KR102624451B1 - Server, method and computer program for providing route information for logistics - Google Patents

Abstract

The server that provides route information for logistics includes a receiving unit that receives transportation data from a terminal of a transportation company; a transport route derivation unit that inputs the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes; a matrix generator that generates at least one three-dimensional matrix for each time period based on transition type information; a simulation unit that simulates a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and a transport route change unit that changes the first transport route to a second transport route based on the results of the simulation, wherein the at least one three-dimensional matrix for each time period includes at least one unit route among the plurality of unit routes. Includes. The matrix generator may update the three-dimensional matrix for each time period based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans.

Description

Server, method, and computer program for providing route information for logistics {SERVER, METHOD AND COMPUTER PROGRAM FOR PROVIDING ROUTE INFORMATION FOR LOGISTICS}

The present invention relates to a server, method, and computer program that provides route information for logistics.

Logistics refers comprehensively to the act of moving goods or services from producers to consumers or specific regions, or to inventory plans, delivery plans, and procurement plans for this purpose. Logistics services create clusters of transportation points at the district/dong level in administrative districts for ease of transportation. In the conventional case, the person in charge of the logistics company directly set the number of delivery points and density within the cluster, so it was dependent on human resources-based know-how.

Recently, logistics transportation plans are being established in real time using algorithms, and in this regard, Korean Patent Publication No. 2010-0062848, a prior art, discloses a logistics operation method and system.

In the past, static data models were mainly used to establish logistics transportation plans. Hereinafter, the process of providing route information using a logistics transportation plan established based on a conventional static data model will be described with reference to FIG. 1.

1 is an exemplary diagram for explaining a process of providing route information for conventional logistics. The conventional route information server calculates the travel time between each point in the form of a two-dimensional matrix in order to plan the transportation route from the center to the depot through each point while the logistics vehicle is loading goods at each point. It was created with .

Afterwards, the conventional route providing server applied a meta heuristic algorithm to the matrix to search the route in a way that minimized the total travel time within the load limit of each logistics vehicle. At this time, the vehicle path problem was solved by ensuring that the sum of the matrix element values on the path was minimized. A metaheuristic algorithm is a high-level heuristic technique that is not limited to the information of a specific problem and can be applied to a variety of problems.

Referring to FIG. 1, in the first step 100, the conventional route information providing server generates a two-dimensional matrix 101 of the travel times between the center, depot, and point. Here, when moving from the first point (0) to the second point (5) as 0→5, the two-dimensional matrix value corresponding to 0→5 is composed of (0, 5).

In the second step 110, the conventional route information providing server searched for a local solution 111 for each logistics vehicle using a metaheuristic algorithm based on travel time. At this time, local solutions were explored excluding situations where the center was the destination or the garage was the departure point.

If the conventional route information providing server selects the 0 → 5 route for the red logistics vehicle (112), the conventional route information providing server temporarily excludes point 5 from the destination and then selects point 5 as the departure point. So, we searched for the next route within the highlighted section.

Afterwards, when the conventional route information providing server selects the 5 → 6 route for the red logistics vehicle (113), based on 'cumulative load of recovered goods: 60% and cumulative travel time: 4 hours' for the red logistics vehicle , As the load reached the limit, even if more could be transported, no more could be loaded, so route 6 → 8 (garage) was selected as the next route.

Through this process, the conventional route information server compares the conditions by accumulating whether the logistics vehicle can be loaded at each transportation point. If the conventional route information providing server selects two-dimensional matrix values in the order of (0, 5), (5, 6), (6, 8) from the matrix in relation to the route of the red logistics vehicle, the conventional route The information provision server checks whether the total sum of the two-dimensional matrix values is '4+2+2=8', which minimizes the travel time, and when the travel time becomes minimum, changes the transportation route of the red logistics vehicle from 0→5→ It was derived in the order of 6 → 8.

In the third step 120, the conventional route information providing server fills in the cells not highlighted in the second step 110, and in the case of the red logistics vehicle, derives a route of 0 → 5 → 6 → 8, and for the yellow logistics vehicle For logistics vehicles, a path of 0→1→4→5→7 is derived, for blue logistics vehicles, a path of 0→3→9 is derived, and for purple logistics vehicles, a path of 0→2→10 is derived. By deriving, the final local solution (121) was obtained with the logistics transportation cost, which is the total travel time between points, being '8+11+5+12=36'.

However, the conventional logistics transport plan has the disadvantage of not reflecting real-time elements, resulting in a somewhat unrealistic logistics transport plan that does not fully consider realistic transport constraints.

Receiving transport data from a first terminal, inputting the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes, each of the plurality of units based on at least one transition type information The present invention seeks to provide a server, method, and computer program for generating at least one three-dimensional matrix for each time period including at least one unit path among the unit paths.

The purpose is to provide a server, method, and computer program that simulates a route change of a first transportation route by combining the values of at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone. .

It is intended to provide a server, method, and computer program for changing a first transport route to a second transport route based on the results of simulation.

However, the technical challenges that this embodiment aims to achieve are not limited to the technical challenges described above, and other technical challenges may exist.

As a means for achieving the above-described technical problem, an embodiment of the present invention includes a receiving unit that receives transportation data from a terminal of a transportation company; a transport route derivation unit that inputs the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes; a matrix generator that generates at least one three-dimensional matrix for each time period based on transition type information; a simulation unit that simulates a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and a transport route change unit that changes the first transport route to a second transport route based on the results of the simulation, wherein the at least one three-dimensional matrix for each time period includes at least one unit route among the plurality of unit routes. The matrix generator may provide a server that updates the three-dimensional matrix for each time period based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans.

