WO2022094968A1 - 多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质 - Google Patents

多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质 Download PDF

Info

Publication number
WO2022094968A1
WO2022094968A1 PCT/CN2020/127255 CN2020127255W WO2022094968A1 WO 2022094968 A1 WO2022094968 A1 WO 2022094968A1 CN 2020127255 W CN2020127255 W CN 2020127255W WO 2022094968 A1 WO2022094968 A1 WO 2022094968A1
Authority
WO
WIPO (PCT)
Prior art keywords
corner
palletizing
boxes
height
existing
Prior art date
Application number
PCT/CN2020/127255
Other languages
English (en)
French (fr)
Inventor
段文杰
张致伟
丁有爽
邵天兰
Original Assignee
梅卡曼德（北京）机器人科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 梅卡曼德（北京）机器人科技有限公司 filed Critical 梅卡曼德（北京）机器人科技有限公司
Publication of WO2022094968A1 publication Critical patent/WO2022094968A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
    • B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
    • B65G57/00—Stacking of articles
    • B65G57/02—Stacking of articles by adding to the top of the stack
    • B65G57/16—Stacking of articles of particular shape
    • B65G57/20—Stacking of articles of particular shape three-dimensional, e.g. cubiform, cylindrical
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
    • B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
    • B65G43/00—Control devices, e.g. for safety, warning or fault-correcting
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
    • B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
    • B65G57/00—Stacking of articles
    • B65G57/02—Stacking of articles by adding to the top of the stack
    • B65G57/03—Stacking of articles by adding to the top of the stack from above
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B65—CONVEYING; PACKING; STORING; HANDLING THIN OR FILAMENTARY MATERIAL
    • B65G—TRANSPORT OR STORAGE DEVICES, e.g. CONVEYORS FOR LOADING OR TIPPING, SHOP CONVEYOR SYSTEMS OR PNEUMATIC TUBE CONVEYORS
    • B65G2201/00—Indexing codes relating to handling devices, e.g. conveyors, characterised by the type of product or load being conveyed or handled
    • B65G2201/02—Articles
    • B65G2201/0235—Containers
    • B65G2201/0258—Trays, totes or bins

Definitions

  • the present application relates to the field of intelligent logistics, and more particularly, to a hybrid palletizing method for multi-size boxes, a hybrid palletizing device for multi-size boxes, a hybrid palletizing device for multi-size boxes, and non-volatile computer-readable storage medium.
  • Palletizing refers to placing boxes in the space corresponding to containers such as pallets and cages, and has important applications in the fields of logistics and warehousing.
  • the essence of palletizing is the division and planning of space.
  • Traditional palletizing algorithms such as Manufacturer's Pallet Loading Problem (MPLP) algorithm, because it plans the palletizing space in layers, the single-layer space is divided by The straight line is divided equally, so it can only be applied to the case where the box size is the same; another example is the retailer's Pallet Loading Problem (DPLP) algorithm.
  • MPLP Manufacturer's Pallet Loading Problem
  • DPLP Pallet Loading Problem
  • Adjustment although it can be used for cases with different box sizes, but its adjustment algorithm needs to be optimized based on a large number of boxes of the same size during the adjustment process, and thus cannot handle the situation where there are only one or two boxes of each kind. That is, the MPLP algorithm and the DPLP algorithm cannot be calculated when there are many types of boxes but the number of boxes of each size is small, so they cannot be applied to the palletizing scene of boxes of multiple sizes and a small number of boxes of each size.
  • MPLP Manufacturer's Pallet
  • Embodiments of the present application provide a hybrid palletizing method for multi-size boxes, a hybrid palletizing device for multi-size boxes, a hybrid palletizing device for multi-size boxes, and a non-volatile computer-readable storage medium for multi-size and each Intelligent palletizing of boxes of various sizes and small quantities.
  • the hybrid palletizing method for multi-size boxes includes: acquiring the corner coordinates and corner orientations of each corner of the existing stacking type, and the boxes to be stacked can be placed along the existing plane facing the corners; and the corner orientations are combined with at least two placement methods of the boxes to be stacked to obtain multiple controls, and the collection of multiple controls forms an original control space; simulating placing the boxes to be stacked into the original control space In each of the controls; remove the unfeasible controls in the original control space according to the simulation to obtain an effective control space; obtain the optimal control in the effective control space according to the simulation; and place the to-be-coded box in the optimal control.
  • the hybrid palletizing apparatus for multi-size boxes of embodiments of the present application includes one or more handlers.
  • One or more of the processors are used to: obtain the corner coordinates and corner orientations of each corner of the existing stack, and the boxes to be stacked can be placed along the corners and the existing plane; Combining with at least two ways of placing the boxes to be stacked to obtain multiple controls, the collection of a plurality of the controls forms an original control space; simulating placing the boxes to be stacked into each control in the original control space; Remove infeasible controls in the original control space according to the simulation to obtain an effective control space; obtain the optimal control in the effective control space according to the simulation; and control to place the to-be-stacked box in the optimal control .
  • the hybrid palletizing device for multi-size boxes includes a first acquisition module, a second acquisition module, a simulation module, a third acquisition module, a fourth acquisition module, and a control module.
  • the first acquisition module is used to acquire the corner coordinates and corner orientations of all corners of the existing stack, along which the boxes to be stacked can be placed along the existing plane.
  • the second acquisition module is configured to combine the corner coordinates and the corner orientation of each corner with at least two placement methods of the box to be coded to obtain multiple controls, and the collection of multiple controls forms an original control space .
  • the simulation module is used for simulating each control of placing the to-be-coded box into the original control space.
  • the third acquisition module is used for removing infeasible controls in the original control space according to the simulation to obtain an effective control space.
  • the fourth obtaining module is used for obtaining the optimal control in the effective control space according to the simulation.
  • the control module is used to control placing the boxes to be stacked in the optimal control.
  • the non-volatile computer-readable storage medium of the embodiments of the present application contains a computer program that, when executed by one or more processors, causes the mixed palletizing apparatus for multi-size boxes to perform the following mixed code for multi-size boxes Palletizing method: Obtain the corner coordinates and corner orientation of each corner of the existing stacking type, and the boxes to be stacked can be placed along the existing plane of the corners; At least two placement methods are combined to obtain multiple controls, and the collection of multiple controls forms an original control space; simulate placing the boxes to be stacked into each control in the original control space; remove the original control space according to the simulation Controlling infeasible controls in the space to obtain an effective control space; obtaining an optimal control in the effective control space according to a simulation; and placing the to-be-stacked box in the optimal control.
