the existence or addition of numbers, steps, operations, components, parts, or combinations thereof is not precluded in advance.

In this specification, a "part" includes a unit realized by hardware, a unit realized by software, and a unit realized using both. In addition, one unit may be implemented using two or more hardware, and two or more units may be implemented by one hardware.

Some of the operations or functions described as being performed by the terminal or device in this specification may be instead performed by a server connected to the terminal or device. Similarly, some of the operations or functions described as being performed by the server may also be performed in a terminal or device connected to the corresponding server.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram of an apparatus for performing a vehicle dispatch according to an embodiment of the present invention. Referring to FIG. 1 , the dispatch execution apparatus 100 includes a GBP model 110 including an order information generating unit 120 , a storage unit 130 , and a simulation unit 140 , a dispatch time determining unit 150 , and a dispatcher. An indication unit 160 may be included.

The GBP (Generator-Buffer-Processor) model 110 is a representative DEVS (Discrete Event System Specification) modeling structure, and in general, the work of generating a workload (Generator), waiting in the queue (Buffer), and processing (Processor) It means a discrete event model that is modeled as a flow. Here, the generation of the workload (Generator) is performed by the order information generating unit 120 , the waiting (Buffer) is performed by the storage unit 130 , and the business processing (Processor) is performed by the simulation unit 140 . can be performed. The order information generating unit 120 , the storage unit 130 , and the simulation unit 140 may function as an environment of the reinforcement learning model.

The GBP model 110 may perform event-based simulation using an external transition function, an internal transition function, an output function, and a time progress function.

According to the present invention, as an environment of the reinforcement learning model, parameters (eg, state information, rewards, etc.) reflecting the field situation using the GBP model 110 that can be updated together with the reinforcement learning model are set to the agent ( Agent), it is possible to improve the learning speed of the reinforcement learning model and the performance of prediction of dispatch timing.

The order information generator 120 may generate virtual order information based on past transportation data. A process of generating the virtual order information will be described in detail with reference to FIG. 2 .

2 is an exemplary diagram for explaining a process of generating virtual order information according to an embodiment of the present invention. Referring to FIG. 2 , the order information generating unit 120 uses a transportation route 201 including loading and unloading locations of a logistics vehicle and past transportation data including order information 202 to provide spatiotemporal characteristics of the past transportation data. The reflected virtual order information 220 may be generated.

The order information generating unit 120 may generate virtual order information similar to past transportation data using a Poisson distribution that can approximate an event that occurs on average L times during a unit time.

For example, since the order information generating unit 120 has a different frequency of occurrence of orders depending on time and day of the week, the time and day of the week through the probability mass function of the non-homogeneous Poisson model that follows a different parameter λ according to the time and day of the week. You can calculate the order frequency according to

Here, the inhomogeneous Poisson model is a parametric statistical model having a constant trend because the error rate is not constant, unlike the homogeneous Poisson process having a constant error rate, and can be expressed by Equation 1. In this case, λ denotes an average of occurrences of all quantities, and Equation 1 indicates a probability value of an event occurring at a specific time and day of the week with respect to an integer x greater than or equal to 0.

For example, the order information generating unit 120 may calculate the order frequency according to the space through the probability mass function of the spatial Poisson model that follows a parameter λ, which is different according to the space.

This is because, when at least one of the loading and unloading stations is not constant, the frequency of occurrence in space as well as time must be modeled. Specifically, when at least one of the loading and unloading locations is not fixed, the region of interest may be divided into a plurality of grids having a predetermined size, and the probability of occurrence of an order in each grid may be approximated using a spatial Poisson model. This may mean a probabilistic process capable of sampling the frequency of occurrence of an event according to the ratio of the entire area to the area corresponding to each grid.

Here, the probability mass function of the spatial Poisson model can be expressed as Equation (2). In this case, λ denotes an average of occurrences of all quantities, and Equation 2 indicates a probability value of an event occurring in a region corresponding to a specific grid with respect to an integer x greater than or equal to 0.

The order information generating unit 120 is a virtual order based on the order frequency according to time and day calculated through the probability mass function of the non-homogeneous Poisson model and the order frequency according to the space calculated through the probability mass function of the spatial Poisson model. information can be generated.