Another embodiment of the present invention includes a receiving unit that receives transportation data from a terminal of a transportation company; a transport route derivation unit that inputs the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes; a matrix generator that generates at least one three-dimensional matrix for each time period based on transition type information; a simulation unit that simulates a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and a transport route change unit that changes the first transport route to a second transport route based on the results of the simulation, wherein the at least one three-dimensional matrix for each time period includes at least one unit route among the plurality of unit routes. The transition type information further includes at least one of logistics characteristic information, driver characteristic information, and transportation point characteristic information, and the matrix generator is based on at least one of the logistics characteristic information, driver characteristic information, and transportation point characteristic information. The at least one additional 3D matrix for each time slot is further generated, and the simulation unit inputs the actual transportation results of the logistics vehicle and the at least one additional 3D matrix for each time slot into an unloading time prediction model to determine unloading of the logistics vehicle. A server that predicts time can be provided.

Another embodiment of the present invention includes receiving transportation data from a terminal of a transportation company; Inputting the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes; generating at least one three-dimensional matrix for each time period based on transition type information; Simulating a route change of the first transportation route by combining values of the at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and changing the first transport route to a second transport route based on the results of the simulation, wherein the at least one three-dimensional matrix for each time period includes at least one unit route among the plurality of unit routes. , The step of generating the 3D matrix for each time slot may include updating the at least one 3D matrix for each time slot based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans. You can.

Another embodiment of the present invention includes receiving transportation data from a terminal of a transportation company; Inputting the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes; generating at least one three-dimensional matrix for each time period based on transition type information; Simulating a route change of the first transportation route by combining values of the at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and changing the first transport route to a second transport route based on the results of the simulation, wherein the at least one three-dimensional matrix for each time period includes at least one unit route among the plurality of unit routes. , the transition type information further includes at least one of logistics characteristic information, driver characteristic information, and transportation point characteristic information, and the step of generating the three-dimensional matrix for each time period includes the logistics characteristic information, driver characteristic information, and transportation point characteristics. A step of further generating the at least one additional 3D matrix for each time slot based on at least one of the information, wherein the simulating step includes an actual transportation result of the logistics vehicle and the at least one additional 3D matrix for each time slot. It may include the step of predicting the unloading time of the logistics vehicle by inputting into an unloading time prediction model.

In another embodiment of the present invention, the computer program, when executed by a computing device, receives transportation data from a terminal of a transportation company, inputs the received transportation data into a transportation dispatch model, and includes a plurality of unit routes. 1 Deriving a transportation route, generating at least one 3-dimensional matrix for each time zone based on transition type information, and generating a 3-dimensional matrix for at least one time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone A sequence of instructions for simulating a route change of the first transport route by combining the values of the matrix and changing the first transport route to a second transport route based on a result of the simulation, and 3 for each time slot. The sequence of instructions for generating a dimensional matrix includes a sequence of instructions for updating the three-dimensional matrix for at least one time period based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans. It is possible to provide a computer program stored on a computer-readable recording medium that includes a computer program. The at least one 3D matrix for each time period may include at least one unit path among the plurality of unit paths.

The above-described means for solving the problem are merely illustrative 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 of the invention.

According to one of the above-described means for solving the problem of the present invention, transport data is input into a transport dispatch model to derive a first transport route including a plurality of unit routes, and each of the above is based on at least one transition type information. A server that establishes a logistics transportation plan that can be operated in the field in real time based on a transportation route that takes time zones into account by generating at least one 3D matrix for each time zone that includes at least one unit path among a plurality of unit paths; Methods and computer programs can be provided.

Simulate a route change in the first transportation route by combining the values of at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies the objective function for the logistics transportation time zone, and simulate the route change of the first transportation route based on the results of the simulation. By changing the route to a second transport route, a server, method, and computer program can be provided to derive an optimal transport route by reflecting various constraints, geographical factors, traffic conditions, etc.

We provide servers, methods, and computer programs that can minimize unfair issues and risks of transportation accidents that may occur in the field by reducing total logistics transportation costs and using a transportation plan in which the work intensity of dispatched logistics vehicles is evenly established. can be provided.

Based on the actual transportation results of the logistics vehicle and the transportation difference information between past transportation plans, a 3-dimensional matrix for at least one time period is updated, and self-learning is performed based on the transportation difference information, allowing a sophisticated transportation plan to be established. Servers, methods, and computer programs may be provided.

1 is an exemplary diagram for explaining a process of providing route information for conventional logistics.
Figure 2 is a configuration diagram of a route information providing server according to an embodiment of the present invention.
3A to 3D are exemplary diagrams for explaining a process of performing a simulation of a route change using at least one 3D matrix for each time period according to an embodiment of the present invention.
FIG. 4 is an exemplary diagram illustrating a process for generating at least one 3D matrix for each time period by clustering a plurality of unit paths according to an embodiment of the present invention.
FIG. 5 is an exemplary diagram illustrating a process of performing simulation for a path change using at least one sparse matrix according to an embodiment of the present invention.
FIG. 6 is an exemplary diagram illustrating a process of performing a simulation of a route change based on a second scenario in consideration of transportation labor intensity according to an embodiment of the present invention.
FIGS. 7A and 7B are exemplary diagrams for explaining a process of predicting the unloading time of a logistics vehicle based on the actual transportation result of the logistics vehicle and transportation difference information between the past transportation plan according to an embodiment of the present invention.
8A to 8C are exemplary diagrams for explaining a process of deriving a second transportation route based on a first scenario and a second scenario according to an embodiment of the present invention.
Figure 9 is a flow chart of a method of providing route information for logistics in a route information providing server according to an embodiment of the present invention.

Below, with reference to the attached drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily implement the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. In order to clearly explain the present invention in the drawings, parts that are not related to the description are omitted, and similar parts are given similar reference numerals throughout the specification.

Throughout the specification, when a part is said to be "connected" to another part, this includes not only the case where it is "directly connected," but also the case where it is "electrically connected" with another element in between. . In addition, when a part is said to "include" a certain component, this does not mean excluding other components unless specifically stated to the contrary, but may further include other components, and one or more other features. It should be understood that it does not exclude in advance the presence or addition of numbers, steps, operations, components, parts, or combinations thereof.