  • the hybrid palletizing method for multi-size boxes, the hybrid palletizing device for multi-size boxes, the hybrid stacking device for multi-size boxes, and the non-volatile computer-readable storage medium utilize the corners of each corner of the existing stacking type Coordinates and corner orientations to realize the planning of the effective control space, and use simulation to obtain the optimal control in the effective control space, and place the boxes to be coded in the optimal control, without the need to plan the space hierarchically like the MPLP algorithm.
  • the space is divided into straight lines, and there is no need to optimize based on a large number of boxes of the same size as the DPLP algorithm, so it can be applied to the palletizing scene of boxes of multiple sizes and a small number of each size.
  • FIG. 1 is a schematic flowchart of a mixed palletizing method for multi-size boxes according to some embodiments of the present application
  • FIG. 2 is a schematic diagram of different placement methods of boxes to be coded in some embodiments of the present application under the same corner coordinates and corner orientations;
  • FIG. 3 is a schematic structural diagram of a mixed palletizing device for boxes of multiple sizes according to some embodiments of the present application
  • FIG. 4 is a schematic diagram of a box to be coded in some embodiments of the present application placed in the same corner coordinates and with multiple corners facing downward;
  • FIG. 5 is a schematic structural diagram of a mixed palletizing device for boxes of multiple sizes according to some embodiments of the present application.
  • 6 and 7 are schematic flowcharts of a mixed palletizing method for multi-size boxes according to certain embodiments of the present application.
  • FIG. 8 is a schematic diagram of a plan model of a stack-type top-level box according to certain embodiments of the present application.
  • FIG. 9 is a schematic diagram of a plan model of a stack-type top-level box according to certain embodiments of the present application.
  • FIG. 10 is a schematic flowchart of a mixed palletizing method for multi-size boxes according to certain embodiments of the present application.
  • FIG. 11 is a schematic diagram of converting an existing stacking pattern into a contour map performed by a hybrid palletizing method for multi-size boxes according to some embodiments of the present application;
  • FIG. 12 is a schematic diagram of converting a contour map into a coordinate system in the hybrid palletizing method of multi-size boxes according to some embodiments of the present application;
  • FIG. 13 is a schematic flowchart of a mixed palletizing method for multi-size boxes according to certain embodiments of the present application.
  • Figure 14 is a schematic diagram of the height and placement of the actual stack shape in some embodiments of the present application.
  • 15 is a schematic flowchart of a mixed palletizing method for multi-size boxes according to certain embodiments of the present application.
  • 16 is a schematic diagram of placing two vectors with a simulated difference of 90° around each corner by the hybrid palletizing method for multi-size boxes according to some embodiments of the present application;
  • 17 to 19 are schematic flowcharts of a mixed palletizing method for boxes of multiple sizes according to certain embodiments of the present application.
  • FIG. 20 is a schematic diagram of a connection state between a non-volatile computer-readable storage medium and a processor according to some embodiments of the present application.
  • the mixed palletizing method provided in the embodiments of the present application.
  • the method provided by the embodiment of the present application may be executed by the processor in combination with the communicating manipulator, and the processor controls the manipulator to operate after obtaining the calculation result.
  • an embodiment of the present application provides a mixed palletizing method for boxes of multiple sizes.
  • This hybrid palletizing method for multi-size boxes includes:
  • the corner coordinates can be one-dimensional coordinates or two-dimensional coordinates.
  • One-dimensional coordinates are applied to the scene of stacking a row of boxes along a straight line, and two-dimensional coordinates are used for stacking in pallet spaces, stacking platforms, cage cars, and other placement spaces The box scene.
  • the corner orientation includes a direction extending to a certain point in space using the corner coordinates as a starting point. If the corner coordinates are one-dimensional coordinates, the corner orientation can be selected from the positive and negative directions of the one-dimensional coordinate axis.
  • the corner orientation can be selected from the positive and negative directions of the two coordinate axes in the two-dimensional coordinates, and optionally can also be selected from the direction that forms a certain angle with any of the two coordinate axes.
  • the corner coordinates are 45° with the x-axis, and 135° with the x-axis.
  • the selection range of the corner direction is not limited to the above.
  • the selection of the corner orientation depends on whether the box can be placed with reference to this orientation. For example, if a certain edge of a box of a certain size is placed in a certain direction of the corner coordinates, and there is no obstacle that affects the placement of the box, the direction can be determined as the corner coordinates. For another example, if there is a plane in an area bounded by a certain direction of corner coordinates that can accommodate boxes of any size, the direction is determined as the corner orientation.
  • a stacking model can be established by an optical sensor to simulate a real space in a computer, thereby identifying whether there is a plane where the boxes to be stacked can be placed.
  • the placement methods of the boxes to be stacked include: as shown in the lower figure of Figure 2, taking the long side of the box as a reference, that is, the long side of the box and the corner orientation coincide; or, as shown in the figure below. 2 As shown in the figure above, the short side of the box is used as the benchmark, that is, the short side of the box is coincident with the corner.
  • the infeasible control may include that after the boxes are stacked to the control, it is not conducive to the stable control of the stacking type; or, there are obstacles on the moving path from stacking the boxes to the control, so that the boxes cannot be stacked, etc.
  • the optimal control may be determined according to criteria such as whether the stacking of boxes into the control is conducive to the stability of the stacking shape and whether it is conducive to the leveling of the stacking shape.
  • an embodiment of the present application provides a mixed palletizing device 100 for boxes of multiple sizes.
  • the hybrid palletizing apparatus 100 for multi-size cases includes one or more handlers 10 .
  • the mixed palletizing method for multi-size boxes according to the embodiment of the present application can be applied to the mixed palletizing apparatus 100 for multi-size boxes according to the embodiment of the present application.
  • one or more processors 10 are used to execute the methods in 01 to 06.
  • one or more processors 10 can be used to: obtain the corner coordinates and corner orientations of the corners of the existing stack, so that the boxes to be stacked can be placed on the existing plane along the corners; At least two placement methods are combined to obtain multiple controls, and the collection of multiple controls forms the original control space; simulate placing the boxes to be stacked into each control in the original control space; remove the unfeasible controls in the original control space according to the simulation to obtain the effective control space; obtain the optimal control in the effective control space according to the simulation; and place the boxes to be coded in the optimal control.
  • one or more processors 10 combine the coordinates and orientations of the corners of the existing stack with at least two placement methods of the boxes to be stacked to obtain multiple controls, and the collection of multiple controls forms the original control space.
  • the box to be coded on a corner coordinate can be placed with the long side as the reference, or can be placed with the short side as the reference.
  • the corner orientation of the corner coordinate can be one of eight orientations of east, west, south, north, southeast, northeast, southwest, and northwest.
  • sixteen controls can be obtained by combining eight corner orientations and two placement methods of a corner coordinate.
  • the control of serial numbers 1 and 6 can be obtained.
  • the control of serial numbers 2 and 7 can be obtained.