In this way, the order information generating unit 120 may provide the advantage of being able to secure transportation data similar to the actual situation by generating virtual order information based on the transportation distribution that reflects the past transportation data in space and time.

Returning to FIG. 1 again, the storage unit 130 may store virtual order information and vehicle information of a logistics vehicle in a buffer. A process of storing virtual order information and vehicle information of a logistics vehicle in the buffer will be described in detail with reference to FIG. 3 .

3 is an exemplary view for explaining a process of storing virtual order information and vehicle information of a logistics vehicle in a buffer according to an embodiment of the present invention. Referring to FIG. 3 , the storage unit 130 may store virtual order information 301 and vehicle information 302 of a logistics vehicle currently operable in the buffer 300 .

In addition, the buffer 300 may manage the virtual order information updated by the update unit 143 and the vehicle information whose status is changed. For example, the storage unit 130 may manage vehicle information including location information, loading rate, and business hours information of the updated logistics vehicle required for automatic dispatch.

The simulation unit 140 may simulate a dispatch time of the logistics vehicle. The simulation unit 140 may include an extraction unit 141 , a determination unit 142 , and an update unit 143 . Here, the determiner 142 may function as an agent of the reinforcement learning model.

State information including order information and vehicle information is given to the agent, and the agent may act on whether the logistics vehicle is dispatched based on the state information to satisfy a preset reward function.

The agent may determine a policy that satisfies the reward function and act based on the determined policy. A policy means, for example, specifying for all states and actions the probability of taking an action based on given state information. Here, the agent may select an action to maximize the reward function, and the preset reward function is, for example, 'total cost/number of available logistics vehicles', 'total transport distance/total transport time', 'top/ It may be derived based on the 'waiting time at the drop-off point'.

The simulation unit 140 may learn the reinforcement learning model through simulation. Reinforcement learning model means a model that learns which action is best to take in the current state, learns through trial and error, and reduces trial and error through accumulated experience. Here, the simulation unit 140 functions as a part of the environment of the reinforcement learning model, so that when the agent selects an action, a reward is given to the agent so that the agent learns in a direction in which the reward is maximized.

The extractor 141 may extract the first characteristic information by inputting the vehicle information into the first learning model, and may extract the second characteristic information by inputting the order information into the second learning model.

The determination unit 142 may determine whether to dispatch the distribution vehicle based on the virtual order information and vehicle information. For example, the determiner 142 may determine whether to dispatch a distribution vehicle based on order information and vehicle information stored in a buffer using an external state transition function.

The determination unit 142 may determine any one of dispatch or standby of the distribution vehicle based on the first characteristic information and the second characteristic information. A process of determining whether to dispatch a logistics vehicle will be described in detail with reference to FIG. 4 .

4 is an exemplary view for explaining a process of determining whether to dispatch a logistics vehicle based on order information and vehicle information according to an embodiment of the present invention. Referring to Figure 4, the extraction unit 141 is a first learning model CNN (CNN) ( 400, the first characteristic information 401 may be extracted by input to the Convolution Neural Network) model. In addition, the extraction unit 141 is a second learning model RCNN (Recurrent Convolution Neural Network, 410) as a continuous matrix of the remaining order information, unloading site information, and quantity information of the remaining orders according to the passage of time constituting the order information. The second characteristic information 411 may be extracted by inputting it into the model.

The determiner 142 may finally perform classification using an MLP (Multilayer Perceptron) layer that merges the first characteristic information 401 and the second characteristic information 411 . Here, the MLP layer is a neural network that solves the classification problem, and the decision unit 142 can approximate the agent's value function or policy to make a decision on whether to dispatch a logistics vehicle 422 . .

For example, when the extracted final characteristic information is '1', the determiner 142 may determine the dispatch of the distribution vehicle. As another example, when the extracted final feature information is '0', the determiner 142 may determine the waiting of the logistics vehicle.

That is, the simulation unit 140 corresponding to the processor of the GBP model 110 may determine whether to dispatch the distribution vehicle based on the execution value determined by the determination unit 142 .