In this specification, 'part' includes a unit realized by hardware, a unit realized by software, and a unit realized using both. Additionally, one unit may be realized using two or more pieces of hardware, and two or more units may be realized using one piece of hardware.

In this specification, some of the operations or functions described as being performed by a terminal or device may instead be performed on a server connected to the terminal or device. Likewise, some of the operations or functions described as being performed by the server may also be performed on a terminal or device connected to the server.

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

Figure 2 is a configuration diagram of a route information providing server according to an embodiment of the present invention. Referring to FIG. 2, it may include a receiving unit 210, a transport route deriving unit 220, a clustering performing unit 230, a matrix generating unit 240, a simulation unit 250, and a transport route changing unit 260. there is.

The receiving unit 210 may receive transportation data from the first terminal. Here, the first terminal may be a terminal that inputs transportation data from a logistics company. Transportation data may include transportation resource information and transportation order information. Transportation resource information includes, for example, origin (center), depot (destination), logistics vehicle information, transportation point, logistics type (type and weight of goods, etc.), and transportation order information includes, for example, daily order. Information, weekly order information, etc.

The transportation route derivation unit 220 may input the received transportation data into a transportation dispatch model to derive a first transportation path including a plurality of unit paths as a path for the logistics vehicle to be transported. Here, the transportation dispatch model can be created using local search techniques, genetic algorithms, reinforcement learning, etc., but is not limited to this. The local search technique may refer to a search technique that selects and expands from the current point (node) to the point with the highest cost savings among expandable neighboring points. A unit route may refer to each route connecting a plurality of points existing within one area for transporting logistics.

The matrix generator 240 may generate at least one 3D matrix for each time period, each including at least one unit path among a plurality of unit paths, based on at least one transition type information. Here, the transition type information may include the travel distance between each point on the first transportation route, the travel time between each point, the loading weight of the logistics vehicle, logistics characteristic information, driver characteristic information, and transportation point characteristic information.

The simulation unit 250 may simulate a route change of the first transportation route by combining the values of at least one 3D matrix for each time zone based on the first transportation scenario that satisfies the objective function for the logistics transportation time zone. The process of generating and simulating at least one 3D matrix for each time period will be described in detail with reference to FIGS. 3A to 3C.

3A to 3C are exemplary diagrams for explaining a process of performing a simulation of a route change using at least one 3D matrix for each time period according to an embodiment of the present invention.

Referring to FIG. 3A, the matrix generator 240 generates at least one of the travel distance 311 between each point on the first transportation route 300, the travel time between each point 312, and the loading weight 313 of the logistics vehicle. Based on this, at least one 3D matrix 310 for each time period can be generated. Here, the situation where the center is the destination or departure point may be excluded, and the time zone may include morning time, afternoon time, lunch time, rush hour, dawn time, etc. The coordinates of the three-dimensional matrix 310 for each time zone are composed of, for example, (origin_point (node) dimension, destination_point (node) dimension, time dimension, transition type (e.g., travel distance, etc.)) The order of coordinates may be changed and is not limited to this. For example, when moving from the first transportation point (0) to the second transportation point (5) as 0 → 5, the matrix value corresponding to the first time zone, the rush hour, and the first travel time is, for example, ( It can consist of values 0, 5, 1, 1).

Through this process, the correlation between points and changes in transportation resources over time can be organized into matrix coordinates, allowing efficient calculations and dynamic programming to be performed when searching for the optimal route.

Referring to FIG. 3B, the objective function 320 for the logistics transportation time zone may be designed as the logistics transportation cost (x, y, z). Here, x may represent the travel distance, y may represent the travel time, and z may represent the loading and unloading work time.

For example, in the case of normal morning and afternoon times, the objective function 320 can be designed as logistics transportation cost (x, y), through which the travel distance and travel time can be optimized simultaneously. This is to take into account that the difference in traffic conditions between morning and afternoon hours is low, and that there are few unusual events during loading and unloading due to universal business hours.

For another example, in the case of commuting and leaving hours, the objective function 320 may be designed to correspond to logistics transportation costs (y, z), through which travel time and work time can be optimized simultaneously. This is to take into account the fact that, in the case of commuting and finishing hours, traffic congestion is severe, time volatility is large, and many unusual events occur from the start to the end of work and when loading and unloading.

For another example, in the early morning hours, the objective function 320 can be designed as logistics transportation costs (x, z), and travel distance and work time can be optimized simultaneously. This is to minimize distances to reduce idle costs as traffic conditions are smooth in the early morning hours, and to take into account the fact that many unusual events occur during loading and unloading.

For example, if the optimal logistics transport schedule corresponds to 'Center: Arrival at 5 a.m.' → 'Point 1: Arrival at 8 a.m.' → 'Point 3: Arrival at 9 a.m.', the simulation unit 250 is ' Total logistics transportation based on the objective function: Dawn logistics transportation cost (x, z): 100'+'Start-up logistics transportation cost (y, z): 500'+'Start-up logistics transportation cost (y, z): 900' Costs can be derived.

Referring to FIG. 3C, the simulation unit 250 combines the values of the three-dimensional matrix 340 for at least one time zone based on the first transportation scenario that satisfies the objective function 320 for the logistics transportation time zone to create a first A route change in the transportation route 330 can be simulated (350).

