Cargo three-dimensional boxing method and platform based on path planning and order splitting and medium

Abstract

The invention relates to a three-dimensional cargo boxing method and platform based on path planning and order splitting and a medium, and the method comprises the steps: carrying out the initial planning of a distribution path of a to-be-distributed cargo according to preset distribution resource information and a distribution planning strategy, and generating initial planning information; based on the logistics path corresponding to the initial planning information, path construction iteration and simulated annealing optimization iteration are carried out by using an ant colony and simulated annealing hybrid algorithm, and current planning information is generated; determining a distribution sequence corresponding to a logistics path corresponding to the current planning information and pre-distributed goods corresponding to each node, and performing three-dimensional boxing planning on all the pre-distributed goods based on a preset three-dimensional boxing strategy and an inverted sequence of the distribution sequence, and determining whether the planned current three-dimensional boxing planning information meets the requirement of boxing all the pre-distributed cargos into the corresponding distribution vehicles or not, and repeatedly executing to generate the current planning information and plan the current three-dimensional boxing planning information until the boxing requirement is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent logistics and warehouse management, and particularly relates to a goods three-dimensional packing method, platform and medium based on path planning and order splitting. BACKGROUND

[0002] In the related art, the distribution link of a logistics enterprise involves three stages of vehicle and goods matching (order allocation), path planning and goods loading and unloading (goods loading). In the prior art, due to the defects of a logistics decision system, the management and control of the three stages of the distribution link are in a completely fragmented state. Optimizing the distribution path alone can increase the daily order allocation quantity, but the vehicle turnover rate decreases to cause the loading rate to drop. Conversely, focusing on improving the loading density will cause the distribution route to be circuitous, thereby reducing the on-time rate.

[0003] In the related art, the three stages are independently controlled, which causes the following defects: the pallet space is not reasonably utilized, resulting in pallet waste, the number of loadable goods of each truck is reduced, and the vehicle use cost is indirectly increased; the end distribution is not considered during loading and unloading, but relies on manual experience, and the position of the code plate is randomly placed during packing, thereby causing the corresponding goods to be screened out to achieve distribution during end distribution, which is low in distribution efficiency, and the path planning is unreasonable, thereby reducing the transportation and distribution efficiency.

[0004] At present, there is a problem of low vehicle compartment space utilization rate and low end loading and unloading distribution efficiency in the related art for the distribution decision of the logistics distribution stage, and an effective solution has not been proposed. SUMMARY

[0005] Embodiments of the present application provide a goods three-dimensional packing method, platform and medium based on path planning and order splitting, to at least solve the problem of low vehicle compartment space utilization rate and low end loading and unloading distribution efficiency in the related art for the distribution decision of the logistics distribution stage.

[0006] In a first aspect, the embodiments of the present application provide a three-dimensional packing method for goods based on path planning and order splitting, comprising: performing initial planning of a delivery path for goods to be delivered according to preset delivery resource information and a delivery planning strategy, and generating initial planning information, wherein the initial planning information comprises delivery information representing a delivery scheme of an enabled delivery vehicle delivering required goods to a plurality of nodes in the planning, one of the delivery schemes corresponds to one logistics path, and the delivery planning strategy comprises a nearest neighbor strategy, a savings algorithm strategy, and a random insertion strategy; performing path construction iteration and simulated annealing optimization iteration using a preset ant colony and simulated annealing hybrid algorithm based on the logistics path corresponding to the initial planning information, and generating current planning information, wherein in the path construction process, the corresponding nodes are determined based on pheromone and distance heuristics, and the extension of the corresponding logistics path is based on dynamic capacity association constraints, and the simulated annealing optimization comprises neighborhood operations; determining a delivery order corresponding to the logistics path corresponding to the current planning information and pre-delivery goods corresponding to each node, performing selectable splitting planning, goods pallet planning, and pallet loading planning on all the pre-delivery goods based on a preset three-dimensional packing strategy and the reverse order of the delivery order, and determining whether the current three-dimensional packing planning information generated by the planning meets the requirement of packing all the pre-delivery goods in the corresponding delivery vehicle; before the current three-dimensional packing planning information does not meet the requirement of packing the pre-delivery goods in the corresponding delivery vehicle, repeatedly performing the steps of generating the current planning information and planning the current three-dimensional packing planning information until the corresponding current three-dimensional packing planning information meets the requirement of packing the corresponding pre-delivery goods in the corresponding delivery vehicle, and performing a packing operation based on the corresponding current three-dimensional packing planning information, wherein in the process of repeatedly performing the generation of the current planning information, at least one of evaporation mechanism and enhancement mechanism is used to update the pheromone.

[0007] In a second aspect, the embodiments of the present application provide a service platform, comprising a memory and a processor, characterized in that the memory stores a computer program, and the processor is configured to run the computer program to perform the steps of the three-dimensional packing method for goods based on path planning and order splitting as described in the first aspect.

[0008] In a third aspect, the embodiments of the present application provide a storage medium having a computer program stored thereon, which is executed by a processor to implement the three-dimensional packing method for goods based on path planning and order splitting as described in the first aspect.

[0009] Compared with the related art, the three-dimensional cargo packing method, platform and medium based on path planning and order splitting provided by the embodiment of the present application performs initial planning of the delivery path for the goods to be delivered according to the preset delivery resource information and delivery planning strategy, and generates initial planning information, wherein the initial planning information includes delivery information of a delivery plan that characterizes the delivery of the required goods by the enabled delivery vehicles to the planned multiple nodes, one of the delivery plans corresponds to a logistics path, and the delivery planning strategy includes the nearest neighbor strategy, the saving algorithm strategy and the random insertion strategy; based on the logistics path corresponding to the initial planning information, a preset ant colony and simulated annealing hybrid algorithm is used Perform path construction iteration and simulated annealing optimization iteration to generate current planning information. During the path construction process, the corresponding nodes are determined based on pheromone and distance heuristic factors, and the logistics path corresponding to the dynamic capacity association constraint is extended. The simulated annealing optimization includes neighborhood operations. Determine the delivery order corresponding to the logistics path corresponding to the current planning information and the pre-delivered goods corresponding to each node. Based on the preset three-dimensional packing strategy and the reverse order of the delivery order, perform optional splitting planning, cargo pallet planning and pallet loading planning for all the pre-delivered goods, and determine whether the planned current three-dimensional packing planning information satisfies all the pre-delivered goods. Packing in the corresponding delivery vehicle; before the current three-dimensional packing planning information does not satisfy the requirement of packing the pre-delivered goods in the corresponding delivery vehicle, repeatedly execute the steps of generating the current planning information and planning the current three-dimensional packing planning information until the corresponding current three-dimensional packing planning information satisfies the requirement of packing the corresponding pre-delivered goods in the corresponding delivery vehicle, repeatedly execute the process of generating the current planning information, at least adopt one of the volatilization mechanism and the enhancement mechanism to update the pheromone, solve the problem of low vehicle compartment space utilization and terminal handling and unloading distribution efficiency in the distribution decision-making of the related technology in the logistics distribution stage, and adopt the method of first performing Considering capacity-constrained vehicle path planning, the cargo and vehicle allocation plan under the minimum vehicle scale is determined, and the logistics path including the node access sequence is generated simultaneously. Then, the distribution and unloading sequence constraints of the logistics path based on path planning and the three-dimensional cargo packing planning with unloading sequence constraints using pallets as the medium are carried out. The cargo location preprocessing with optional splitting rules, pallet loading and car loading that maximize space utilization are adopted in sequence to complete the three-dimensional packing optimization of the cargo. In the three-dimensional packing process, the backtracking idea is used to conduct collaborative backtracking optimization of the path-planned logistics path, realizing the effective effect of introducing the cargo order splitting strategy, improving loading efficiency and reducing transportation costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings: Figure 1 is a hardware structure block diagram of a terminal of the three-dimensional cargo packing method based on path planning and order splitting of an embodiment of the application; Figure 2 is a flow chart of the three-dimensional cargo packing method based on path planning and order splitting according to an embodiment of the application; Figure 3 is a structure block diagram of the three-dimensional cargo packing device based on path planning and order splitting according to an embodiment of the application. DETAILED DESCRIPTION

[0011] In order to make the objects, technical solutions and advantages of the present application clearer, the present application is described and explained below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. Based on the embodiments provided in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application. In addition, it can be understood that although the efforts made in this development process can be complex and lengthy, some design, manufacture or production changes made on the basis of the technical content disclosed in the present application by those of ordinary skill in the art related to the content disclosed in the present application are only routine technical means and should not be understood as insufficient disclosure of the present application.

[0012] In the present application, the phrase "embodiment" means that the specific features, structures or characteristics described in combination with the embodiment can be included in at least one embodiment of the present application. The appearance of this phrase at various places in the specification does not necessarily mean the same embodiment, nor is it an independent or alternative embodiment to other embodiments. It is explicitly and implicitly understood by those of ordinary skill in the art that the embodiments described in the present application can be combined with other embodiments without conflict.