  • the control of serial numbers 9 and 14 can be obtained.
  • the control of the serial numbers 10 and 15 can be obtained. And so on, you can get sixteen controls.
  • an embodiment of the present application provides a mixed palletizing device 200 for boxes of multiple sizes.
  • the mixed palletizing device 200 for boxes of multiple sizes includes a first acquisition module 211 , a second acquisition module 212 , a simulation module 213 , a third acquisition module 214 , a fourth acquisition module 215 and a control module 216 .
  • the mixed palletizing method for multi-size boxes according to the embodiment of the present application can be applied to the mixed palletizing device 200 for multi-size boxes according to the embodiment of the present application.
  • the first acquisition module 211, the second acquisition module 212, the simulation module 213, the third acquisition module 214, the fourth acquisition module 215, and the control module 216 can be used to execute 01, 02, 03, 04, 05, and 06, respectively.
  • the second acquisition module 212 can be configured to combine the corner coordinates and corner orientations of each corner with at least two placement methods of the boxes to be stacked to obtain multiple controls, and the collection of multiple controls forms an original control space.
  • the simulation module 213 is used to simulate each control of placing the boxes to be stacked in the original control space.
  • the third acquisition module 214 is used for removing infeasible controls in the original control space according to the simulation to obtain an effective control space.
  • the fourth obtaining module 215 is used to obtain the optimal control in the effective control space according to the simulation.
  • the control module 216 is used to place the boxes to be stacked in optimal control.
  • the hybrid stacking method for multi-size boxes, the hybrid stacking device 100 for multi-size boxes, and the hybrid stacking device 200 for multi-size boxes utilize the corner coordinates and corner orientations of existing stacking shapes to achieve effective control space planning, and obtain the optimal control in the effective control space through the simulation of the actual situation, and place the boxes to be coded in the optimal control, without the need to plan the space hierarchically like the MPLP algorithm and then control the single space straight line, etc. There is no need for optimization based on a large number of boxes of the same size as the DPLP algorithm, so it can be applied to palletizing scenarios with multiple sizes and a small number of boxes of each size.
  • the mixed palletizing method for boxes of multiple sizes further includes:
  • the photographing device can be driven to take pictures of the stacking (that is, the current situation of the stacking body), and then the photographing data fed back by the photographing device can be obtained, a model can be established according to the photographing data, and then the model can be identified to complete the existing stacking type. Obtain.
  • one or more processors 10 are further configured to execute the method in 07 . That is, one or more processors 10 are used to take pictures of the pallets and build models to obtain existing pallet patterns.
  • the mixed palletizing device 200 for boxes of multiple sizes further includes a fifth acquisition module 217 , and the fifth acquisition module 217 is further configured to execute the method in 07 . That is, the fifth acquisition module 217 is used to take pictures of the stack and build a model to obtain the existing stack type.
  • the device for taking pictures of the pallet may be an industrial camera, a 3D camera, a network camera, or other device with a camera function.
  • the hybrid palletizing device 100 for multi-size boxes or the hybrid palletizing device 200 for multi-size boxes can establish a communication connection with the photographing device through a cable, and the photographing device transmits the photographed and palletized information to the hybrid palletizing device of multi-size boxes through the cable.
  • the palletizing equipment 100 or the mixed palletizing device 200 for boxes of multiple sizes, one or more processors 10 or the fifth acquisition module 217 builds a model based on the information, so as to acquire the existing pallet type.
  • the photographing device can communicate with the mixed palletizing device 100 of multi-size boxes or the mixed palletizing device 200 of multi-size boxes through Bluetooth, wireless local area network (Wireless Fidelity, WiFi), near field communication (Near Field Communication, NFC), wireless Optical communication (LiFi--Light Fidelity, LiFi) establishes a communication connection, and the photographing device transmits the photographed stacking information to the mixed palletizing equipment 100 of multi-size boxes or the mixed code of multi-size boxes through Bluetooth, WIFI, NFC, LiFi, etc.
  • the stacking device 200, one or more processors 10 or the fifth acquisition module 217 builds a model based on this information, thereby acquiring the existing stacking pattern.
  • the device for photographing the pallets may be a device with a photographing function installed in the mixed palletizing device 100 of multi-size boxes or the mixed palletizing device 200 of multi-size boxes.
  • the device directly transmits the photographed and palletized information to one or more processors 10 or the fifth acquisition module 217 through a line, so as to establish a model and acquire the existing pallet shape.
  • the photographing device can be installed on a certain side of the actual stack to take pictures from a certain perspective of the stack to help model and identify the stack type.
  • the above-mentioned angle of view may be the operation angle of view of the manipulator that actually performs the palletizing operation.
  • the manipulator may be as shown in FIG. 3 .
  • the stacking type is to be stacked from bottom to top, it is necessary to refer to the top layer of the stack body for reference, then the top layer of the stack type is the operating angle of the manipulator, and then the camera device is installed above the stack body;
  • the outermost layer of the stacking body for stacking it is necessary to refer to the outermost layer of the stacking body for stacking, then the outermost layer of the stacking type is the operating perspective of the manipulator.
  • the photographing device is installed on the outer side of the stack relative to the accommodating space.
  • the type of model in some embodiments may be a point cloud model or a bounding box model; the model may be established based on the overall, exposed surfaces of the stacking type and the surfaces closest to the photographing device, or may be based on Models are established on the outermost surfaces of the stacking type, and of course, the model can also be established based on the overall stacking type.
  • 07 take pictures of the pallet and build a model to obtain the existing pallet shape, which may include:
  • one or more processors can be used to execute the methods in 071, 072, 073 and 074, that is, one or more processors 10 can be used to photograph the pallets for each palletizing the current stacking type and build an overall model to obtain the existing stacking type; or photograph the first stacking type of the first stacking, and obtain the existing stacking type based on the first stacking type and the position and height of the boxes placed in the subsequent stacking process; or The current stacking type is photographed for each stacking, and only the plane model of the top-level box of the stacking type is established to obtain the existing stacking type; , add the stacked boxes at the corresponding positions of the first stacking type, and correct the added boxes based on the plane model of the top-level box photographed to obtain the existing stacking type.
  • the fifth obtaining module 217 may include an obtaining component 2170 , and the obtaining component 2170 may be used to execute the methods in 071 , 072 , 073 and 074 . That is, the acquisition component 2170 is used for photographing the current stacking pattern of each palletizing and establishing an overall model to obtain the existing palletizing pattern; or for photographing the first palletizing pattern for the first palletizing, based on the first palletizing pattern and subsequent palletizing The position and height of the boxes stacked in the process can be used to obtain the existing stacking type; or the current stacking type of the stacking is photographed for each stacking, and only the plane model of the top-level box of the stacking type is established to obtain the existing stacking type.