For example, the simulation unit 250 may simultaneously apply a meta-heuristic algorithm to all three-dimensional matrices 340 for each time period corresponding to transition type information based on the first transportation route 330. The simulation unit 250 can utilize all of the various proven meta-heuristic algorithms to solve vehicle routing problems. Representative meta-heuristic algorithms include Genetic Algorithms, Tabu Search, Simulated Annealing, Ant Algorithm, Savings, The Sweep Algorithm, Includes Cheapest Insertion, Nearest Insertion, etc. The first transportation scenario may be designed to satisfy the objective function so that the logistics transportation cost is minimized by combining the values of the 3-dimensional matrix 340 for each time period related to each transition type information. For example, if the moving time of the logistics vehicle to date is '4 minutes', the moving distance is '4km', and the delivered weight is 600kg, the simulation unit 250 takes this into consideration and creates a three-dimensional matrix 340 for each time period. ) can be combined to perform a simulation 350 for the first transportation route 330.

At this time, if the sum of accumulated times in the travel time matrix is outside the time zone of the time dimension, the simulation unit 250 may perform simulation 350 using the next time dimension. In other words, by setting the objective function differently depending on the time dimension, optimization can be performed to derive the desired transportation route for each time zone.

For example, the cost of logistics transportation during the early morning hours can be designed as a weighted sum of travel distance, travel time, and load deviation. If the simulation unit 250 selects the 0 → 5 path in the three-dimensional matrix for each time zone, point 5 in all time dimensions of each transition type information may be temporarily excluded from the destination, and the simulation unit 250 The following route can be explored within the highlighted section, using point 5 as the starting point.

Afterwards, during the simulation 350, if the cumulative travel time of a specific logistics vehicle is outside the set time zone, the simulation unit 250 continues the search using a tuple, which is an attribute value related to a matrix of another time dimension. You can. Through this, routing close to real-time can be performed.

When the simulation unit 250 selects route 5 → 6 in the three-dimensional matrix for each time zone, the accumulated travel time may change depending on the rush hour, so the simulation unit 250 selects the next route with point 6 as the starting point. , You can search within the highlighted section, but in the next time dimension that includes cumulative time. At this time, the logistics transportation cost during the commuting time zone can be set as a time sum considering the characteristics of the time zone.

Referring to FIG. 3D, the simulation unit 250 performs the first transportation based on the first transportation scenario that minimizes the logistics transportation costs related to the travel distance, travel time, and loading weight for the logistics vehicle in consideration of the logistics transportation time zone. Simulation of route changes can be repeatedly performed.

For example, the simulation unit 250 repeatedly performs the simulation 370 for changing the route of the first transport route 360 based on the first transport scenario, thereby creating a second transport route that satisfies the objective function for each time period. It can be derived. At this time, the more detailed the time dimension is, the more similar it is to real-time travel speed, which prevents the routing process from establishing a plan that cannot realistically be transported within time.

Through this, in the past, only the travel time matrix during early morning hours was used to establish a logistics transportation plan, so logistics vehicles could not move as fast as the logistics transportation plan during rush hour, but a realistic logistics transportation plan was established through the technology of the present invention. You can do it.

Returning to FIG. 2, when the transport route derivation unit 220 inputs the received transport data into the transport distribution model to derive the first transport route to be transported by the logistics vehicle, the clustering performing unit 230 evenly distributes the cargo volume. Clustering can be performed on multiple unit routes based on conditions, loading conditions for each logistics vehicle, preferred area information, etc.

The matrix generator 240 generates at least one 3D matrix for each time zone including a plurality of unit paths each clustered based on at least one transition type information, and generates a 3D matrix for each time zone from the generated at least one 3D matrix for each time zone. At least one sparse matrix can be created.

The simulation unit 250 may perform a simulation of a route change of the first transportation route by combining the values of at least one sparse matrix based on the first scenario. When clustering is performed on the first transport route, the process of creating and simulating a 3D matrix for each time period will be described in detail with reference to FIGS. 4 and 5.

FIG. 4 is an exemplary diagram illustrating a process for generating at least one 3D matrix for each time period by clustering a plurality of unit paths according to an embodiment of the present invention.

Referring to FIG. 4, when the receiver 210 receives transport data from the first terminal, the transport route derivation unit 220 inputs the received transport data into the transport dispatch model to determine the first transport route to be transported by the logistics vehicle. (400) can be derived. At this time, the matrix generator 240 generates at least one 3D matrix 410 for each initial time period including a plurality of unit paths included in the first transportation path 400 based on at least one transition type information. You can save it. For example, in the three-dimensional matrix 410 for each initial time zone, the matrix values of (3, 4, 4, 0) are the travel times for the 3rd departure point (node), the 4th destination point (node), and the 4th evening time zone. and transition type.

The clustering performing unit 230 may perform clustering 420 on a plurality of unit routes included in the first transport route 400 based on conditions for equal distribution of cargo volume, loading conditions for each logistics vehicle, preferred area information, etc. . For example, the clustering unit 230 performs the first clustering (421) based on the conditions for equal distribution of cargo volume, and then performs the second clustering (422) based on the loading conditions for each logistics vehicle and preferred region information. can do.

The matrix generator 240 may generate at least one 3D matrix 430 for each time period including a plurality of unit paths each grouped based on at least one transition type information.

The simulation unit 250 changes the route of the first transportation route 400 by combining the values of the 3D matrix 430 for at least one time zone based on the first transportation scenario that satisfies the objective function for the logistics transportation time zone. It can be simulated.

In this way, the plurality of unit routes included in the first transport route 400 are clustered based on transition type information between each point to generate a 3-dimensional matrix 430 for each time period, so that the clustering described above is not considered. Compared to the initial 3D matrix 410 for each time period, the amount of information is reduced by, for example, 1/3, and the modeling speed is improved by, for example, three times. Through this process, optimal transportation is achieved through an objective function that clusters transportation locations, adheres to preferred areas for each vehicle, considers equal load capacity for each vehicle, and takes into account all inconvenient routes such as rivers, construction, and accident areas. A path can be derived.

FIG. 5 is an exemplary diagram illustrating a process of performing simulation for a path change using at least one sparse matrix according to an embodiment of the present invention.