[0013] Unless otherwise defined, technical terms and scientific terms used in the present application shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terms "a", "an", "one", "this", and similar referents in the context of describing the application are to be construed to be inclusive, not exclusive. For example, the use of the term "a" or "one" or "the" in referring to a feature or element of the application does not preclude the presence of additional such features or elements, or the use of additional such features or elements, unless otherwise indicated. The terms "including", "containing", "having", and any variations thereof, are intended to cover a non-exclusive inclusion, for example, a process, method, system, product or apparatus that comprises a list of steps or units (elements) is not necessarily limited to those listed steps or units but can include additional steps or units not expressly listed or can include additional steps or units inherent to such process, method, product or apparatus. The term "multiple stages" refers to more than or equal to two stages. The term "and / or" describes an associated relationship between associated objects, which means that there can be three relationships, for example, "A and / or B" can mean that A exists alone, A and B exist together, and B exists alone. The terms "first", "second", "third", and the like merely distinguish similar objects, and do not represent a specific order for the objects.

[0014] The following describes specific embodiments of the method for three-dimensional packing of goods based on path planning and order splitting according to the embodiments of the present application:

[0015] The method provided by the embodiments can be executed in a terminal, a computer or a similar computing device. Taking the case of running on a terminal, Figure 1 is a hardware structure block diagram of a terminal for the method for three-dimensional packing of goods based on path planning and order splitting according to the embodiments of the present application. As shown in Figure 1 , the terminal can include one or more (only one is shown in Figure 1 ) processor 102 (the processor 102 can include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data. Optionally, the terminal can also include a transmission device 106 for communication functions and an input and output device 108. Those skilled in the art can understand that Figure 1 the structure shown is only for illustration, and does not limit the structure of the terminal. For example, the terminal can include more or fewer components than those shown in Figure 1 , or have a different configuration than that shown in Figure 1 .

[0016] The memory 104 can be used to store computer programs, such as software programs of application software and modules, such as the computer program corresponding to the method for three-dimensional packing of goods based on path planning and order splitting in the embodiments of the present application. The processor 102 can execute various functional applications and data processing, i.e., implement the above method, by running the computer program stored in the memory 104. The memory 104 can include a high-speed random access memory, and can further include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some examples, the memory 104 can further include a memory remotely arranged with respect to the processor 102, which can be connected to the terminal 10 through a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0017] The transmission device 106 is used to receive or send data via a network. The specific examples of the above network can include a wireless network provided by a communication provider of the terminal 10. In one example, the transmission device 106 includes a network adapter (NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet in a wireless manner.

[0018] The embodiments of the present application provide a method for three-dimensional packing of goods based on path planning and order splitting running on the above terminal, Figure 2 The flowchart of the method for three-dimensional packing of goods based on path planning and order splitting according to the embodiments of the present application is shown in FIG. 1, which includes the following steps: Figure 2

[0019] In step S201, initial planning of a delivery path is performed on goods to be delivered according to preset delivery resource information and a delivery planning strategy, and initial planning information is generated. The initial planning information includes delivery information representing a delivery scheme of an enabled delivery vehicle to a plurality of nodes for delivering required goods, and one delivery scheme corresponds to one logistics path. The delivery planning strategy includes a nearest neighbor strategy, a saving algorithm strategy, and a random insertion strategy.

[0020] ​In the embodiment, the delivery resource information includes, but is not limited to, alternative delivery vehicles, loading capacity of the delivery vehicles (for example, maximum volume, maximum load) and use cost corresponding to the delivery vehicles; in the embodiment, when processing the comprehensive decision considering vehicle path planning and three-dimensional packing of goods, the comprehensive decision is divided into two stages of planning of a capacitated vehicle routing problem (CVRP) and three-dimensional packing planning considering unloading sequence of goods, and when performing the three-dimensional packing planning considering unloading sequence of goods, it is necessary to determine pre-delivery goods, delivery nodes and delivery sequence of each delivery vehicle, that is, according to the volume weight corresponding capacity information of the goods in the delivery resource information, the goods required by different delivery nodes are reasonably allocated to each enabled delivery vehicle, and then the number of required delivery vehicles, which goods each delivery vehicle delivers and the delivery path of each delivery vehicle are determined. It can be understood that a delivery path is formed in the process that a delivery vehicle delivers pre-delivery goods to corresponding nodes in a delivery sequence, and a delivery path indicates which nodes a delivery vehicle needs to deliver goods to and the delivery sequence. In the embodiment, in the initial planning stage, three delivery planning strategies are used for initial planning, but the number of delivery schemes and corresponding logistics paths generated by each delivery planning strategy is determined according to a set proportion, in the embodiment, 30% is generated by using the nearest neighbor method strategy, 20% is generated by using the saving algorithm, and the remaining 50% is generated by using the random insertion method strategy; in the embodiment, when using any delivery planning strategy for initial planning, after selecting the next node each time, the total volume and load of the current delivery vehicle need to be calculated to constrain the effectiveness of the initial planned logistics path and reduce unnecessary backtracking decision overhead.

[0021] In the embodiment, through initial planning, a set of initial feasible schemes satisfying the vehicle capacity constraint are generated to establish a foundation for subsequent path decision optimization.

[0022] In step S202, based on the logistics path corresponding to the initial planning information, a preset ant colony and simulated annealing hybrid algorithm is used for path construction iteration and simulated annealing optimization iteration to generate current planning information, wherein in the path construction process, the corresponding node is determined based on pheromone and distance heuristic factors, and the extension of the corresponding logistics path is based on dynamic capacity association constraint, and the simulated annealing optimization includes neighborhood operation.

[0023] In the embodiment, after the initial planning generates a set of initial feasible schemes satisfying the vehicle capacity constraint, the set of initial feasible schemes is firstly subjected to path construction optimization based on the ant colony algorithm, that is, further optimization is performed on the initial feasible schemes to generate further optimized logistics paths; in the embodiment, when the path construction optimization based on the ant colony algorithm is performed, the corresponding selection probability is determined according to the pheromone concentration and the distance heuristic factor, and the next visited node is determined according to the selection probability; in the embodiment, the height of the distance heuristic factor is determined by the distance from the current node, and the node with a short distance corresponds to a high heuristic factor; in the embodiment, during the path construction process based on the ant colony algorithm, dynamic capacity management is performed, that is, during the extension of the logistics path (determination of the next node), it is considered whether the current remaining capacity of the distribution vehicle can accommodate the pre-distribution goods required by the next node (the node determined based on the pheromone concentration and the distance heuristic factor), and if it can, the next node is considered to be added to the current logistics path. It can be understood that the process of path construction using the ant colony algorithm is clear to those skilled in the art and does not constitute an unclear limitation on the embodiments of the present application.

[0024] In the embodiment, after the path construction using the ant colony algorithm, the simulated annealing algorithm is used for perturbation operation to locally optimize the logistics path constructed using the ant colony algorithm; in the embodiment, the simulated annealing perturbation is realized by using neighborhood operation, and the selected neighborhood operation includes but is not limited to 2-opt exchange and node exchange across distribution vehicles; at the same time, during the neighborhood operation, corresponding dynamic capacity management is also performed; in the embodiment, during the annealing perturbation iteration process, the perturbation acceptance rate is determined according to the annealing temperature, that is, the perturbation acceptance rate is adjusted according to the annealing temperature, for example: in the high-temperature stage, the logistics path generated by the perturbation operation is accepted with a higher perturbation acceptance rate, and in the low-temperature stage, the logistics path generated by the perturbation operation is accepted with a lower acceptance rate.

[0025] In step S203, the distribution order corresponding to the logistics path corresponding to the current planning information and the pre-distribution goods corresponding to each node are determined, all pre-distribution goods are subjected to selectable disassembly planning, goods pallet planning and pallet loading planning based on the preset three-dimensional packing strategy and the inverse order of the distribution order, and it is determined whether the planned current three-dimensional packing planning information satisfies the packing of all pre-distribution goods in the corresponding distribution vehicle.

[0026] In the embodiment, the three-dimensional packing (3D Bin Packing, 3D-BP) aims to place various shapes and sizes of goods as efficiently as possible in limited space; in the embodiment, the three-dimensional packing is performed according to the reverse order of the delivery sequence of the logistics path corresponding to the current planning information, and the maximum volume filling scheme is tried to find, in the three-dimensional packing stage, three steps of processing are needed, that is, the pre-delivery goods are subjected to the selectable splitting planning (goods preprocessing), the goods pallet planning (pallet loading), and the pallet loading planning (loading of the vehicle compartment), and through the processing of the three stages, it is determined whether the corresponding filling scheme can complete the packing of the corresponding pre-delivery goods, that is, it is determined whether the current three-dimensional packing planning information planned satisfies the packing of all pre-delivery goods in the corresponding delivery vehicle; in the embodiment, in the selectable splitting planning process, the goods are split according to the reverse order of the delivery path, that is, the goods are simulated to be loaded on the corresponding pallet in the process of the reverse order of the delivery path, and the goods that exceed the limit (exceeding the length or height) are split to enable the goods associated with the corresponding node to be loaded on the corresponding pallet; then, the pallet loading is performed, that is, the pallet dynamic space matching is performed, the goods that complete splitting and do not split are placed on the pallet according to the corresponding order (the reverse order of the delivery path), and the goods are sorted according to the descending order of height and then the descending order of volume, and after each sorting and placing, the remaining space of the pallet is traversed to select the smallest adaptive space that satisfies the placement of the next goods, and in the process of selecting the smallest adaptive space, the goods can be selected to try position rotation, thereby realizing dynamic space adaptation; in the embodiment, in the vehicle compartment loading (pallet loading) stage, the global space and the unloading sequence need to be coordinated, the pallets are arranged in the reverse order of the delivery path, and the row-first filling strategy is adopted to load all pallets of the delivery vehicle.