  • the processor 10 or the acquisition component 2170 performs the modeling of the stack shape.
  • the device for taking pictures of the pallets may be a device with a photographing function, such as a 3D smart camera, an industrial camera, and a network camera.
  • the photographing device after palletizing a box to be palletized, the photographing device will take pictures of the current pallet pattern, so that the processor 10 can perform modeling. In this way, it can be ensured that the model is completely consistent with the actual palletizing situation, and has a high degree of reduction, thereby ensuring the high accuracy of the mixed palletizing operation.
  • the photographing device first takes pictures of the initial stack shape, and the processor 10 or the acquisition component 2170 builds an initial overall model to obtain the existing stack shape.
  • the photographing device After the hybrid palletizing device 100 or the hybrid palletizing device 200 for multi-size boxes has palletized a to-be-stacked box, the photographing device will take pictures of the new pallet type again, and one or more processors 10 or acquisition components 2170 Then, a new overall model is established according to the newly captured image to update the existing stack type.
  • the photographing device will be synchronized with the mixed palletizing device 100 of multi-size boxes or the mixed palletizing device 200 of multi-size boxes, that is, the mixed palletizing device 100 of multi-size boxes or the mixed palletizing device 200 of multi-size boxes
  • the photographing device After a palletizing operation is performed, the photographing device will perform a photographing operation, so that one or more processors 10 or acquisition components 2170 can update the pallet shape.
  • the photographing device only photographs the first palletizing type based on which the mixed palletizing device 100 of the multi-size boxes or the mixed palletizing device 200 of the multi-size boxes performs the first palletizing, so that the processor 10 or the acquisition component 2170 Modeling is performed to obtain the existing pallet pattern, and one or more processors 10 or acquisition components 2170 estimate the pallet pattern after each stacking operation through the location and height of each palletized box in the existing pallet pattern and subsequent palletizing process.
  • There is a stacking type without triggering the photographing device to take pictures that is, the photographing device only takes a picture once during the palletizing process, and the processor 10 or the acquisition component 2170 only performs modeling once.
  • the model established for the first time can be the overall model of the stacking type, or the plane model of the top-level box of the stacking type.
  • the first stacking type is shown in Figure 8, which is the plane model of the top-level box of the stacking type
  • the solid line part of the pallet in the figure is the stacked box
  • the numbers in it are the height of the box, including: boxes with a height of 50, And a box with a height of 70, if it is decided to fill a box with a height of 60 to the area shown by the dotted line based on the first stacking type, the height of the box to be filled in this area will be increased by 60 on the basis of the current height of 0 in this area as the first operation.
  • the photographing device will photograph the current palletizing type of the mixed palletizing device 100 of multi-size boxes or the mixed palletizing device 200 of multi-size boxes after each palletizing, and then one or more processors 10 or
  • the component 2170 establishes the current stacking shape model after each stacking based on the shooting results of each time, and the established stacking shape is the plane model of the top-level box of the current stamping shape, which can reduce the computational complexity of the modeling data and increase the modeling efficiency, and To a certain extent, the accuracy of modeling can be ensured and the accuracy of mixed palletizing can be improved to a certain extent. For example, as shown in Fig.
  • the modeling based on the photographed data of the current stack type can be shown in (a), which can be represented as including boxes with heights of 70 and heights of 50 in the pallet respectively.
  • the boxes are stacked between the boxes with a height of 70 and the edge of the pallet, and the results obtained by taking pictures and modeling after stacking are shown in (b); Stacked on the box with the height of 70 in the picture (b), the modeling result of the photo after the stacking operation can be shown in the figure (c).
  • the current stacking model will be updated to the picture (d) after the photo is taken.
  • the photographing device will first photograph the mixed palletizing device 100 of the multi-size boxes to perform the first palletizing model modeling before the first palletizing.
  • the stacking equipment 100 will add the stacked boxes at the corresponding positions of the first stacking type, and estimate the current stacking type model.
  • the photographing device will also photograph the top-level box, and then one or more processors 10 establish a corresponding plane model, and correct the added box based on the plane model to obtain the existing stacking shape.
  • the actual palletizing situation is detected by taking pictures, and the previously established model is adjusted by means of the highest surface modeling, so as to ensure the accuracy of the mixed palletizing to the highest degree.
  • the corner coordinates of each corner of the existing stack shape are obtained, including:
  • 012 Use the computational geometry method based on the contour map to obtain the coordinates of each inflection point of the contour map as the corner coordinates of each corner of the existing stack.
  • one or more processors 10 can be used to execute the methods in 011 and 012, that is, one or more processors 10 can be used to convert the existing stack shape into a contour map; Using the computational geometry method to obtain the coordinates of each inflection point of the contour map as the corner coordinates of each corner of the existing stack shape.
  • the contour map in the embodiment of the present application reflects the height of the upper surface of the top-level box of the stacking type from the plane on which the stacking type is placed, which can optionally be calculated according to the depth image, Or it can be obtained according to the established point cloud model.
  • the first obtaining module 211 may include a converting unit 2110 and a first obtaining unit 2111 .
  • the conversion unit 2110 can be used to execute the method in 011, and the first acquisition unit 2111 can be used to execute the method in 012. That is, the conversion unit 2110 is used to convert the existing stack shape into a contour map, and the first acquisition unit 2111 is used to obtain the coordinates of each inflection point of the contour map by using the computational geometry method based on the contour map, as the existing contour map. Corner coordinates of each corner of the stack.
  • the size of the three boxes on the left side of the left picture is 55 cm
  • the size of the remaining two boxes from left to right is a box with a side length of 55 cm * 55 cm * 75 cm (length and width).
  • 55cm, 55cm and 75cm in height) and 110cm*75cm*100cm of boxes length, width and height are 110cm, 75cm and 100cm respectively.
  • the picture on the right is a contour map transformed according to the existing pallet type, and the height in the corresponding area is marked to express the stacking situation of the palletized boxes in the area. Among them, the area with no height marked is represented as the area where the boxes are not stacked, and its height is 0.
  • the inflection point coordinates in the contour map are the coordinates of a point where the extension direction of the contour line changes.
  • one or more processors 10 or the first obtaining unit 2111 obtain each part of the contour map by using a computer geometry method according to the contour map shown in the left figure of FIG. 12 .
  • Inflection point coordinates that is, to establish a coordinate system for the contour map, so as to obtain the coordinates of each inflection point.
  • the coordinate system is established with the lower left corner of the contour map as the reference point, and the coordinates of each inflection point of the contour map can be obtained.
  • the points marked with dotted lines are the contours Part of the inflection point of the line graph. From left to right, the coordinates of each inflection point from low to high are (55, 55), (55, 110), (110, 35), (110, 55), (110, 110), (220, 35) , (220, 110). Take the obtained inflection point coordinates as the corner coordinates of each corner of the existing stack.