The clustering performing unit 230 may perform clustering 420 on a plurality of unit routes included in the first transport route 400 based on conditions for equal distribution of cargo volume, loading conditions for each logistics vehicle, preferred area information, etc. . For example, the clustering unit 230 performs the first clustering (421) based on the conditions for equal distribution of cargo volume, and then performs the second clustering (422) based on the loading conditions for each logistics vehicle and preferred region information. can do.

At this time, in the case of the first clustering (421) and the second clustering (422), clustering does not take into account the transportation time of the logistics vehicle, so when the logistics vehicle performs actual logistics transportation, the logistics transportation is not completed within business hours. Problems may arise that may not be possible.

Accordingly, the clustering performing unit 230 may re-perform clustering 500 based on the degree of proximity between clusters for a plurality of unit paths on which clustering has been performed. For example, the clustering unit 230 creates at least one cluster 501 (e.g., a first cluster located in the upper left corner and a second cluster located in the upper right corner) for a plurality of unit paths on which the secondary clustering 422 has been performed. ), the final cluster (502) can be derived by re-performing the clustering. The final cluster 502 may be clustered based on the degree of contiguity between clusters, so that at least two or more lower clusters are clustered into upper clusters. At this time, when the clustering unit 230 divides the space by connecting points on the plane into triangles, the clustering unit 230 uses Delaunay triangulation, which is a division that maximizes the minimum value of the interior angles of these triangles, based on the degree of proximity between clusters. By expanding the colony, it is possible to prevent abnormal colonies from occurring.

The matrix generator 240 may generate at least one 3D matrix 510 for each time period, each of which includes a plurality of finally clustered unit paths, based on at least one transition type information.

The matrix generator 240 may generate at least one 3D sparse matrix 520 for each time slot from the generated 3D matrix for each time slot. Here, the sparse matrix may be in the form of a matrix where most of the matrix values are '0'.

The simulation unit 250 simulates the route change of the first transportation route 400 by combining the values of at least one 3D sparse matrix for each time zone based on the first transportation scenario that satisfies the objective function for the logistics transportation time zone. You can. At this time, the simulation unit 250 may search for a path for the vehicle path problem using a metaheuristic algorithm. In the 3D matrix for each time period, all transition type information between points (all points are connected) connected by the gray line expressed at the initial point on the first transportation route 400 can be recorded. Accordingly, the simulation unit 250 can search for a path along each gray line in order to combine them and simulate them.

In the case of the first cluster 421 and the second cluster 422, transition type information corresponding to the gray line within each cluster is recorded, and the simulation unit 250 records the 3D matrix values for each time period generated within each cluster. Path changes can be simulated through a combination of . The final cluster 502 not only records transition type information between valid points like the secondary cluster 422, but also can have an intersection of multiple clusters including the same point at the same time. Accordingly, the simulation unit 250 can search for points connected by gray lines within the final cluster 502 and search for an optimal path by going back and forth between the clusters expressed in green and the clusters expressed in blue.

For example, the simulation unit 250 may combine the values of the 3D sparse matrix for each time period and then use the Cheapset Insertion algorithm to simulate a route change to minimize the travel distance according to the following flow.

Since the travel distance matrix is a three-dimensional matrix in which the coordinate of the last transition type information in the overall sparse matrix is '0', the cost for the initial path connecting the departure node x, for each vehicle, to the destination node y, is the value of the entire sparse matrix. It can correspond to the values (x, y, 0, 0). Here, the cost for the initial path can be defined as C _xy = (x, y, 0, 0). Afterwards, an intermediate point can be searched that minimizes the increase in cost when included in the vehicle's initial route. This can search for nodes that satisfy argmin{C _xi +C _iy -C _xy } for all values on the movement distance matrix. That is, through the process of searching for point i where the cost of the x->i->y path increases the least compared to the cost of the x->y path, the simulation unit 250 determines that when the cost function is a simple travel distance, You can simply check by iterating through the values whose coordinates of the last transition type information of the sparse matrix are 0. However, in this process, since the travel time must be tracked separately from the cost, the simulation unit 250 accumulates and adds the (x, y, 0, 1) values of the sparse matrix, and exceeds the predefined time standard. Each time you do this, you can check the next search by iterating through the values where the time dimension coordinate is 1. In other words, when calculating C _xy in the above equation, this may mean checking the coordinate values of (x, y, 1, 0).

At this time, the simulation unit 250 may simulate a route change considering the vehicle routing problem (VRP) 530 based on the generated 3D sparse matrix 520 for each time period. Here, the vehicle routing problem 530 refers to the problem of deriving a route that can be transported at minimum cost or minimum distance when k vehicles of the same type are located in a logistics warehouse to provide logistics transportation services to n points. do.

The transport route change unit 260 may change the first transport route 400 to the second transport route 540 based on the results of the simulation. For example, the transport route change unit 260 changes the first transport route 400 to the second transport route 540, so that it takes about 160 minutes for a blue logistics vehicle, and for a red logistics vehicle, It can take about 90 minutes, and in the case of a green logistics vehicle, it can take about 120 minutes.

Through this process, it is possible to simulate to explore a more optimized route through a global optimization method that takes into account not only the loading conditions and geographical conditions of the logistics vehicle, but also the vehicle route problem.

Returning to FIG. 2, the simulation unit 250 derives a temporary transportation route based on the first transportation scenario as a result of the simulation, and creates a temporary transportation route based on the second transportation scenario that satisfies the objective function for transportation labor intensity. Simulation of path changes can be performed. The process of performing a simulation for changing the route of a temporary transport route based on a second transport scenario that satisfies the objective function for transport labor intensity will be described in detail with reference to FIG. 6.