[0027] Step S204, before the current three-dimensional packing planning information does not satisfy the packing of the pre-delivery goods in the corresponding delivery vehicle, the steps of generating the current planning information and planning the current three-dimensional packing planning information are repeatedly executed until the corresponding current three-dimensional packing planning information satisfies the packing of the corresponding pre-delivery goods in the corresponding delivery vehicle, and the packing operation is performed according to the corresponding current three-dimensional packing planning information, wherein in the process of repeatedly generating the current planning information, at least one of the evaporation mechanism and the enhancement mechanism is used to update the pheromone.

[0028] In the embodiment, after completing a three-dimensional packing planning once, if the current three-dimensional packing planning information planned does not satisfy the packing of all pre-delivery goods in the corresponding delivery vehicle, it means that the corresponding loading fails, at this time, the path backtracking mechanism is triggered, the planned logistics path is adjusted, and the pheromone used for vehicle path planning is updated by using the setting update mechanism, for example, the pheromone of the logistics path of the loading failure is volatilized, and the pheromone of the logistics path of the loading success and the corresponding low cost (set as the optimal logistics path) is enhanced, then the CVRP re-planning and the three-dimensional packing cycle operation are performed until the current three-dimensional packing planning information meeting the demand is generated, and then the three-dimensional packing of goods is performed according to the corresponding current three-dimensional packing planning information.

[0029] Through the steps S201 to S204, the initial planning information is generated by using the preset delivery resource information and the delivery planning strategy to plan the delivery path of the goods to be delivered; the current planning information is generated by using the preset ant colony and simulated annealing hybrid algorithm to perform path construction iteration and simulated annealing optimization iteration based on the logistics path corresponding to the initial planning information, in the path construction process, the corresponding node is determined based on the pheromone and distance heuristic factor, and the extension of the corresponding logistics path is based on the dynamic capacity association constraint; the delivery order corresponding to the logistics path of the current planning information and the pre-delivery goods corresponding to each node are determined, all pre-delivery goods are planned for the selectable split, the goods pallet planning and the pallet loading planning based on the preset three-dimensional packing strategy and the reverse order of the delivery order, and it is determined whether the current three-dimensional packing planning information planned satisfies the packing of all pre-delivery goods in the corresponding delivery vehicle; before the current three-dimensional packing planning information does not satisfy the packing of the pre-delivery goods in the corresponding delivery vehicle, the steps of generating the current planning information and planning the current three-dimensional packing planning information are repeatedly executed until the corresponding current three-dimensional packing planning information satisfies the packing of the corresponding pre-delivery goods in the corresponding delivery vehicle, which solves the problem of low vehicle compartment space utilization rate and end unloading delivery efficiency in the related art of delivery decision in the logistics delivery stage, the vehicle path planning considering the capacity constraint is performed first to determine the goods and vehicle allocation scheme under the minimum vehicle scale, the logistics path containing the node access order is generated synchronously, and then the goods three-dimensional packing planning based on the delivery unloading sequence constraint of the logistics path and the pallet unloading sequence constraint of the pallet as the medium is performed, the goods three-dimensional packing optimization is completed by sequentially performing the goods location preprocessing by introducing the selectable split rule, the pallet loading and the compartment loading for maximizing the space utilization rate, and in the three-dimensional packing process, the logistics path of the path planning is cooperatively backtracked and optimized by using the backtracking idea, the effective effect of introducing the goods order splitting strategy, improving the loading efficiency and reducing the transportation cost is achieved.

[0030] In some embodiments, the current planning information is generated by the following steps:

[0031] Step 21, determining the current generated intention planning information, wherein the intention planning information comprises one of the following: initial planning information, current planning information generated by completing the previous path construction iteration and annealing disturbance update iteration.

[0032] Step 22, after determining the current pheromone and current distance heuristic factor corresponding to the logistics path corresponding to the intention planning information, performing local search optimization on the logistics path corresponding to the intention planning information according to the current pheromone and the current distance heuristic factor, and performing capacity constraint management on the generated logistics path corresponding to the cargo capacity in the local search optimization, to generate a first planning logistics path.

[0033] In some optional embodiments, after determining the current generated intention planning information, the following steps are further implemented:

[0034] Step 221, in the case where the intention planning information comprises the initial planning information, determining the current pheromone as the initialization pheromone.

[0035] In the present embodiment, the initialization pheromone is set according to the reciprocal of the distance between nodes, so as to preferentially guide the exploration of adjacent nodes.

[0036] Step 222, in the case where the intention planning information comprises the current planning information generated by completing the previous path construction iteration and annealing disturbance update iteration, determining the current pheromone as the pheromone updated once by using one of the evaporation mechanism and the enhancement mechanism.

[0037] In the present embodiment, the next node is selected according to the selection probability determined according to the pheromone concentration and the distance heuristic factor, and the selection probability formula is: wherein, represents the pheromone, the more pheromones accumulated on the logistics path, the higher the probability of selecting the logistics path; in the present embodiment, the pheromone is updated by using the pheromone evaporation and enhancement mechanism, and the pheromone is updated once when triggering the path loading collaborative backtracking optimization, in the present embodiment, after each iteration (backtracking once), the path edge (the path between two nodes) pheromone of all logistics paths is reduced by 40% according to the evaporation coefficient (RHO=0.6), and 60% is reserved, then the path edge of the elite logistics path (such as the lowest 10% of the cost) is enhanced, and the pheromone increment △τ is determined by the solution quality (inversely proportional to the logistics path cost L k ) and the temperature decay factor, and the specific formula is: wherein Q is pheromone intensity constant, T0 is initial temperature, T is current temperature; in this embodiment, the pheromone enhancement is also affected by simulated annealing temperature decay, and the specific formula is Δτ = (Q / L) * (T0-T) / T0.

[0038] In some optional embodiments, in the local search optimization, the capacity constraint management is performed on the generated logistics path corresponding to the cargo capacity, and the management is achieved by the following steps:

[0039] Step 223, respectively determining the first capacity of the pre-delivery cargo corresponding to the searched next node and the current cargo carrying capacity margin of the delivery vehicle, wherein the first capacity at least includes one of volume and weight.

[0040] Step 224, judging whether the cargo carrying capacity margin can accommodate the pre-delivery cargo corresponding to the first capacity, and in the case that the cargo carrying capacity margin can accommodate the pre-delivery cargo corresponding to the first capacity, adding the next node to the corresponding first planning logistics path.

[0041] Step 225, in the case that the cargo carrying capacity margin cannot accommodate the pre-delivery cargo corresponding to the first capacity, rejecting to add the next node, and constructing the corresponding first planning logistics path based on all the nodes that have been searched.

[0042] In this embodiment, in the local search optimization process of the logistics path corresponding to the intention planning information, dynamic capacity management is achieved, that is, in the extension process of the logistics path, if the remaining capacity of the current delivery vehicle cannot accommodate the demand of the next node, the delivery vehicle path is immediately terminated and returned to the warehouse, and then a new delivery vehicle is started, for example, the delivery vehicle 1 has loaded the node F (1.5 cubic meters, 600 kg), the remaining volume of the delivery vehicle is 2.25 cubic meters, the remaining load is 2400 kg, the volume of the next candidate node G is 2.3 cubic meters, and the weight is 1300 kg, although the total weight is 1900 kg (≤3000), but the total volume is 3.8 cubic meters (>3.75), therefore, the node G cannot be added, the delivery vehicle 1 path is closed as “warehouse→F→warehouse”, and the node G is assigned to the delivery vehicle 2.

[0043] Step 23, based on the loading capacity corresponding to the delivery vehicle, performing simulated annealing optimization corresponding to the neighborhood operation on the first planning logistics path to generate the current planning information corresponding to the planning, wherein the neighborhood operation includes one of the following: 2-opt exchange, node exchange across delivery vehicles.