  • the corner orientation of each corner of the existing stack type is obtained, including:
  • steps 013 and 014 can help to filter out some directions from various directions of the corner coordinates to determine the corner orientation, thereby helping to reduce the computational complexity of the simulated placement.
  • one or more processors 10 can be used to perform the methods in 013 and 014, that is, one or more processors 10 can be used to convert the existing stack shape into a contour map, and convert the contour lines
  • the height of each inflection point in the figure is taken as the corner height of each corner of the existing stack type; and the direction in which the height in the predetermined area adjacent to each corner is not higher than the corner height of the corresponding corner is taken as the corner orientation of the corner.
  • the conversion unit 2110 can also be used for executing the method in 013
  • the first obtaining unit 2111 is also used for executing the method in 014 . That is, the conversion unit 2110 can also be used to convert the existing stack shape into a contour map, and use the height of each inflection point of the contour map as the corner height of each corner of the existing stack shape.
  • the first obtaining unit 2111 may also be configured to use the direction in which the height of each corner in the adjacent predetermined area is not higher than the corner height of the corresponding corner as the corner orientation of the corner.
  • the size of the predetermined area can be determined randomly; preferably, it can be determined according to the size of the box to be stacked; specifically, the average value of the size of the boxes to be stacked can be used as the size of the predetermined area; or, if there is only one box to be stacked, you can This box size is taken as the predetermined area size.
  • the shape of the predetermined area is a rectangle.
  • one or more processors 10 or conversion units 2110 convert the existing stacking pattern into a contour map and identify the height of the corresponding area to express the stacking situation of the palletized boxes in the area.
  • One or more processors 10 or the first obtaining unit 2111 can obtain the height of each inflection point of the contour map through the height of the corresponding area identifier, and use it as the corner height of each corner of the existing stack.
  • the one or more processors 10 or the first obtaining unit 2111 compare the height of the corner with the height of the predetermined area adjacent to the corner. The direction that is not higher than the height of the corner can be regarded as the corner orientation of the corner.
  • the corner heights of point A include 0 cm, 55 cm, 75 cm and 110 cm.
  • all directions from the east to the south of the corner coordinates satisfy that the height within the predetermined area is not higher than the height of the corner, that is, the height within the predetermined area is toward the area of 0 cm.
  • Each direction is not higher than 0 cm, and the corner orientation can be selected from the east to south area, which can specifically include east, southeast, and south.
  • the corner orientation of point A can be east, southeast, south, southwest, and south.
  • the corner of point A When the corner height of point A is 75 cm, the 0 cm area is not higher than 75 cm, the 55 cm area is not higher than 75 cm, and the 75 cm area is not higher than 75 cm, then the corner of point A can be oriented from the above area. It can be selected from any one of the regions, which can be north, northeast, east, southeast, south, southwest, and west.
  • the height of the corner of point A is 110 cm, the four areas of 0 cm, 55 cm, 75 cm, and 110 cm are not higher than 110 cm, and the corner orientation of point A can be any direction.
  • the height of the corner of point P includes 40 cm and 0 cm, where the 0 cm area includes the gap between the stacked boxes in the existing stack and the unplaced area on the right side of the 40 cm box. .
  • the corner height of 40 cm it can be judged that the directions in both the 0 cm area and the 40 cm area satisfy the corner height of point P within the predetermined area of not higher than 40 cm, that is, the corner orientation of point P can be in the 0 cm area.
  • select in the 40 cm area in some embodiments, it can include west, northwest, north, northeast, east, southeast, south, because there is a 50 cm higher than the corner height 40 cm in the neighborhood (predetermined area) in the southwest direction of point P.
  • the corner heights of the Q point include 0 cm and 50 cm.
  • the 0 cm area is not higher than this value, but since the 0 cm area is a narrow gap between two boxes that cannot meet the requirements of the above predetermined area, there is no corner orientation for this corner height.
  • the direction within the 50 cm area meets the requirement of not being lower than the height within the predetermined area, that is, the corner orientation of point Q can include any direction within the 50 cm area, optionally south, southeast, east , Northeast, North.
  • the direction of the boundary of the stacking area is not regarded as the corner orientation, and boxes cannot be stacked in this direction.
  • the corner heights of point M include 0 cm and 30 cm.
  • corner height of 0 cm its southwest, west, northwest, north, northeast, east, and southeast orientations all meet the requirement that the height in the predetermined area is not higher than the corner height.
  • corner height of 30 cm all directions meet the corner orientation requirements.
  • obtaining the corner orientation of each corner of the existing stack further including:
  • 015 Convert the existing stack shape into a contour map, and use the height of each inflection point of the contour map as the corner height of each corner of the existing stack shape;
  • 016 Simulate the placement of two vectors that are 90° apart around each corner.
  • steps 015, 016, and 017 can help to select a direction with a greater possibility of stably placing the box from various directions of the corner coordinates, thereby helping to improve the effective control space in the original control space.
  • the size of the preset area can be randomly determined; preferably, it can be determined according to the size of the boxes to be stacked; specifically, the average value of the size of the boxes to be stacked can be used as the size of the preset area; or, if there is only one box to be stacked , you can use the box size as the default area size.
  • the shape of the preset area is a rectangle.
  • one or more processors 10 can be used to perform the methods in 015, 016 and 017, that is, one or more processors 10 can be used to convert the existing stack shape into a contour map, and the etc.
  • the height of each inflection point of the high-line graph is used as the corner height of each corner of the existing stack; two vectors with a difference of 90° are placed around each corner; when the heights of the two vectors themselves are the same as the corner heights of the corresponding corners, And when the height of the preset area between the 90° difference between the two vectors is not higher than the corner height of the corresponding corner, the orientation of the two vectors is taken as the corner orientation of the corresponding corner.
  • the first acquisition module 211 may further include a simulation unit 2112 .
  • the conversion unit 2110 is further configured to execute the method in 015, the simulation unit 2112 is further configured to execute the method in 016, and the first acquisition unit 2111 is further configured to execute the method in 017. That is, the conversion unit 2110 is used to convert the existing stack shape into a contour map, and the height of each inflection point of the contour map is used as the corner height of each corner of the existing stack shape.
  • the vectors are placed around each corner, and the first acquisition unit 2111 is used to set the height of the preset area between the two vectors when the height of the two vectors themselves is the same as the corner height of the corresponding corner, and the difference between the two vectors is 90°. When the height is not higher than the corner height of the corresponding corner, the orientation of the two vectors is taken as the corner orientation of the corresponding corner.
  • Point B is the determined corner coordinates, and its corner heights are 0, 55, 75, and 110, respectively, and the xy axes are two vectors that differ by 90°.