FIG. 6 is an exemplary diagram illustrating a process of performing a simulation of a route change based on a second scenario in consideration of transportation labor intensity according to an embodiment of the present invention. Referring to FIG. 6 , the simulation unit 250 may perform a simulation on the first transport route by combining the values of the 3D matrix 600 for at least one time period to derive a temporary transport route 610. At this time, the simulation unit 250 repeats the simulation of the route change of the temporary transport route 610 based on the second transport scenario that minimizes the labor intensity equal cost for a plurality of logistics vehicles in consideration of transport labor intensity. It can be done. According to the present invention, it is possible to derive a second transportation route 620 that can minimize transportation labor intensity while satisfying transportation service conditions by changing the temporary transportation route 610 to minimize the objective function for transportation labor intensity. do. Here, the simulation unit 250 may repeatedly perform the simulation until the set execution time and execution number are reached to derive optimized simulation results. Labor intensity balancing cost is, for example, a value that is minimized as the transportation labor intensity is equally distributed between multiple logistics vehicles, including transportation time deviation, number of transportation points, and maximum transportation work time for multiple logistics vehicles. It can be calculated based on etc.

When performing a simulation, the simulation unit 250 may select the maximum transportation work time among labor intensity equal costs as a simulation parameter. Here, since the model that performs simulation cannot have derivatives, an optimization technique such as the 'Nelder-Mead' technique is used to determine the number of times that one logistics vehicle can transport with simulation parameters that can minimize the labor intensity equal cost. You can select the ‘maximum transportation working time’. Here, the 'Nelder-Mead' technique is a numerical technique to find the minimum or maximum value, and is mainly used in nonlinear optimization problem situations where the derivative is unknown. For example, the simulation unit 250 may repeatedly search for the convergence point of the cost hyperplane using the 'Nelder-Mead' technique.

The simulation unit 250 may repeatedly perform a simulation of the route change of the temporary transport route 610 by changing the value of the selected simulation parameter. For example, the simulation unit 250 performs simulation according to a transportation scenario that allows other logistics vehicles to perform more transportation work by changing the value of the 'maximum transportation work time' that one logistics vehicle can transport to decrease. can do. In this case, the simulation unit 250 may perform simulation with the number of the plurality of logistics vehicles fixed so that the number of the plurality of logistics vehicles does not increase.

The simulation unit 250 may terminate the simulation when the labor intensity equal cost converges and does not change through repeated simulations or when the set execution time and number of execution times are satisfied.

The transport route change unit 260 may change the temporary transport route 610 to the second transport route 620 based on a second transport scenario in which labor intensity and equal costs are minimized as a result of the simulation. Here, as a result of the simulation, improved values in 'transportation time deviation' and 'maximum transportation work time' can be derived compared to the temporary transportation route 610.

Through this process, the second transportation route 620 corresponding to the final route considering transportation labor intensity is derived, thereby optimizing transportation labor intensity while maintaining logistics transportation costs such as the number of logistics vehicles and travel distance. Drivers' transportation labor intensity may vary. At this time, by providing a dynamic and optimized transportation route by taking into account transportation time deviations, etc., it is possible to instruct dispatch considering transportation labor intensity even during vehicle mapping.

For example, in the case of the temporary transportation route 610, the blue logistics vehicle took about 160 minutes, the red logistics vehicle took about 90 minutes, and the green logistics vehicle took about 120 minutes, but the transportation route change unit As (260) changes the temporary transport route 610 to the second transport route 620, the blue logistics vehicle takes about 140 minutes, the red logistics vehicle takes about 90 minutes, and the green logistics vehicle takes about 130 minutes. It may take about a minute.

Returning to FIG. 2 , the matrix generator 240 may update at least one 3D matrix for each time period based on the simulation results, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans. The process of updating the 3D matrix for at least one time period will be described in detail with reference to FIGS. 7A and 7B.

FIGS. 7A and 7B are exemplary diagrams for explaining a process of predicting the unloading time of a logistics vehicle based on the actual transportation result of the logistics vehicle and transportation difference information between the past transportation plan according to an embodiment of the present invention.

Referring to FIG. 7A, the matrix generator 240 creates at least one three-dimensional matrix 740 for each time period based on the transportation difference information between the simulation result, the actual transportation result 700 of the logistics vehicle, and the past transportation plan 710. ) can be updated.

Through this, the accuracy of travel time and routing options can be improved. For example, in the case of travel time, modeling may be performed based on a correction value for the difference in travel time/travel distance between points based on the driver's skill level, point location information, etc. Additionally, in the case of routing options, modeling may be performed based on the tendency to prefer the routing option based on distance information that matches each of the multiple routing option paths.

For example, the simulation unit 250 explores routing options through reverse simulation so that the transport difference information between the actual transport results 700 and the past transport plan 710 is reduced, and the matrix generation unit 240 searches for customer and The travel distance matrix and travel time matrix can be updated by applying routing options for each operator. For example, in the morning time zone, the simulation unit 250 generates an error in the travel time as the travel time increases by about 20% compared to the past transportation plan 710, so a factor that increases the travel time by about 20% Reverse simulation can be performed for . Here, as a result of reverse simulation, it is possible to derive plan 1) inferring that the speed of the logistics vehicle is 20% slower than expected, or plan 2) moving through a route that does not require a U-turn. If it is inferred that the logistics vehicle on the green route has a long driving experience and does not drive at low speeds, the simulation unit 250 may select the second option to search for the optimal route.

For example, the matrix generator 240 regenerates (730) the travel distance matrix (731) and the travel time matrix (732) of the green route logistics vehicles in the morning hours with the U-turn prohibition option based on the second plan, The 3D matrix 740 can be updated for each time period.

The matrix generator 240 may generate at least one 3D matrix for each time period based on logistics characteristic information, driver characteristic information, and transportation point characteristic information.

The simulation unit 250 may predict the unloading time 722 of the logistics vehicle by inputting the actual transportation results of the logistics vehicle and at least one 3D matrix 721 for each time period into the unloading time prediction model 720. Here, the alighting time matrix can be updated by learning through a neural network that predicts the alighting time based on transition type information based on actual transportation data. Through this, the accuracy of the prediction time for getting off can be improved.