[0044] In this embodiment, the annealing disturbance performs path optimization through the following two neighborhood operations: First, 2-opt exchange (2-opt local flip), specifically, two nodes in the first planning path are selected, and the order therebetween is reversed to shorten the distance, for example, the total distance of the original path "warehouse -> A -> B -> C -> warehouse" is 50 kilometers, and after flipping B and C, the path becomes "warehouse -> A -> C -> B -> warehouse", the distance is reduced to 45 kilometers, at the same time, the capacity of all vehicles is checked after the flip, if the node composition is not changed, the capacity is not exceeded; if the change of node composition or node order causes the vehicle to exceed the limit, the disturbance is rejected; Second, node exchange across delivery vehicles, a node is randomly selected from delivery vehicle A and exchanged with a node in delivery vehicle B, for example, the path of delivery vehicle A is "warehouse -> X -> Y -> warehouse" (volume 3.4 cubic meters), and the path of delivery vehicle B is "warehouse -> Z -> warehouse" (volume 2.1 cubic meters), after exchanging node Y (1.8 cubic meters) and node Z (2.1 cubic meters), the volume of delivery vehicle A becomes 3.4-1.8+2.1=3.7 cubic meters (legal), and the volume of delivery vehicle B becomes 2.1-2.1+1.8=1.8 cubic meters (legal), so the exchange is accepted, if any of the delivery vehicles exceeds the limit after the exchange, a rollback operation is performed.

[0045] Through steps 21 to 23, the generation of the current planning information or the current planning information after path backtracking is realized according to the initial planning information, and the logistics path that needs three-dimensional packing and delivery is determined.

[0046] In some embodiments, after the simulated annealing optimization corresponding to the neighborhood operation is performed, the following steps are also implemented: Step 31, determining the candidate logistics path corresponding to the delivery vehicle after the neighborhood operation is completed, and calculating the total capacity of the pre-delivery goods required by all nodes of the candidate logistics path.

[0047] Step 32, judging whether the total capacity exceeds the total carrying capacity of the delivery vehicle, and in the case that the total capacity does not exceed the total carrying capacity of the delivery vehicle, accepting the completed neighborhood operation and taking the candidate logistics path as the logistics path of the current planning information corresponding to the planning.

[0048] Step 33, in the case that the total capacity exceeds the total carrying capacity of the delivery vehicle, rejecting the completed neighborhood operation, and taking the first planning logistics path before the simulated annealing optimization as the logistics path of the current planning information corresponding to the planning.

[0049] In the embodiment, if the neighborhood operation of 2-opt exchange is adopted, the total capacity of each delivery vehicle is checked after the flipped path, and if the node composition is not changed after the flipping, the total capacity of the delivery vehicle is not over the limit; if the change of the node composition or the node sequence causes the total capacity of the delivery vehicle to exceed the limit, the disturbance is rejected; if the neighborhood operation of node exchange across the delivery vehicles is adopted, after the exchange, the total capacity of the delivery vehicle participating in the exchange does not exceed the limit, the corresponding exchange is accepted, and if the total capacity of at least one delivery vehicle exceeds the limit after the exchange, the disturbance is not accepted, and a rollback operation is performed, that is, the state before the exchange is rolled back.

[0050] Through steps 31 to 33, the disturbance operation of simulated annealing and the capacity constraint management are realized to update the logistics path of the current planning information to obtain the expected logistics path, so that the subsequent three-dimensional packing according to the logistics path can realize the packing with maximum space utilization, thereby reducing the distribution cost.

[0051] In some embodiments, according to the preset distribution resource information and the distribution planning strategy, the initial planning of the distribution path of the goods to be distributed is performed to generate initial planning information, and the following steps are implemented: Step 41, after obtaining the target information of the nodes to be distributed and the distribution capacity of the goods to be distributed, candidate vehicles and the carrying capacity corresponding to the candidate vehicles are obtained from the distribution resource information, wherein the candidate vehicles are one of the delivery vehicles to be activated.

[0052] Step 42, based on the target information of the nodes and the selected distribution planning strategy, the initial planning of the distribution path is performed, and during the planning, the total capacity of the pre-distribution goods corresponding to all the selected nodes for the corresponding candidate vehicle is accumulated.

[0053] Step 43, when the difference between the total capacity and the carrying capacity of the candidate vehicle is less than the distribution capacity corresponding to the next selected node, the planning of the visited nodes for the corresponding candidate vehicle is terminated, a first distribution path corresponding to the selected candidate vehicle is constructed with the selected nodes, and the initial planning operation under the selected distribution planning strategy is repeated until the first distribution path with a preset proportion is generated, wherein each distribution planning strategy corresponds to a preset proportion.

[0054] In the embodiment, three distribution planning strategies are used for initial planning, but the number of distribution schemes and first distribution paths generated by each distribution planning strategy is determined according to a set proportion. In the embodiment, 30% is generated by using the nearest neighbor method strategy, 20% is generated by using the saving algorithm, and the remaining 50% is generated by using the random insertion method strategy. Each method needs to strictly check the maximum volume and maximum load constraints of the distribution vehicle. In the embodiment, the nearest neighbor method is to start from the warehouse, and each time the nearest unvisited node is selected to join the current path. For example, the coordinates of nodes A, B, C and D are (5, 10), (8, 15), (12, 8) and (3, 20) respectively, and the warehouse is located at (0, 0). First, the distances from all nodes to the warehouse are calculated. It is assumed that node C (distance 12.8) is the nearest, so the first distribution path is initially “warehouse C”. Then the distances from node C to other nodes are calculated. If node A (distance about 8.6) is the nearest, the first distribution path is updated to “warehouse C A”. At this time, the total volume and load of the distribution vehicle need to be checked. It is assumed that the volume of node C is 1.8 cubic meters, the weight is 900 kilograms, the volume of node A is 1.5 cubic meters, and the weight is 800 kilograms. The total volume is 3.3 cubic meters (≤3.75), and the total load is 1700 kilograms (≤3000). Therefore, it is legal, and the next node selection follows the same rule until a new vehicle is started when more nodes cannot be added. It should be noted that in the embodiment, the dimensions of the parameters related to position coordinates and distance are adapted. For example, the two parameter values in the coordinates correspond to longitude and latitude, and the distance parameter is in meters or kilometers. The example does not specify the corresponding dimension, which does not mean that the parameter is unreasonable. The parameter data used in the embodiment is matched with the simulation system and is set according to a set proportion with the actual position or distance. In the embodiment, the saving algorithm generates the first distribution path by merging the short side. For example, the distance from node A to the warehouse is 10 kilometers, the distance from node B to the warehouse is 15 kilometers, and the distance from node A to B is 8 kilometers. The saved distance after merging is 10+15-8=17 kilometers. If the total volume of the merged path is 3.5 cubic meters (node A: 1.5, node B: 2.0) and the load is 2000 kilograms (node A: 800, node B: 1200), the planning is accepted. Otherwise, the path needs to be split. In the embodiment, the random insertion method randomly selects a node and tries to insert it into the best position. For example, the initial first distribution path is “warehouse C warehouse”, node B is randomly selected and inserted after C to form “warehouse C B warehouse”, and the total volume (1.8+2.0=3.8>3.75) found over limit, therefore rejected insertion, instead opened a new delivery vehicle to perform "warehouse B warehouse" path planning; need to note that no matter what delivery planning strategy is adopted, the total volume and load of the current delivery vehicle need to be calculated in real time every time a node is added, so as to avoid the backtracking overhead caused by illegal initial planning scheme in subsequent stages through constraint processing.

[0055] Step 44, sequentially perform initial planning operation under each delivery planning strategy, and all the preset proportion of the first delivery path generated under each delivery planning strategy is taken as the logistics path corresponding to the initial planning information.

[0056] Through steps 41 to 44, the initial planning of the delivery path of the goods to be delivered is realized according to the delivery resource information and the delivery planning strategy, and the initial planning information is generated.

[0057] In some embodiments, based on the preset three-dimensional packing strategy and the reverse order of the delivery sequence, all pre-delivery goods are subjected to optional split planning, goods pallet planning and pallet loading planning, including the following steps: Step 51, after determining the three-dimensional space of the pallet associated with the delivery vehicle loaded with the corresponding pre-delivery goods and the pallet, traversing the size parameters of all pre-delivery goods corresponding to the logistics path corresponding to the current planning information in the reverse order of the corresponding delivery sequence, determining whether the three-dimensional space can accommodate the pre-delivery goods according to the size parameters traversed in sequence and the size of the three-dimensional space.

[0058] Step 52, based on the preset split strategy, the pre-delivery goods that cannot be accommodated in the three-dimensional space are pre-split into corresponding sub-goods, and then according to the size parameters of the split sub-goods and the size parameters of all pre-delivery goods that can be accommodated in the three-dimensional space, the space region in the three-dimensional space where all target goods are pre-assigned is determined, wherein the split strategy includes one of splitting along the height direction according to the cutting ratio and splitting along the longest side direction according to the cutting ratio, and the target goods include sub-goods and pre-delivery goods that can be accommodated in the three-dimensional space.