Returning to FIG. 1 again, the update unit 143 dispatches a logistics vehicle through a preset automatic dispatch model based on whether the logistics vehicle is dispatched, and updates the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle. have.

The updater 143 may calculate a transportation schedule based on the dispatch result of the distribution vehicle using the internal state transition function and update the location and state of the distribution vehicle as vehicle information based on the transportation schedule.

For example, the updater 143 may update the state of the logistics vehicle as vehicle information based on the transportation schedule, and the state related to the vehicle information stored in the buffer based on the location, current loading rate, surplus business hours, and the like. Here, the updater 143 may manage the moving time of the vehicle based on the order information using the time progress function.

As another example, the update unit 143 may update the status related to the order information stored in the buffer in the order of waiting → distributing → loading complete → in transit → disembarking complete (end) as order information based on the transportation schedule. have.

Through this process, the updater 143 may transfer the status of the updated vehicle information and order information related to time, space, available information, etc. of the logistics vehicle to the buffer by using the output function.

The dispatch time determining unit 150 may determine the dispatch time of the logistics vehicle with respect to the real-time order information based on the simulation result. For example, the dispatch timing determiner 150 may determine the dispatch timing of the logistics vehicle by using the learned reinforcement learning model.

The dispatch time determining unit 150 does not immediately instruct the dispatch for real-time order information based on the simulation result, and minimizes the number of logistics vehicles within a range that can delay dispatch for real-time order information as much as possible, It is possible to determine the time of dispatch of the logistics vehicle as the time point at which the travel distance and cost are minimized. Accordingly, it is possible to provide an optimal time for dispatching vehicles even in industries that can be mixed or combined, or industries in which loading and unloading are frequently repeated.

The dispatch instruction unit 160 may instruct dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle.

5 is an exemplary view for explaining a process of simulating a dispatch time of a logistics vehicle according to an embodiment of the present invention. Referring to FIG. 5 , the dispatching apparatus 100 may determine whether to dispatch a distribution vehicle by determining virtual order information stored in a buffer and vehicle information of a distribution vehicle using an external state transition function. At this time, if the distribution of the distribution vehicle is not determined, an internal state transition may occur only for vehicle information of the dispatched distribution vehicle. A transportation route 500 that enables delivery to a logistics vehicle may be calculated.

The dispatch execution apparatus 100 may calculate the transportation schedule 510 based on the dispatch result of the distribution vehicle. For example, the dispatching device 100 calculates the dispatch time for 'No. 0 order' for the logistics vehicle 511 of 'Seoul 02cha 1234' as '19:30', and 'No. 1 The dispatch time for 'order' may be calculated as '19:45', and the dispatch time for 'No.2 order' may be calculated as '20:10'.

The dispatch execution apparatus 100 may update 520 the vehicle information and the order information of the logistics vehicle over time based on the transportation schedule calculated using the internal state transition function. For example, the dispatching device 100 updates the state of the logistics vehicle related to the location of the logistics vehicle, the current loading rate, and the surplus business hours, and orders in the order of waiting, dispatching, loading, loading, etc. state can be updated.

The vehicle dispatching apparatus 100 may manage the travel time for each order by using the time progress function.

The dispatch execution device transmits the updated information related to the status of the logistics vehicle and the order status to the buffer using the output function, so that the information related to the status and order status of the logistics vehicle stored in the buffer is continuously updated (530). have.

The dispatch execution apparatus 100 may be executed by a computer program stored in a medium including a sequence of instructions for performing dispatch of a logistics vehicle. When the computer program is executed by the computing device, the order information generating unit 120 generates virtual order information based on past transportation data, and the storage unit 130 generates virtual order information and vehicle information of the logistics vehicle. is stored in the buffer, the decision unit 142 of the simulation unit 140 determines whether to dispatch a distribution vehicle based on the virtual order information and vehicle information, and the update unit 143 of the simulation unit 140 determines whether Distributes the logistics vehicle through a preset automatic dispatch model based on whether the logistics vehicle is dispatched by the system, and simulates the dispatch time of the logistics vehicle to update the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle, and determines the dispatch time Based on the simulation result by the unit 150 to determine the dispatch time of the distribution vehicle for real-time order information, and instruct the dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle by the dispatch instruction unit 160, and order The information generation unit 120, the storage unit 130, and the simulation unit 140 function as an environment of the reinforcement learning model, and the determination unit 142 may include a sequence of commands to function as an agent of the reinforcement learning model. have.