Referring to FIG. 7B, the simulation unit 250 may input (751) the customer company's cargo characteristic information, driver characteristics, and point characteristics into the unloading time prediction model 750. Here, the get-off time prediction model 720 may be a prediction model based on deep learning (Deep Neural Network).

Customer cargo characteristics may include, for example, delivery volume, visit day, visit time zone, weather, usage area, peak/off-peak season, etc.

Driver characteristics and branch characteristics include, for example, vehicle capacity, driver skill level, road type, POI-transit point name search address matching, driver visit experience, means of disembarkation, pick-up/security authentication procedures, availability of free parking space, dedicated parking space. This can include presence or absence, travel distance, whether it is on the first floor, etc.

The simulation unit 250 inputs (751) the customer's cargo characteristics, driver characteristics, and point characteristics into the unloading time prediction model (750) and outputs (752) the predicted unloading time. Here, the predicted unloading time may include parking/stopping time, cargo unloading time (after GPS stop is confirmed), and unloading time (after electronic handover is confirmed).

The learning unit (not shown) may self-train an unloading time prediction model periodically based on transportation difference information derived based on past transportation plans and actual transportation results.

This route information providing server 200 may be executed by a computer program stored in a medium containing a sequence of commands that provide route information for logistics. When executed by a computing device, the computer program receives transportation data from a first terminal, inputs the received transportation data into a transportation dispatch model to derive a first transportation route including a plurality of unit paths, and generates at least one transition. Based on the type information, at least one 3-dimensional matrix for each time zone is generated, each of which includes at least one unit path among a plurality of unit paths, and at least based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone. The route change of the first transport route can be simulated by combining the values of a three-dimensional matrix for each time period, and may include a sequence of commands to change the first transport route to the second transport route based on the results of the simulation. .

8A to 8C are exemplary diagrams for explaining a process of deriving a second transportation route based on a first scenario and a second scenario according to an embodiment of the present invention.

Referring to FIG. 8A, the route information providing server 200 may receive transport data from the first terminal and input the received transport data into a transport dispatch model to derive the first transport route through which the logistics vehicle will be transported.

The route information providing server 200 may generate at least one 3D matrix 800 for each time zone based on transition type information between each point of a plurality of unit routes included in the first transport route.

For example, if the transition type information corresponds to cargo volume 801, the coordinates of the cargo volume matrix may be composed of (origin point, arrival point, time dimension, transition type). For example, z-coordinate 4: cargo volume matrix during evening hours, z-coordinate 3: cargo volume matrix during rush hours, z-coordinate 2: cargo volume matrix during afternoon hours, z-coordinate 1: cargo volume matrix during rush hours, z-coordinate 0: morning hours. It may consist of a cargo volume matrix.

For another example, if the transition type information corresponds to travel distance/travel time 802, the coordinates of the travel distance/travel time matrix may be composed of (start point, arrival point, time dimension, transition type). . For example, z-coordinate 4: evening time travel distance/travel time matrix, z-coordinate 3: travel time zone travel distance/travel time matrix, z-coordinate 2: afternoon time travel distance/travel time matrix, z-coordinate 1: work time travel time zone. Distance/travel time matrix, z-coordinate 0: It can be composed of a travel distance/travel time matrix during the early morning hours.

Referring to FIG. 8B, the route information providing server 200 may perform a simulation 820 for a transportation route based on the first scenario.

For example, if the first scenario is designed based on the customer company's 'cargo volume' scenario 821, the route information providing server 200 may perform modeling according to the characteristics of the transportation time zone. For example, Agency A can unload pallets by operating a forklift only during the morning hours, so it can be taken into consideration that pallet unit packaging and volume calculation are required when using cargo pallets. If the pallet is not used, it is possible to increase the efficiency of cargo loading space, but it can be considered that the work time increases due to manual unloading when the pallet is not used. Additionally, the route information providing server 200 may perform modeling according to ‘cargo characteristics’. For example, if agency C's cargo requires careful handling, it may be taken into consideration that it cannot be loaded at the same time as agency A's cargo.

By considering these matters, the route information providing server 200 considers that afternoon transportation and manual unloading are performed at point A in the case of coordinates (A, B, 4), and In the case, considering that morning transport is carried out at point A, and forklift pallet unloading is carried out, for coordinates (A, C, 1) and (A, C, 4), A and C are the same due to the impossibility of loading. Simulation can be performed taking into account that large numbers that cannot be calculated are assigned.

For another example, when the first scenario is designed based on the 'transportation route scenario' 822, the route information providing server 200 may perform modeling according to the traffic situation. This can be taken into account that travel time increases and the optimal route frequently changes during commuting and leaving hours. Additionally, the route information providing server 200 may perform modeling according to route characteristics. This can be taken into account that logistics vehicles cannot make a U-turn on a road below the lane, are located on the same road surface in the order of point A → point B, and that a U-turn or detour route is required from point B → point A.

Through this process, the route information providing server 200 considers that it takes about 1 hour from point A to point B in the case of coordinates (A, B, 4), and determines the location of coordinates (B, A, 4). In this case, the simulation can be performed by considering that a detour route is required because a U-turn is not possible, and that in the case of coordinates (A, B, 1), the time increases due to traffic congestion during rush hour.

Referring to FIG. 8C, as a result of the simulation through this process, the route information providing server 200 provides a total transportation route of 11 hours, with 4 hours in the off-hours time, 2 hours in the afternoon time, and 5 hours in the rush hour, based on the first scenario. A temporary transport route 830 including time can be derived. At this time, since the driver's work hours of 8 hours are exceeded, the route information providing server 200 provides 1 hour in the afternoon time, 2 hours in the afternoon time, and rush hour based on the second scenario that satisfies the driver's 8 hours of work time. The second transport route 840 can be derived as the optimal transport route to satisfy a total of 8 hours (5 hours).