[0059] In the preprocessing stage of the three-dimensional packing of the present embodiment, cargo splitting is the key to solving the problem of large size or over-limit. When the cargo cannot be directly loaded into the pallet due to height or plane size, the splitting mechanism will be triggered according to the physical constraints, and the cargo with vertical height over-limit will be cut along the height direction. For example, when the maximum effective height of a truck is 2.99 meters, a cargo with a height of 3.5 meters will be split into an upper half of 2.99 meters and a lower half of 0.51 meters, and the volume and weight of the two will be distributed in proportion to the cutting, while the original length and width attributes are retained. For plane size over-limit, it is preferred to split along the longest side. For example, a cargo with a length of 2.5 meters cannot be placed on a pallet with a maximum length of 1.5 meters, so it is split into two sub-cargos of 1.5 meters and 1.0 meter, and automatic rotation is attempted for adaptation. The first sub-cargo can be rotated to 1.5 meters wide and 1.0 meter long, which can be placed on the pallet, and the second sub-cargo can also be adapted after rotation. In the present embodiment, the splitting process strictly follows the reverse processing principle: in the loading sequence, the node cargo of the later delivery is processed first. For example, the delivery path is warehouse→A→B→C, and the actual loading sequence is C→B→A, to ensure that the cargo of the earlier delivery is not damaged by the subsequent cargo. The sub-cargos after splitting will inherit the parent identification and delivery priority. If node B is split into B-1 and B-2, they will be stored together and the original path order will be maintained to avoid increasing the complexity of unloading due to scattered placement. Space coordinates will also be pre-assigned for the split sub-cargos. For example, if the parent cargo is originally planned to be placed in the lower left corner of the pallet, the sub-cargos may occupy the associated areas in the lower left and upper right, respectively, and fast positioning is achieved through space marking.

[0060] In step 53, the pre-assigned space regions are selected in reverse order of the corresponding delivery sequence, and the three-dimensional space is filled and occupied by the selected space regions according to the pallet rule of minimizing the fragmentation area of the remaining space in the three-dimensional space, to complete the pallet loading of the corresponding cargo. The pallet rule includes one of occupying vertical space in descending order of height, minimum space adaptation, and rotating placement.

[0061] In some optional embodiments, filling and occupying the three-dimensional space with the selected space regions according to the pallet rule of minimizing the fragmentation area of the remaining space in the three-dimensional space includes the following steps: determining a target adaptation region for accommodating the currently selected target cargo in the current corresponding remaining space of the three-dimensional space, and assigning the space region corresponding to the currently selected target cargo to the target adaptation region. The target adaptation region is the region that produces the least fragmented residual area in the remaining space when the target cargo is accommodated in the pallet.

[0062] In the tray loading stage of the embodiment, the space utilization is maximized by multi-dimensional rules. Specifically, the goods are arranged in descending order of height before loading, and the goods with a height of 1.5 meters are given priority over the goods with a height of 1.0 meter. In this way, the vertical space is quickly occupied, and subsequent adjustment is reduced. For goods with the same height, the larger volume is given priority for loading, for example, the goods with a size of 1.5*1.0*1.0 meters are loaded first than the goods with a size of 1.0*0.8*1.0 meters, so as to reduce the risk of forced secondary splitting of large-volume goods due to insufficient space. During the loading process, a dynamic maintenance of the remaining space list is adopted, and the minimum space adaptation principle is adopted to sequentially traverse all the remaining spaces, select the smallest region that can accommodate the current goods, and avoid fragmentation of large spaces by small goods. For example, the plane size of the tray is 1.5*1.5 meters, if the remaining space contains two regions of 1.0*1.0 meters and 0.8*0.8 meters, a 0.7*0.7 meter good will be preferentially placed in the latter. Each good will try two placement directions of 0 degrees and 90 degrees, and automatically select the scheme that can be completely placed and the remaining space is more regular. For example, a 1.2*0.8 meter good needs to occupy a 0.3 meter wide residual space on the right side of the tray in the 0 degree direction, while after rotation, the 0.8*1.2 meter placement may make the remaining space form a complete 0.7*1.5 meter region, thereby being more conducive to subsequent loading. After the goods are placed, three new remaining spaces are generated: right side remaining, upper remaining, and vertical upper space. The adjacent spaces are merged in real time, for example, two consecutive 0.5*1.5 meter right side spaces are automatically combined into 1.0*1.5 meters, reducing the impact of fragmentation on loading efficiency. When a good cannot be placed in the current tray, it is temporarily stored and subsequent goods are processed. After most goods are loaded, the good is reattempted. If it still fails, a new tray is started and the good is placed at the first position.

[0063] In step 54, according to the reverse order of the delivery sequence and using the row-first filling strategy, the tray with completed code plate is arranged in the vehicle compartment of the delivery vehicle.

[0064] In the tray loading stage of the embodiment, the global space and unloading sequence need to be coordinated. The trays are arranged in reverse order of the delivery path, and the node that arrives later corresponds to the tray far from the vehicle door. For example, the delivery sequence of multiple trays is A→B→C→D, and tray D is placed at the innermost side of the vehicle compartment, followed by C, forming a linear layout of “D-C-B-A”, ensuring that there is no need to move the front tray during unloading. During loading, the row-first filling strategy is adopted: the trays are arranged row by row along the length direction of the vehicle compartment, and a new row is started after each row occupies a width of 3 meters. Two 1.5-meter-wide trays are placed in the first row, and a new row is started after a length of 3 meters is occupied, forming a stepped space utilization. If the remaining space at the end of the row is less than 0.8 meters, a new row is directly started at the next position to avoid space waste.

[0065] It should be noted that the pallets can be allowed to be stacked, for the scenario of allowing stacking, the vertical direction is handled in layers, the first layer of pallets is uniformly 1.2 meters high, the second layer is stacked from a height of 1.2 meters, and the total height does not exceed 3 meters; the pallets with greater weight are always placed on the bottom layer, the goods with lower gravity center are placed on the upper layer, the transportation stability is ensured by calculating the centroid distribution of the pallets; the overlap of the bounding boxes of the pallets is checked in real time by using a set collision detection algorithm, if the new pallet intersects with the placed pallet in the three-dimensional space, the new pallet is automatically offset horizontally or vertically until a legal position is found; for special unloading requirements, a sliding channel is reserved for the inner pallet, for example, a roller track is preinstalled at the bottom of the pallet, or a structure that can horizontally translate is designed, to ensure that any pallet can be taken out without disassembling other goods.

[0066] Through steps 51 to 54, the planning of the three-dimensional packing process is realized.

[0067] The following describes a three-dimensional packing method of goods based on path planning and order splitting according to a preferred embodiment of the present application: In the model aspect, the three-dimensional packing problem of goods considering goods splitting and vehicle paths is divided into two stages in the embodiment of the present application, a CVRP model with capacity constraints is established in the first stage, and a three-dimensional packing model is established in the second stage.