  • the x-axis can be placed along the eight directions of east, west, south, north, southeast, northeast, southwest, and northwest, respectively.
  • the x-axis faces east and the y-axis faces north: if both the xy-axis fall in the height area of 75, then the height of xy itself is 75, the corresponding corner height is 75, and, between xy The height in the preset area is not higher than 75, and the current xy direction is the corner orientation; if the y-axis falls on the edge of the 110-height area due to a slight deviation, and the x-axis is still in the 75-height area, the current x itself is at a height of 75 corresponds to the corner height of 75, and the height of y itself is 110.
  • the corresponding corner height is different from the x-axis, and the current xy direction cannot be used as the corner orientation; if the x-axis falls on the edge of the 0-height area due to slight differences, the y-axis If it still falls within the height area of 75, then the height of the current x itself is 0 and the height of the corresponding corner is 0, and the height of the y-axis itself is 75, and the height of the corresponding corner is 75. Orientation cannot be used as a corner orientation.
  • the corner height corresponding to the x-axis is 75 and the y-axis is 75.
  • the corresponding corner height is 110, and the corresponding corner heights of the two are different, so the current xy direction cannot be used as the corner orientation.
  • the hybrid palletizing method for multi-size boxes, the hybrid palletizing device 100 for multi-size boxes, and the hybrid palletizing device 200 for multi-size boxes according to the embodiments of the present application further determine the corner by simulating placement of the boxes to be stacked at the same corner coordinates.
  • the various postures that the boxes to be stacked can be placed are perfected, thereby improving the stacking accuracy of the mixed palletizing device 100 for multi-size boxes or the mixed stacking device 200 for multi-size boxes.
  • 04 remove infeasible controls in the original control space according to the simulation to obtain an effective control space
  • 041 remove the controls that cannot be stacked in the simulation, and the stacking is not stable , and there is no effective fill trajectory control, in order to obtain an effective control space.
  • one or more processors 10 can be used to execute the method in 041, that is, one or more processors 10 are also used to remove the control that the stacking is not stable in the simulation, and the control that there is no effective filling track , in order to effectively control the space.
  • the third obtaining module 214 further includes a second obtaining unit 2140 .
  • the second acquisition unit 2140 is used to execute the method in 041, that is, the second acquisition unit 2140 is used to remove the control that cannot be stacked, the control that is not stacked smoothly, and the control that does not have an effective filling trajectory in the simulation, so as to obtain effective control space.
  • the one or more processors 10 or the second acquisition unit 2140 simulate placement of each control in the original control space formed by the multiple control sets of the bin to be coded based on the model. Effective control space is obtained by screening the results of simulated placement. Assuming that the original control space has 5 controls, the boxes to be coded are placed in these 5 controls in sequence. In an example, if the simulative placement of the boxes to be coded into the 1st to 4th controls, the boxes to be coded can be filled. Enter the 1st to 4th controls, and simulate placing the boxes to be coded into the 5th control, the boxes to be coded cannot be filled into the 5th control.
  • one or more processors 10 or the second obtaining unit 2140 will remove the fifth control in the original control space.
  • one or more processors 10 or the second acquisition unit 2140 will remove the original control space. Control, for example, if the box to be coded is placed in the 1st to 4th controls, although the box to be coded can be filled into the 1st to 4th controls, the box to be coded is filled into the 1st control. After the middle, the to-be-stacked box cannot be placed stably and has a certain inclination angle, which is prone to collapse.
  • One or more processors 10 or the second acquisition unit 2140 will also remove the first control in the original control space.
  • the box to be coded can be placed in the original space, and the box to be coded can be placed stably, at this time, it is necessary to consider whether the box can be moved from the current position to the actual place during the actual placement process.
  • One or more processors 10 or the second acquisition unit 2140 will also remove the control in the original control space, for example, if simulating placing the to-be-coded box into the 2-4th control, even though the to-be-coded box can be Filled into the 2-4th control according to the predetermined attitude, and after the to-be-coded box is filled into the 2-4th control, the to-be-coded box can be placed stably without an inclination angle, and it is not easy to cause collapse, but if When there is no route for placing the box to be loaded into the second control, one or more processors 10 or the second acquisition unit 2140 will also remove the second control in the original control space. Therefore, after removing the controls that cannot be filled and the boxes cannot be
  • the hybrid palletizing method for multi-size boxes, the hybrid palletizing device 100 for multi-size boxes, and the hybrid palletizing device 200 for multi-size boxes simulate removal by one or more processors 10 or the second acquisition unit 2140 Infeasible control to get effective control space.
  • Various possibilities that are not suitable for stacking in the actual stacking process are eliminated in advance, thereby improving the stacking efficiency of the mixed stacking device 100 for boxes of multiple sizes.
  • 05 Obtain optimal control in the effective control space according to the simulation, including:
  • 051 Acquire the first degree of matching between the first parameter set of the box to be coded and the first preset parameter set after the simulation is performed, where the first parameter set includes the contact area, support surface area, and support surface of the box to be coded and adjacent boxes at least one of percentage, number of supported boxes, distance from edge or specified container corner;
  • one or more processors 10 may be used to perform the methods in 051 , 053 , 055 and 057 . That is, one or more processors 10 can be used to obtain the first matching degree between the first parameter set of the box to be coded and the first preset parameter set after the simulation is executed, and the first parameter set includes the contact between the box to be coded and the adjacent box.
  • the second matching degree between the two, the second parameter set includes at least one of the height of the new stack type, the volume ratio, and the unevenness of the top surface; and according to the first matching degree, the second matching degree, the first matching degree prediction
  • the preset first weight and the preset second weight of the second matching degree are used to obtain the total matching degree of each control; the control with the highest total matching degree among the multiple controls is regarded as the optimal control.
  • the fourth acquisition module 215 further includes a third acquisition unit 2150 and a first control unit 2151 .
  • the third acquisition unit 2140 is used for executing the methods in 051 , 053 and 055
  • the first control unit 2151 is used for executing the method in 057 .
  • the third obtaining unit 2150 is configured to obtain the first matching degree between the first parameter set of the box to be coded and the first preset parameter set after the simulation is executed, and the first parameter set includes the contact area between the box to be coded and the adjacent box, At least one of the area of the supporting surface, the percentage of the supporting surface, the number of supporting boxes, the distance from the edge or the specified container corner; obtain the difference between the second parameter set and the second preset parameter set of the new stack type formed after the simulation is performed.
  • the second matching degree, the second parameter set includes at least one of the height, volume ratio, and top surface unevenness of the new stack type; and preset according to the first matching degree, the second matching degree, and the first matching degree
  • the first weight and the preset second weight of the second matching degree obtain the total matching degree of each control.
  • the first control unit 2151 is configured to use the control with the highest total matching degree among the multiple controls as the optimal control.