Figure 9 is a flowchart of a method of providing route information for logistics in a route information providing server according to an embodiment of the present invention. The method of providing route information for logistics in the route information providing server 200 shown in FIG. 9 includes steps processed in time series according to the embodiments shown in FIGS. 2 to 8C. Therefore, even if the content is omitted below, it also applies to the method of providing route information for logistics in the route information providing server 200 according to the embodiment shown in FIGS. 2 to 8C.

In step S910, the route information providing server 200 may receive transportation data from the first terminal (not shown).

In step S920, the route information providing server 200 may input the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes.

In step S930, the route information providing server 200 may generate at least one 3D matrix for each time period, each of which includes at least one unit path among a plurality of unit paths, based on at least one transition type information.

In step S940, the route information providing server 200 simulates a route change of the first transport route by combining the values of at least one three-dimensional matrix for each time period based on the first transport scenario that satisfies the objective function for the logistics transport time slot. can do.

In step S950, the route information providing server 200 may change the first transport route to the second transport route based on the results of the simulation.

In the above description, steps S910 to S950 may be further divided into additional steps or combined into fewer steps, depending on the implementation of the present invention. Additionally, some steps may be omitted or the order between steps may be switched as needed.

The method of providing route information for logistics from the route information providing server described through FIGS. 1 to 9 can also be implemented in the form of a computer program stored on a medium executed by a computer or a recording medium containing instructions executable by a computer. It can be. Additionally, the method of providing route information for logistics from the route information providing server described with reference to FIGS. 1 to 9 may also 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 non-volatile media, removable and non-removable media. Additionally, computer-readable media may include computer storage media. Computer storage media includes both volatile and non-volatile, 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 description of the present invention described above is for illustrative purposes, and those skilled in the art will understand that the present invention can be easily modified into other specific forms without changing the technical idea or essential features of the present invention. will be. Therefore, the embodiments described above should be understood in all respects as illustrative and not restrictive. For example, each component described as unitary may be implemented in a distributed manner, and similarly, components described as distributed may also be implemented in a combined form.

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

200: route information providing server
210: receiving unit
220: Transportation route derivation unit
230: Clustering execution unit
240: Matrix generation unit
250: Simulation unit
260: Transportation route change unit

Claims (5)

In the server that provides route information for logistics,
A receiving unit that receives transport data from a transport company terminal;
a transport route derivation unit that inputs the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes;
a matrix generator that generates at least one three-dimensional matrix for each time period based on transition type information;
a simulation unit that simulates a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and
A transport route change unit that changes the first transport route to a second transport route based on the results of the simulation,
The at least one three-dimensional matrix for each time period includes at least one unit path among the plurality of unit paths,
The matrix generator updates the three-dimensional matrix for each time period based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans.
In the server that provides route information for logistics,
A receiving unit that receives transport data from a transport company terminal;
a transport route derivation unit that inputs the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes;
a matrix generator that generates at least one three-dimensional matrix for each time period based on transition type information;
a simulation unit that simulates a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and
A transport route change unit that changes the first transport route to a second transport route based on the results of the simulation,
The at least one three-dimensional matrix for each time period includes at least one unit path among the plurality of unit paths,
The transition type information further includes at least one of logistics characteristic information, driver characteristic information, and transportation point characteristic information,
The matrix generator further generates the at least one additional three-dimensional matrix for each time period based on at least one of the logistics characteristic information, driver characteristic information, and transportation point characteristic information,
The simulation unit predicts the unloading time of the logistics vehicle by inputting the actual transportation results of the logistics vehicle and the additional 3D matrix for each time zone into an unloading time prediction model.
Receiving transport data from a transport company terminal;
Inputting the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes;
generating at least one three-dimensional matrix for each time period based on transition type information;
Simulating a route change of the first transportation route by combining values of the at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and
Changing the first transport route to a second transport route based on the results of the simulation,
The at least one three-dimensional matrix for each time period includes at least one unit path among the plurality of unit paths,
The step of generating a 3D matrix for each time period is,
A route information providing method comprising updating the three-dimensional matrix for at least one time period based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans.
Receiving transport data from a transport company terminal;
Inputting the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes;
generating at least one three-dimensional matrix for each time period based on transition type information;
Simulating a route change of the first transportation route by combining values of the at least one three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone; and
Changing the first transport route to a second transport route based on the results of the simulation,
The at least one three-dimensional matrix for each time period includes at least one unit path among the plurality of unit paths,
The transition type information further includes at least one of logistics characteristic information, driver characteristic information, and transportation point characteristic information,
The step of generating a 3D matrix for each time period is,
Further generating the at least one additional three-dimensional matrix for each time period based on at least one of the logistics characteristic information, driver characteristic information, and transportation point characteristic information,
The simulation step is,
A server comprising the step of predicting the unloading time of the logistics vehicle by inputting the actual transportation results of the logistics vehicle and the at least one additional 3D matrix for each time period into an unloading time prediction model.
A computer program stored in a computer-readable recording medium containing a sequence of instructions providing route information for logistics,
When the computer program is executed by a computing device,
Receive transportation data from the transportation company's terminal,
Entering the received transport data into a transport dispatch model to derive a first transport route including a plurality of unit routes,
Generate at least one three-dimensional matrix for each time period based on transition type information,
Simulating a route change of the first transportation route by combining values of the three-dimensional matrix for each time zone based on a first transportation scenario that satisfies an objective function for the logistics transportation time zone,
A sequence of instructions for changing the first transport route to a second transport route based on the results of the simulation,
The sequence of instructions for generating the 3D matrix for each time slot updates the at least one 3D matrix for each time slot based on the results of the simulation, actual transportation results of the logistics vehicle, and transportation difference information between past transportation plans. Contains a sequence of instructions,
A computer program stored in a computer-readable recording medium, wherein the at least one three-dimensional matrix for each time period includes at least one unit path among the plurality of unit paths.