[0068] Establishment of the CVRP model with capacity constraints: In the embodiment, the following assumptions are made for the CVRP model with capacity constraints: (1) the sizes of the goods to be packed are all smaller than the size of the vehicle compartment; (2) the goods to be packed and the vehicle compartment are all cuboids with uniform mass; (3) the goods can be placed at any position in the vehicle compartment; (4) there are no fragile or dangerous goods in the packed goods, i.e., the goods can support weight and multi-layer loading; (5) the vehicle travels in a straight line distance from the location nodes of each two customers; (6) the vehicle can meet the volume constraint and weight constraint of the loaded goods; (7) each node has only one vehicle to complete the delivery; (8) each vehicle starts from the distribution center and ends at the distribution center; (9) since the information of the goods in the vehicle is basically determined when the goods are loaded into the vehicle; (10) each user has only one good; then, the following mathematical model with minimum vehicle transportation cost and vehicle usage cost is constructed: The constraint conditions are as follows: 1. Each node is served by a unique delivery vehicle, and the constraint formula is: 2. Consistency between the logistics path and the pre-delivery goods allocation, and the constraint formula is: 3. Association between the delivery vehicle usage and the logistics path, and the constraint formula is: 4. Flow balance: 5. Volume and weight capacity constraints, and the constraint formulas are: , ; 6, sub-loop cancellation, constraint is: , ; Wherein, i represents the i-th node; L represents the node set, L={0, 1, 2,..., m}, 0 represents the warehouse; k represents the k-th distribution vehicle; K represents the distribution vehicle set, K={1, 2,..., k}; v i Represents the volume of the pre-distribution goods of the i-th node; g i Express the weight of the i-th node with the distribution goods; c ij Represents the driving cost from the i-th node to the j-th node, and c ii =0; V represents the maximum volume allowed to be transported by a single distribution vehicle, G represents the maximum load of a single distribution vehicle; F represents the fixed use cost of each distribution vehicle; x ijk Is a decision variable, x ijk ∈{0.1}, indicates whether the k-th distribution vehicle drives from the i-th node to the j-th node, if yes, x ijk =1, otherwise, x ijk =0; y k Is a decision variable, y k ∈{0.1} indicates whether the k-th distribution vehicle is enabled, if yes, y k =1, otherwise, y k =0; z ik Is a decision variable, z ik ∈{0.1}, indicates whether the pre-distribution goods of the i-th node are loaded by the k-th distribution vehicle, if yes z ik =1, otherwise, z ik =0; p ik Is a decision variable, p ik ≥1, indicates the order position of the k-th distribution vehicle visiting node i; Construction of three-dimensional packing model: First, for the construction of three-dimensional packing model, the following assumptions are made: (1) the size of the to-be-packed goods is smaller than the size of the truck; (2) the to-be-packed goods and the truck are cuboids with uniform mass; (3) each order goods can be placed in any position in the truck; (4) the loaded goods do not have fragile or dangerous goods, i.e. can support weight and multi-layer loading; (5) the box cannot be placed in the air, and the bottom of the box must be supported; (6) each node needs to distribute only one goods, i.e. the goods number corresponds to the node number; (7) the weight of the pallet is not considered; Before constructing the mathematical model of three-dimensional packing, through analysis, it is determined that the three-dimensional packing of the embodiment of the application is divided into two stages, the first stage is to pack the goods into pallets, and the second stage is to pack the pallets into the interior of the carriage, that is, the packing is divided into two stages, the first stage is to pack the goods into the pallets as containers, and the second stage is to pack the pallets into the carriage as containers, meanwhile, the constraint conditions of the real packing are considered, and the model construction idea is as follows: (1) the total volume of the goods (including the pallets) loaded in the vehicle cannot exceed the maximum volume of the vehicle, and the total weight of the goods loaded cannot exceed the maximum weight limit of the vehicle; (2) each goods in the pallet cannot exceed the boundary of the pallet when placed; (3) each pallet in the carriage cannot exceed the boundary of the vehicle when placed; (4) each goods in the vehicle cannot overlap each other, that is, at least one dimension does not overlap in the spatial coordinate system; (5) the weight of the pallet is ignored, and the pallet has unlimited bearing capacity; (6) the goods in the vehicle need to meet the rule of "first in, last out", that is, both the unloading sequence of the goods on the pallet and the unloading sequence of the pallet are considered; A total objective function corresponding to the pallet transfer model is established, and the total objective function is as follows: maxf is the total objective function, so as to balance the volume filling rate of the pallets and the number of pallets used, and represents a penalty coefficient; In the three-dimensional packing stage of the embodiment, the following constraints are set: (1) pallet volume constraint, the constraint formula is: ; (2) each goods must be completely packed into the interior of the pallet and cannot exceed the boundary of the pallet, the constraint formulas are respectively: , , ; (3) any two goods in the pallet do not overlap, a large value is given to M as a constraint, and the constraint formulas are respectively: , , , , , , ; (4) the loading and unloading sequence problem is considered, if o qg =1, it indicates that the delivery vehicle travels from node i to node j, that is, the goods q of node i is unloaded before the goods g of node j, that is, the goods g of node j cannot block the goods q of node i, and the constraint formula is: , ; (5) the pallet loading model is constructed, , the total mass of the goods loaded in the carriage cannot exceed the total bearing capacity of the carriage, and the total volume of all the goods in the carriage cannot exceed the maximum volume of the carriage, and the constraint formula is: ; wherein, N represents a set of pre-delivery goods, N = {1, 2, …, n}, each pre-delivery good corresponds to a node; p represents the pth pallet, P represents a set of pallets of a delivery vehicle; l q , w q , h q respectively represent the length, width and height of the qth pre-delivery good; l p , w p , h p respectively represent the length, width and height of the pth pallet; x q , y q , z q respectively represent the position of the qth pre-delivery good in the corresponding pallet, α q is a decision variable, α q ∈{0, 1}, indicating whether the qth pre-delivery good is rotated, if yes, α q =1, otherwise, α q =0; are all decision variables, , indicating the relative position relationship of the qth pre-delivery good and the gth pre-delivery good in the pth pallet; is a decision variable, ∈{0.1}, indicating whether the qth pre-delivery good is located on the right side or below the gth pre-delivery good when sent, if yes =1, otherwise, =0; s k =Z + represents the number of pallets used by the kth delivery vehicle, ;ẟ kp is a decision variable, ẟ kp ∈{0,1}, indicating whether the kth vehicle uses the pth pallet, if yes, ẟ kp =1, otherwise ẟ kp =0.

[0069] After establishing the corresponding mathematical model, firstly, the ant colony-simulated annealing hybrid algorithm is used to solve the vehicle routing problem with capacity constraints, the random insertion method and the saving algorithm are fused to generate a feasible path population in the ant colony initialization stage, the node is selected based on pheromone concentration and distance heuristic factor in path construction, and local disturbance is performed through 2-opt flip and cross-vehicle node exchange, and the pheromone of the elite solution is enhanced and the dynamic temperature scheduling jointly drives the search direction; in the three-dimensional packing stage, the system processes the path nodes in reverse order, generates sub-goods by vertically cutting or horizontally splitting the oversized goods according to the longest side, dynamically loads the goods into the pallet through height-volume two-level sorting and merges the remaining space in real time, and if the loading fails, the path backtracking mechanism is triggered, the high-frequency split node is dynamically migrated to a new vehicle according to the split history, the path selection cost is increased, and the pheromone of the failed path is volatilized, and finally a closed-loop optimization process of 'path generation -> three-dimensional loading verification -> feedback adjustment' is formed, until the global solution that meets the distribution economy and loading feasibility is output.

[0070] In this embodiment, three-dimensional packing includes three-dimensional packing goods preprocessing, pallet loading and car loading, and specific reference can be made to the description of steps 51 to 54 above, which will not be repeated here.

[0071] In this embodiment, the three-dimensional packing of goods based on path planning and order splitting is realized by the steps.

[0072] Step 1, population initialization and capacity constraint verification: hybrid heuristic (30% nearest neighbor + 20% saving algorithm + 50% random insertion) generates initial paths, and the truck volume and load are checked in real time, and when the capacity is exceeded, a new vehicle is forced to start or the solution is discarded.

[0073] Step 2, path construction and annealing disturbance: ants select the next node according to pheromone and heuristic factor, and the probability formula is P=(τ^α)*(η^β) / Σ; the generated path is subjected to 2-opt flip and node exchange by simulated annealing, and the probability of accepting inferior solutions is exp(-Δ / T).

[0074] Step 3, pheromone update and elite reservation: only the elite solution enhances the path pheromone, the volatilization rate RHO=0.6, and the enhancement amount is inversely proportional to the solution cost (Δτ=Q / (sol.cost), Q=100); in the temperature scheduling, if there is no improvement for 5 generations in a row, trigger reheating (T*=1.2).

[0075] Step 4, convergence control and restart mechanism: terminate when the temperature drops to a set value or the optimal solution stagnates for more than a maximum threshold time, and reset the pheromone and retain the elite solution when restarting.

[0076] Step 5, cargo reverse split preprocessing; process node cargo in reverse order according to delivery path, call preset split tool for splitting for size over-limit items, vertical cutting is proportionally distributed in height, horizontal splitting is evenly distributed along the longest side, sub-cargo inherits volume / weight and marks parent ID, split times are recorded to split history dictionary for subsequent backtracking times control.

[0077] Step 6, dynamic space matching of pallets; after sorting the goods in descending order of height and then volume, traverse the remaining space of the pallet: try position rotation for each good, select the smallest fitting space, place it and split the right, upper and vertical remaining space, and merge adjacent areas to get the corresponding remaining space; if the current pallet is full, start a new pallet to continue loading.

[0078] Step 7, row priority loading and collision detection of carriages; arrange the pallets along the length of the carriage, change rows after each row is full, limit the maximum number of stacked layers and check the weight when stacking, and use bounding box overlap to determine collision, and shift horizontally until a legal position is found when there is a conflict.

[0079] Step 8, path-loading collaborative backtracking optimization; if the carriage loading fails, trigger path adjustment, migrate the node with the most split times to other vehicles, and increase the path cost in the ACO algorithm, while evaporating pheromones on failed path edges, and finally bind the loading coordinates to the path solution to ensure execution feasibility.

[0080] The embodiment also provides a goods three-dimensional packing device based on path planning and order splitting, which is used to implement the above-mentioned embodiments and preferred embodiments, and will not be described again. As used below, the terms "module", "unit", "sub-unit", etc. can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware is also possible and contemplated.

[0081] Figure 3 is a structural block diagram of a goods three-dimensional packing device based on path planning and order splitting according to the embodiment of the application, as Figure 3 shown, the device includes a planning module 31, a generation module 32, a processing module 33, and a reshaping module 34:

[0082] The planning module 31 is configured to perform initial planning of a delivery path for goods to be delivered according to preset delivery resource information and a delivery planning strategy, and generate initial planning information, wherein the initial planning information includes delivery information representing a delivery scheme in which an enabled delivery vehicle delivers required goods to a plurality of nodes planned, a delivery scheme corresponds to a logistics path, and the delivery planning strategy includes a nearest neighbor strategy, a savings algorithm strategy, and a random insertion strategy.

[0083] The generating module 32 is coupled to the planning module 31, and is configured to generate the current planning information based on the logistics path corresponding to the initial planning information, and by using a preset ant colony and simulated annealing hybrid algorithm for path construction iteration and simulated annealing optimization iteration. In the path construction process, the corresponding node is determined based on pheromone and distance heuristic factors, and the extension of the corresponding logistics path is determined based on dynamic capacity association constraints. The simulated annealing optimization includes neighborhood operation.

[0084] The processing module 33 is coupled to the generating module 32, and is configured to determine the delivery order corresponding to the logistics path corresponding to the current planning information and the pre-delivery goods corresponding to each node, and perform selectable split planning, goods pallet planning, and pallet loading planning on all pre-delivery goods based on a preset three-dimensional packing strategy and the reverse order of the delivery order, and determine whether the current three-dimensional packing planning information satisfies the packing of all pre-delivery goods in the corresponding delivery vehicle.