  • one or more processors 100 or the third acquisition unit 2150 will set a first preset parameter set for the to-be-stacked box before simulating placement of the to-be-stacked box to limit the to-be-stacked box in the existing pallets Ideal placement in type.
  • the first parameter set of the simulated placement of the box to be coded is obtained.
  • the first parameter set includes the contact area between the box to be stacked and the adjacent box, the support surface area, the percentage of the support surface, the number of supported boxes, the distance from the edge or the specified container corner, and the like.
  • One or more processors 10 or the third obtaining unit 2150 obtain the first matching degree by comparing the first preset parameter set with the first parameter set.
  • One or more processors 100 or the third acquisition unit 2150 will also set a second preset parameter set for the new stack type after placement before simulating placement of the boxes to be stacked, so as to define the rational placement situation of the new stack type, After the simulated placement of all the boxes to be stacked is completed, the second parameter set of the current stack type is obtained.
  • the second parameter set includes the height, volume ratio, top surface roughness, etc. of the new stack.
  • One or more processors 10 or the third obtaining unit 2150 obtain the second matching degree by comparing the second preset parameter set with the second parameter set.
  • the one or more processors 10 or the third obtaining unit 2150 may also obtain each item according to the first matching degree, the second matching degree, the preset first weight of the first matching degree, and the preset second weight of the second matching degree. Controls the overall fit. For example, one or more processors 10 simulate placement of each control in the original control space formed by the multiple control sets of the to-be-coded box, and after the simulated placement of the second control, the first matching degree is 90%, If the matching degree is 95%, the preset first weight of the first matching degree is 70%, and the preset second weight of the second matching degree is 30%, the total matching degree of the control can be obtained as 91.5%.
  • the first matching degree is 95%
  • the second matching degree is 90%
  • the preset first weight of the first matching degree is 70%
  • the second matching degree is preset The second weight of 30%
  • the total matching degree of the control can be obtained as 93.5%.
  • one or more processors 1 or the first control unit 2151 will select the control whose total matching degree is 93.5%, that is, the control with the highest total matching degree (the third control) as the optimal control.
  • the hybrid palletizing method for multi-size boxes, the hybrid palletizing device 100 for multi-size boxes, and the hybrid palletizing device 200 for multi-size boxes according to the embodiments of the present application are simulated and effectively controlled by one or more processors 10 or the third acquisition unit 2150 space to obtain the optimal control space.
  • the palletizing methods with low matching degree in the actual palletizing process are filtered in advance, thereby improving the palletizing efficiency of the mixed palletizing device 100 for boxes of multiple sizes.
  • the mixed palletizing method for multi-size boxes provided by the embodiment of the present application further includes:
  • one or more processors 10 can be used to execute the method in 08 , that is, one or more processors 10 can be used to stop performing palletizing when the effective control space is empty.
  • control module 216 further includes a second control unit 2160 .
  • the second control unit 2160 is used to execute the method in 08, that is, the second control unit 2160 is used to stop executing the palletizing when the effective control space is an empty set.
  • all the boxes to be stacked are successfully placed, which means that the effective control space is an empty set, that is, the stacking task is completed, and the stacking is stopped at this time.
  • the boxes to be stacked are not completely placed, but there are remaining boxes, and when there is no placement space for the current stacking type, it means that the effective control space is an empty set, and at this time, the execution of stacking is stopped.
  • an embodiment of the present application further provides a non-volatile computer-readable storage medium 300 including a computer program 301 .
  • the computer program 301 when executed by one or more processors 10, causes the one or more processors 10 to execute the mixed palletizing method of multi-size boxes of any of the above-described embodiments.
  • the computer program 301 when executed by one or more processors 10, it causes the processors 10 to execute the following hybrid palletizing method for boxes of multiple sizes: 01: Obtain the corner coordinates and corner orientations of all corners of the existing pallet, and the orientations along the corners There are planes that can place the boxes to be stacked; 02: Combine the corner coordinates and corner orientations of each corner with at least two placement methods of the boxes to be stacked to obtain multiple controls, and the collection of multiple controls forms the original control space; 03: Simulation Place the to-be-stacked boxes into each control in the original control space; 04: Remove infeasible controls in the original control space according to the simulation to obtain an effective control space; 05: Obtain the optimal control in the effective control space according to the simulation; and 06: Place the to-be-coded box in the optimal control.
  • 01 Obtain the corner coordinates and corner orientations of all corners of the existing pallet, and the orientations along the corners There are planes that can place the boxes to be stacked
  • 02 Combine the corner coordinates and corner orientations of
  • the computer program 301 when executed by one or more processors 10, causes the processors 10 to perform the following hybrid palletizing method for boxes of multiple sizes: 07: Take pictures of the pallets and build models to obtain existing pallet patterns. 071: Photograph the current stacking type for each stacking and build an overall model to obtain the existing stacking type; or 072: Photograph the first stacking type of the first stacking, based on the first stacking type and the boxes placed in the subsequent stacking process or 073: The current stacking type is photographed for each stacking, and only the plane model of the top-level box of the stacking type is established to obtain the existing stacking type; or 074 : Shoot the first palletizing pattern of the first palletizing. Each time the subsequent palletizing is placed, add the stacked boxes to the corresponding position of the first palletizing pattern, and correct the added boxes based on the plane model of the top-level box. Get the existing stack type.
  • the computer program 301 when executed by one or more processors 10, causes the processors 10 to perform the following hybrid palletizing method for boxes of multiple sizes:
  • 011 Convert the existing stack shape into a contour map
  • 012 Use the computational geometry method based on the contour map to obtain the coordinates of each inflection point of the contour map as the corner coordinates of each corner of the existing stack shape.
  • 013 Convert the existing stack shape into a contour map, and use the height of each inflection point of the contour map as the corner height of each corner of the existing stack shape; and 014: Change the heights in the predetermined area adjacent to each corner to The direction of the corner height higher than the corresponding corner is taken as the corner orientation of the corner.
  • 015 Convert the existing stack shape into a contour map, and use the height of each inflection point of the contour map as the corner height of each corner of the existing stack shape;
  • 016 Simulate two vectors with a difference of 90° to swing around each corner Put;
  • 017 When the height of the two vectors themselves is the same as the height of the corresponding corner, and the height of the preset area between the 90° difference between the two vectors is not higher than the corner height of the corresponding corner, the The orientation of the two vectors is taken as the corner orientation of the corresponding corner.
  • computer program 201 when executed by one or more processors 10, causes processor 10 to perform the following hybrid palletizing method for boxes of multiple sizes: 041: remove controls that cannot be stacked in the simulation, controls that do not stack smoothly, and There is no control that effectively fills the trajectory to get an effective control space.