6 is a flowchart of a method for distributing a logistics vehicle in the distributing apparatus according to an embodiment of the present invention. The method for distributing a logistics vehicle in the dispatching apparatus 100 includes steps processed in time series according to the embodiments shown in FIGS. 1 to 5 . Therefore, even if omitted below, it is also applied to the method of distributing the distribution vehicle in the distributing apparatus 100 according to the embodiment shown in FIGS. 1 to 5 .

In step S610 , the vehicle dispatching apparatus 100 may generate virtual order information based on past transportation data by the order information generating unit 120 .

In step S620 , the vehicle dispatching apparatus 100 may store virtual order information and vehicle information of the logistics vehicle in a buffer by the storage unit 130 .

In step S630 , the dispatching device 100 determines whether to dispatch a distribution vehicle based on the virtual order information and vehicle information by the determining unit 142 of the simulation unit 140 , and the update unit of the simulation unit 140 . By 143, a distribution vehicle is dispatched through a preset automatic dispatch model based on whether the logistics vehicle is dispatched, and the dispatch time of the logistics vehicle is simulated to update the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle. can Here, the simulation unit 140 may function as an environment of the reinforcement learning model, and the determination unit 142 may function as an agent of the reinforcement learning model.

In operation S640, the apparatus 100 for performing the dispatch may determine a dispatch time of the logistics vehicle with respect to the real-time order information based on the simulation result.

In step S650 , the apparatus 100 for performing dispatching may instruct dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle.

In the above description, steps S610 to S650 may be further divided into additional steps or combined into fewer steps, according to an embodiment of the present invention. In addition, some steps may be omitted as necessary, and the order between the steps may be switched.

The method of performing the dispatch of the logistics vehicle in the dispatching performing apparatus described through FIGS. 1 to 6 may also be implemented in the form of a recording medium including a computer program stored in a medium executed by a computer or instructions executable by a computer. can In addition, the method for distributing a logistics vehicle in the distributing apparatus described with reference to FIGS. 1 to 6 may be implemented in the form of a computer program stored in a medium executed by a computer.

Computer-readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media, removable and non-removable media. Also, computer-readable media may include computer storage media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data.

The foregoing description of the present invention is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a distributed manner, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

100: dispatch execution device
120: order information generation unit
130: storage
140: simulation unit
141: extraction unit
142: decision part
143: update
150: dispatch time decision unit
160: dispatch instruction unit

Claims (15)