[0085] The reshaping module 34 is coupled to the processing module 33, and is configured to repeat the steps of generating the current planning information and planning the current three-dimensional packing planning information before the current three-dimensional packing planning information does not satisfy the packing of the pre-delivery goods in the corresponding delivery vehicle, until the corresponding current three-dimensional packing planning information satisfies the packing of the corresponding pre-delivery goods in the corresponding delivery vehicle, and performs the packing operation with the corresponding current three-dimensional packing planning information. In the repeated execution of the current planning information generation process, at least one of the evaporation mechanism and the enhancement mechanism is used to update the pheromone.

[0086] In some embodiments, the device is also configured to determine the current generated intention planning information, wherein the intention planning information includes one of the following: the initial planning information, the current planning information generated by completing the previous path construction iteration and annealing disturbance update iteration; after determining the current pheromone and the current distance heuristic factor corresponding to the logistics path corresponding to the intention planning information, performing local search optimization on the logistics path corresponding to the intention planning information according to the current pheromone and the current distance heuristic factor, and performing capacity constraint management on the generated goods capacity corresponding to the logistics path in the local search optimization to generate a first planning logistics path; based on the loading capacity corresponding to the delivery vehicle, performing simulated annealing optimization corresponding to the neighborhood operation on the first planning logistics path to generate the current planning information corresponding to the current planning, wherein the neighborhood operation includes one of the following: 2-opt exchange, node exchange across delivery vehicles.

[0087] In some embodiments, after the device determines the current generated intention planning information, it is further configured to determine the current pheromone as an initialization pheromone if the intention planning information comprises initial planning information; and determine the current pheromone as a pheromone updated once by using one of the evaporation mechanism and the reinforcement mechanism if the intention planning information comprises current planning information generated by completing a previous path construction iteration and an annealing disturbance update iteration.

[0088] In some embodiments, the device is further configured to determine a first capacity of the pre-delivery goods corresponding to the searched next node and a current cargo capacity margin of the delivery vehicle, respectively, wherein the first capacity comprises at least one of volume and weight; determine whether the cargo capacity margin can accommodate the pre-delivery goods corresponding to the first capacity; and add the next node to the corresponding first planning logistics path if the cargo capacity margin can accommodate the pre-delivery goods corresponding to the first capacity; and reject adding the next node and build the corresponding first planning logistics path based on all the nodes currently searched if the cargo capacity margin cannot accommodate the pre-delivery goods corresponding to the first capacity.

[0089] In some embodiments, after the simulated annealing optimization corresponding to the neighborhood operation is performed, the device is further configured to determine a candidate logistics path corresponding to the delivery vehicle after the neighborhood operation is completed, and calculate a total capacity of the pre-delivery goods required by all the nodes of the candidate logistics path; determine whether the total capacity exceeds the total cargo capacity corresponding to the delivery vehicle; accept the completed neighborhood operation and take the candidate logistics path as the logistics path of the current planning information corresponding to the current planning if it is determined that the total capacity does not exceed the total cargo capacity corresponding to the delivery vehicle; and reject the completed neighborhood operation and take the first planning logistics path before the simulated annealing optimization is performed as the logistics path of the current planning information corresponding to the current planning if it is determined that the total capacity exceeds the total cargo capacity corresponding to the delivery vehicle.

[0090] In some embodiments, the planning module 31 further comprises:

[0091] The acquisition unit is configured to acquire candidate vehicles and carrying capacities corresponding to the candidate vehicles from the delivery resource information after acquiring target information of nodes to be delivered and delivery capacities of goods to be delivered, wherein the candidate vehicles are one of the delivery vehicles to be activated;

[0092] The planning unit is coupled to the acquisition unit and configured to perform initial planning of the delivery path based on the target information of the nodes and the selected delivery planning strategy, and accumulate a total capacity of the pre-delivery goods corresponding to all the nodes selected for the corresponding candidate vehicles during the planning process.

[0093] The construction unit is coupled to the planning unit, and is configured to terminate planning of the visited nodes for the corresponding candidate vehicle when a difference between the total capacity and a carrying capacity of the candidate vehicle is less than the to-be-delivered capacity corresponding to the next to-be-selected node, and construct a first delivery path corresponding to the candidate vehicle with the selected node, and repeat the initial planning operation under the selected delivery planning strategy until a preset proportion of the first delivery paths are generated, wherein each delivery planning strategy corresponds to a preset proportion;

[0094] The generation unit is coupled to the construction unit, and is configured to sequentially perform the initial planning operation under each delivery planning strategy, and generate all the first delivery paths of the preset proportion under each delivery planning strategy as the logistics path corresponding to the initial planning information.

[0095] In some embodiments, the processing module 33 further includes:

[0096] The traversal unit is configured to traverse, in a reverse order of the corresponding delivery sequence, all size parameters of the pre-delivery goods corresponding to a logistics path corresponding to the current planning information after determining a tray associated with a delivery vehicle carrying the corresponding pre-delivery goods and a three-dimensional space of the tray, and determine whether the three-dimensional space can accommodate the pre-delivery goods according to the sequentially traversed size parameters and a size of the three-dimensional space.

[0097] The preprocessing unit is coupled to the traversal unit, and is configured to pre-split, based on a preset splitting strategy, a pre-delivery good that cannot be accommodated in the three-dimensional space into corresponding sub-goods, and determine a space region in which all target goods are pre-assigned in the three-dimensional space according to size parameters of the split sub-goods and size parameters of all pre-delivery goods that can be accommodated in the three-dimensional space, wherein the splitting strategy includes one of splitting in a height direction according to a cutting ratio and splitting in a longest side direction according to a cutting ratio, and the target goods include the sub-goods and the pre-delivery goods that can be accommodated in the three-dimensional space.

[0098] The palletizing unit is coupled to the preprocessing unit, and is configured to sequentially select the pre-assigned space regions in a reverse order of the corresponding delivery sequence, and fill and occupy the three-dimensional space with the selected space regions according to a palletizing rule that minimizes fragmentation of a remaining space of the three-dimensional space, to complete palletizing of the corresponding goods, wherein the palletizing rule includes one of occupying a vertical space in a descending order of height, minimum space adaptation, and rotating placement.

[0099] The loading unit is coupled to the palletizing unit, and is configured to arrange the tray on which the palletizing is completed in a vehicle compartment of the delivery vehicle according to the reverse order of the delivery sequence and using a row-first filling strategy.

[0100] In some embodiments, the code plate unit is further configured to determine a target fitting region for accommodating the currently selected target goods in the remaining space corresponding to the current three-dimensional space, and assign the space region corresponding to the currently selected target goods to the target fitting region, wherein the target fitting region is a region in which the fragmentation residual region generated by the remaining space is the least when the target goods are accommodated in the pallet.

[0101] The embodiment also provides a service platform, comprising a memory and a processor, the memory storing a computer program, and the processor being configured to execute the computer program to perform the steps in any of the above method embodiments.

[0102] Optionally, the service platform can further comprise a transmission device and an input / output device, wherein the transmission device is connected to the processor, and the input / output device is connected to the processor.

[0103] Optionally, in the embodiment, the processor can be configured to perform the following steps through the computer program:

[0104] S1, performing initial planning of a delivery path for the goods to be delivered according to preset delivery resource information and a delivery planning strategy, and generating initial planning information, wherein the initial planning information comprises delivery information representing a delivery scheme of the delivery vehicles enabled to deliver the required goods to the planned nodes, one delivery scheme corresponds to one logistics path, and the delivery planning strategy comprises a nearest neighbor strategy, a savings algorithm strategy, and a random insertion strategy.

[0105] S2, performing path construction iteration and simulated annealing optimization iteration using a preset ant colony and simulated annealing hybrid algorithm based on the logistics path corresponding to the initial planning information, and generating current planning information, wherein in the path construction process, the corresponding nodes are determined based on pheromone and distance heuristic factors, and the extension of the corresponding logistics path is based on dynamic capacity association constraints, and the simulated annealing optimization comprises neighborhood operations.

[0106] S3, determining the delivery order corresponding to the logistics path of the current planning information and the pre-delivery goods corresponding to each node, performing selectable split planning, goods code plate planning, and code plate loading planning on all the pre-delivery goods based on a preset three-dimensional packing strategy and the reverse order of the delivery order, and determining whether the current three-dimensional packing planning information planned satisfies the packing of all the pre-delivery goods in the corresponding delivery vehicles.

[0107] S4, before the current three-dimensional packing planning information meets the packing of the pre-delivery goods into the corresponding delivery vehicle, repeating the steps of generating the current planning information and planning the current three-dimensional packing planning information until the corresponding current three-dimensional packing planning information meets the packing of the corresponding pre-delivery goods into the corresponding delivery vehicle, wherein during the repeating of generating the current planning information, at least one of the evaporation mechanism and the enhancement mechanism is used to update the pheromone.

[0108] It should be noted that the specific examples in the embodiments can refer to the examples described in the above embodiments and optional implementation manners, and the embodiments will not be described here.

[0109] In addition, in combination with the above-mentioned embodiment of the goods three-dimensional packing method based on path planning and order splitting, the embodiments of the present application can provide a storage medium for implementation. The storage medium has a computer program stored thereon; the computer program is executed by a processor to implement any one of the above-mentioned embodiments of the goods three-dimensional packing method based on path planning and order splitting.

[0110] The above embodiments only express several implementation manners of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of protection of the present application. Therefore, the scope of protection of the patent of the present application should be subject to the appended claims.

Claims

1. A three-dimensional cargo packing method based on path planning and order splitting, characterized in that: include: Based on the preset distribution resource information and distribution planning strategy, initial distribution path planning is performed for the goods to be distributed, and initial planning information is generated, wherein the initial planning information includes distribution information representing a distribution plan for the activated distribution vehicles to distribute the required goods to the planned multiple nodes, each distribution plan corresponds to a logistics path, and the distribution planning strategy includes a nearest neighbor strategy, a conservation algorithm strategy, and a random insertion strategy; Based on the logistics path corresponding to the initial planning information, a preset ant colony and simulated annealing hybrid algorithm is used to perform path construction iterations and simulated annealing optimization iterations to generate current planning information, wherein, during the path construction process, the corresponding nodes are determined based on pheromone and distance heuristic factors, and the logistics path corresponding to the dynamic capacity association constraint is extended, and the simulated annealing optimization includes a neighborhood operation; Determine the delivery order corresponding to the logistics path corresponding to the current planning information and the pre-delivered goods corresponding to each node; perform optional splitting planning, cargo pallet planning, and pallet loading planning for all pre-delivered goods based on a preset three-dimensional packing strategy and the reverse order of the delivery order; and determine whether the planned current three-dimensional packing planning information satisfies the requirements for packing all pre-delivered goods into the corresponding delivery vehicles; Before the current three-dimensional packing planning information does not satisfy the requirement for packing the pre-delivered goods into the corresponding delivery vehicle, the steps of generating the current planning information and planning the current three-dimensional packing planning information are repeatedly executed until the corresponding current three-dimensional packing planning information satisfies the requirement for packing the corresponding pre-delivered goods into the corresponding delivery vehicle, and the packing operation is performed using the corresponding current three-dimensional packing planning information. In the process of repeatedly executing the generation of the current planning information, at least one of a volatilization mechanism and an enhancement mechanism is used to update the pheromone.

2. The method according to claim 1, characterized in that Generate current planning information, including: Determining currently generated intended planning information, wherein the intended planning information includes one of the following: the initial planning information, the current planning information generated by completing a previous path construction iteration and an annealing perturbation update iteration; After determining the current pheromone and the current distance heuristic factor corresponding to the logistics path corresponding to the intended planning information, performing local search optimization on the logistics path corresponding to the intended planning information based on the current pheromone and the current distance heuristic factor, and performing capacity constraint management on the cargo capacity corresponding to the generated logistics path during the local search optimization to generate a first planned logistics path; Based on the loading capacity corresponding to the delivery vehicle, the first planned logistics path is optimized by simulated annealing corresponding to the neighborhood operation to generate the current planning information corresponding to the current planning, wherein the neighborhood operation includes one of the following: 2-opt exchange and node exchange across delivery vehicles.

3. The method according to claim 2, characterized in that After determining the currently generated intention planning information, the method includes: In the case where it is determined that the intended planning information includes the initial planning information, determining that the current pheromone is an initialization pheromone; When it is determined that the intended planning information includes the current planning information generated by completing a previous path construction iteration and an annealing perturbation update iteration, it is determined that the current pheromone is the pheromone that has completed an update using one of a volatilization mechanism and an enhancement mechanism.

4. The method according to claim 2, characterized in that In local search optimization, capacity constraints are managed for the cargo capacity corresponding to the generated logistics path, including: Determining respectively a first capacity of the pre-delivered goods and a current remaining cargo capacity of the delivery vehicle corresponding to the next node found out, wherein the first capacity includes at least one of volume and weight; determining whether the remaining cargo capacity can accommodate the pre-delivered goods corresponding to the first capacity, and if it is determined that the remaining cargo capacity can accommodate the pre-delivered goods corresponding to the first capacity, adding the next node to the corresponding first planned logistics path; When it is determined that the remaining cargo capacity cannot accommodate the pre-delivered goods corresponding to the first capacity, the next node is refused to be added, and the corresponding first planned logistics path is constructed based on all the nodes currently searched.

5. The method according to claim 2, characterized in that After performing the simulated annealing optimization corresponding to the neighborhood operation, the method further includes: Determining a candidate logistics path corresponding to the delivery vehicle after completing the neighborhood operation, and calculating the total capacity of the pre-delivered goods required by all the nodes of the candidate logistics path; Determining whether the total capacity exceeds the total cargo capacity corresponding to the delivery vehicle, and if it is determined that the total capacity does not exceed the total cargo capacity corresponding to the delivery vehicle, accepting the completed neighborhood operation and using the candidate logistics path as the logistics path of the current planning information corresponding to the current planning; When it is determined that the total capacity exceeds the total cargo capacity corresponding to the delivery vehicle, the completed neighborhood operation is rejected, and the first planned logistics path before simulated annealing optimization is used as the logistics path of the current planning information corresponding to the current planning.

6. The method according to claim 1, characterized in that Based on the preset distribution resource information and distribution planning strategy, the initial distribution route planning for the goods to be distributed is carried out, and the initial planning information is generated, including: After obtaining the target information of the node to be delivered and the delivery capacity of the goods to be delivered, obtaining a candidate vehicle and the carrying capacity corresponding to the candidate vehicle from the delivery resource information, wherein the candidate vehicle is one of the delivery vehicles to be activated; Performing initial delivery route planning based on the target information of the node and the selected delivery planning strategy, and accumulating the total capacity of the pre-delivered goods corresponding to all the nodes selected for the corresponding candidate vehicle during the planning process; When the difference between the total capacity and the carrying capacity of the candidate vehicle is less than the to-be-delivered capacity corresponding to the next node to be selected, the planning of the node to be visited for the corresponding candidate vehicle is terminated, and a first delivery path corresponding to the candidate vehicle is constructed using the selected node, and the initial planning operation under the selected delivery planning strategy is repeated until the first delivery path with a preset proportion is generated, wherein each delivery planning strategy corresponds to one preset proportion; The initial planning operation under each of the distribution planning strategies is performed in sequence, and all the first distribution paths with a preset proportion generated under each of the distribution planning strategies are used as the logistics paths corresponding to the initial planning information.

7. The method according to claim 1, characterized in that Based on the preset three-dimensional packing strategy and the reverse order of the delivery order, all the pre-delivered goods are subjected to optional splitting planning, cargo pallet planning, and pallet loading planning, including: After determining the pallet associated with the delivery vehicle carrying the corresponding pre-delivered goods and the three-dimensional space of the pallet, traversing the size parameters of all the pre-delivered goods corresponding to the logistics route corresponding to the current planning information in reverse order of the corresponding delivery order, and determining whether the three-dimensional space can accommodate the pre-delivered goods based on the traversed size parameters and the area size of the three-dimensional space; After pre-dividing the pre-delivered goods that cannot be accommodated in the three-dimensional space into corresponding sub-goods based on a preset splitting strategy, the pre-allocated spatial areas for all target goods in the three-dimensional space are determined based on the size parameters corresponding to the split sub-goods and the size parameters of all pre-delivered goods that can be accommodated in the three-dimensional space, wherein the splitting strategy includes one of splitting according to a cutting ratio along the height direction and splitting according to a cutting ratio along the longest side direction, and the target goods include the sub-goods that can be accommodated in the three-dimensional space and the pre-delivered goods; Selecting the pre-allocated spatial areas in reverse order of the corresponding delivery order, and using the selected spatial areas to pre-plan filling and occupying the three-dimensional space according to a palletizing rule that minimizes the fragmentation of the remaining space in the three-dimensional space, so as to complete palletizing of the corresponding goods, wherein the palletizing rule includes one of occupying vertical space in descending order of height, minimum space adaptation, and rotation placement; According to the reverse order of the delivery order and using a row-first filling strategy, the pallets that have completed the pallet loading are arranged in the compartment of the delivery vehicle.

8. The method according to claim 7, characterized in that According to the code plate rule that minimizes the fragmentation area of ​​the remaining space of the three-dimensional space, the three-dimensional space is pre-planned for filling and occupying using the selected space area, including: determining a target adaptation area for accommodating the currently selected target goods within the remaining space currently corresponding to the three-dimensional space, and allocating the space area corresponding to the currently selected target goods to the target adaptation area, wherein the target adaptation area is the area with the least fragmentation residual area generated by the remaining space when the target goods are accommodated in the pallet.

9. A service platform comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps of the three-dimensional cargo packing method based on path planning and order splitting according to any one of claims 1 to 8.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the three-dimensional cargo packing method based on path planning and order splitting according to any one of claims 1 to 8 is implemented.

Images