In an apparatus for distributing a logistics vehicle,
an order information generation unit for generating virtual order information based on past transportation data;
a storage unit for storing the virtual order information and vehicle information of the logistics vehicle in a buffer;
Distributing the distribution vehicle through a predetermined automatic distribution model based on a determination unit that determines whether the distribution vehicle is dispatched based on the virtual order information and the vehicle information and whether the distribution vehicle is dispatched, the distribution vehicle a simulation unit for simulating a dispatch time of the distribution vehicle, including an update unit for updating vehicle information stored in the buffer based on a dispatch result of
a dispatch time determining unit configured to determine a dispatch time of the logistics vehicle with respect to real-time order information based on the simulation result; and
A dispatch instructing unit for instructing the dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle
including,
The order information generating unit, the storage unit, and the simulation unit will function as an environment of the reinforcement learning model, and the determining unit will function as an agent of the reinforcement learning model.
The method of claim 1,
State information including the order information and the vehicle information is given to the agent, and the agent acts on whether the logistics vehicle is dispatched based on the state information to satisfy a preset reward function (Action), the dispatching device.
The method of claim 1,
The update unit calculates a transportation schedule based on a dispatch result of the distribution vehicle, and updates the location and state of the distribution vehicle as the vehicle information based on the transportation schedule.
The method of claim 1,
The simulation unit learns the reinforcement learning model through the simulation,
The dispatch time determining unit is to determine the dispatch time of the distribution vehicle by using the learned reinforcement learning model, dispatch performing apparatus.
5. The method of claim 4,
The simulation unit extracts first characteristic information by inputting the vehicle information into a first learning model,
Further comprising an extraction unit for extracting second characteristic information by inputting the order information into a second learning model, the dispatching apparatus.
6. The method of claim 5,
The vehicle information includes at least one of location information, loading information, and business hours information of the logistics vehicle,
The order information will include at least one of loading information, unloading location information, and quantity information of the remaining orders according to the passage of time.
6. The method of claim 5,
Wherein the determining unit determines any one of dispatch or standby of the distribution vehicle based on the first characteristic information and the second characteristic information.
In the method of performing the dispatch of a logistics vehicle in the dispatch performing apparatus,
generating virtual order information based on past transportation data by an order information generating unit;
storing the virtual order information and vehicle information of a logistics vehicle in a buffer by a storage unit;
Determine whether the distribution vehicle is dispatched based on the virtual order information and the vehicle information by the determining unit of the simulation unit, and a preset automatic distribution model based on whether the distribution vehicle is dispatched by the update unit of the simulation unit distributing the logistics vehicle through, and simulating the dispatch time of the logistics vehicle to update the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle;
determining a dispatch time of the logistics vehicle with respect to real-time order information based on a result of the simulation; and
Instructing the dispatch of the logistics vehicle based on the dispatch time of the logistics vehicle
including,
The order information generating unit, the storage unit and the simulation unit function as an environment of the reinforcement learning model, and the determining unit functions as an agent of the reinforcement learning model.
9. The method of claim 8,
State information including the order information and the vehicle information is given to the agent, and the agent acts on whether the logistics vehicle is dispatched based on the state information to satisfy a preset reward function , how to perform dispatch.
10. The method of claim 9,
Updating the vehicle information stored in the buffer comprises:
Calculating a transportation schedule based on a dispatch result of the distribution vehicle, and updating the location and state of the distribution vehicle as the vehicle information based on the transportation schedule.
10. The method of claim 9,
learning the reinforcement learning model through the simulation; and
The method further comprising the step of determining a dispatch time of the logistics vehicle by using the learned reinforcement learning model.
12. The method of claim 11,
extracting first characteristic information by inputting the vehicle information into a first learning model;
The method further comprising the step of inputting the order information into a second learning model to extract second characteristic information.
13. The method of claim 12,
The vehicle information includes at least one of location information, loading information, and business hours information of the logistics vehicle,
Wherein the order information includes at least one of loading information, unloading location information, and quantity information of the remaining orders according to the passage of time.
13. The method of claim 12,
The step of determining whether to dispatch the logistics vehicle,
How to determine any one of dispatch or standby of the distribution vehicle based on the first characteristic information and the second characteristic information.
A computer program stored in a medium including a sequence of instructions for performing dispatch of a logistics vehicle, the computer program comprising:
When the computer program is executed by a computing device,
Generates virtual order information based on the past transportation data by the order information generation unit,
Storing the virtual order information and vehicle information of the logistics vehicle in a buffer by the storage unit,
Determine whether the distribution vehicle is dispatched based on the virtual order information and the vehicle information by the determining unit of the simulation unit, and a preset automatic distribution model based on whether the distribution vehicle is dispatched by the update unit of the simulation unit dispatching the logistics vehicle through, and simulating the dispatch time of the logistics vehicle to update the vehicle information stored in the buffer based on the dispatch result of the logistics vehicle,
Determine the dispatch time of the logistics vehicle with respect to the real-time order information based on the result of the simulation by the dispatch time determining unit,
Instructing the dispatch of the distribution vehicle based on the dispatch time of the distribution vehicle by the dispatch instruction unit,
The order information generating unit, the storage unit, and the simulation unit function as an environment of the reinforcement learning model, the determining unit comprising a sequence of instructions to function as an agent of the reinforcement learning model, the computer program stored in the medium.

Images