  • the computer program 301 when executed by one or more processors 10, causes the processors 10 to execute the following mixed palletizing method for boxes of multiple sizes:
  • 051 Acquire the first degree of matching between the first parameter set of the box to be coded and the first preset parameter set after the simulation is performed, where the first parameter set includes the contact area, support surface area, and support surface of the box to be coded and adjacent boxes At least one of the percentage, the number of supported boxes, the distance from the edge or the specified container corner; 053: Obtain the second matching degree between the second parameter set of the new stack type formed after the simulation is performed and the second preset parameter set , the second parameter set includes at least one of the height, volume ratio, and top surface unevenness of the new stack type; and 055: a first matching degree preset according to the first matching degree, the second matching degree, and the first matching degree The weight and the preset second weight of the second matching degree obtain the total matching degree of each control; 057: Take the control with the highest total matching degree among the multiple controls as the optimal control.
  • the computer program 301 when executed by one or more processors 10, causes the processors 10 to execute the following mixed palletizing method for boxes of multiple sizes: 08: When the effective control space is an empty set, stop executing the palletizing.
  • Any description of a process or method in a flowchart or otherwise described herein may be understood to represent a module, segment or block of code comprising one or more executable instructions for implementing a particular logical function or step of the process. part, and the scope of the preferred embodiments of the present application includes alternative implementations in which the functions may be performed out of the order shown or discussed, including performing the functions substantially concurrently or in the reverse order depending on the functions involved, which It should be understood by those skilled in the art to which the embodiments of the present application belong.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Stacking Of Articles And Auxiliary Devices (AREA)

Abstract

一种码垛方法、设备(100)、装置(200)及可读存储介质。包括：(01)获取垛型角落坐标及朝向，沿角落朝向存在平面能够放置箱子；(02)各角落坐标及朝向与箱子放置方式组合以得到控制，控制的集合形成原始控制空间；(03)模拟放置箱子至各控制中；(04)根据模拟移除不可行控制以得到有效控制(05)；根据模拟获取最优控制；(06)及将箱子放置在最优控制中。

Description

多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质
优先权信息
本申请请求2020年11月3日向中国国家知识产权局提交的、专利申请号为202011210486.3的专利申请的优先权和权益，并且通过参照将其全文并入此处。
技术领域
本申请涉及智能物流领域，更具体而言，涉及一种多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备、多尺寸箱子的混合码垛装置及非易失性计算机可读存储介质。
背景技术
随着机器人智能技术的发展，机器人智能操作逐渐替代人工操作并成为各领域的主流劳动力，例如对箱子的码垛操作。码垛是指将箱子排放在托盘、笼车等容器对应的空间中，在物流、仓储领域具有重要的应用。
码垛的本质是对空间的分割规划，传统的码垛算法例如生产者码垛(Manufacturer’s Pallet Loading Problem，MPLP)算法，由于其将码垛空间进行分层规划，对所分的单层空间通过直线等分，因此只能应用于箱子尺寸一致的情况；再例如零售商码垛(Distributor’s Pallet Loading Problem，DPLP)算法，在直线分割所得空间的基础上，依据箱子尺寸对每一分割所得空间进行调整，虽然可以用于箱子尺寸不同的情况，但是其调整算法需要在调整过程中基于大量同尺寸箱子进行优化，进而无法很好处理每种箱子只有一、两个的情况。即，MPLP算法和DPLP算法在箱子种类比较多但是每种尺寸的箱子的数量不大时，无法进行计算，也就无法适用于多尺寸且每种尺寸数量少的箱子的码垛场景。
发明内容
本申请实施方式提供一种多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备、多尺寸箱子的混合码垛装置及非易失性计算机可读存储介质，用于多尺寸且每种尺寸数量少的箱子的智能码垛。
本申请实施方式的多尺寸箱子的混合码垛方法包括：获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；模拟放置所述待码箱子至所述原始控制空间的各控制中；根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；根据模拟获取所述有效控制空间中的最优控制；及将所述 待码箱子放置在所述最优控制中。
本申请实施方式的多尺寸箱子的混合码垛设备包括一个或多个处理器。一个或多个所述处理器用于：获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；模拟放置所述待码箱子至所述原始控制空间的各控制中；根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；根据模拟获取所述有效控制空间中的最优控制；及控制将所述待码箱子放置在所述最优控制中。
本申请实施方式的多尺寸箱子的混合码垛装置包括第一获取模块、第二获取模块、模拟模块、第三获取模块、第四获取模块和控制模块。第一获取模块用于获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子。第二获取模块用于将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间。模拟模块用于模拟放置所述待码箱子至所述原始控制空间的各控制中。第三获取模块用于根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间。第四获取模块用于根据模拟获取所述有效控制空间中的最优控制。控制模块用于控制将所述待码箱子放置在所述最优控制中。
本申请实施方式的非易失性计算机可读存储介质包含计算机程序，当所述计算机程序被一个或多个处理器执行时，使得多尺寸箱子的混合码垛装置执行如下多尺寸箱子的混合码垛方法：获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；模拟放置所述待码箱子至所述原始控制空间的各控制中；根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；根据模拟获取所述有效控制空间中的最优控制；及将所述待码箱子放置在所述最优控制中。
本申请实施方式的多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备、多尺寸箱子的混合码垛装置及非易失性计算机可读存储介质利用已有垛型各角落的角落坐标及角落朝向来实现有效控制空间的规划，并利用模拟获取有效控制空间中的最优控制，及将待码箱子放置在最优控制中，无需如MPLP算法一样对空间分层规划后对单空间直线等分，也无需如DPLP算法需要基于大量同尺寸的箱子进行优化，从而能够适用于多尺寸且每种尺寸数量少的箱子的码垛场景。
本申请的实施方式的附加方面和优点将在下面的描述中部分给出，部分将从下面的描述中变得明显，或通过本申请的实施方式的实践了解到。
附图说明
本申请的上述和/或附加的方面和优点从结合下面附图对实施方式的描述中将变得明显和容 易理解，其中：
图1是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图2是本申请某些实施方式的待码箱子在同角落坐标及角落朝向下不同放置方式的示意图；
图3是本申请某些实施方式的多尺寸箱子的混合码垛设备的结构示意图；
图4是本申请某些实施方式的待码箱子在同角落坐标及多角落朝向下放置的示意图；
图5是本申请某些实施方式的多尺寸箱子的混合码垛装置的结构示意图；
图6和图7是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图8是本申请某些实施方式垛型最高层箱子的平面模型的示意图；
图9是本申请某些实施方式垛型最高层箱子的平面模型的示意图；
图10是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图11是本申请某些实施方式的多尺寸箱子的混合码垛方法执行将已有垛型转化为等高线图的示意图；
图12是本申请某些实施方式的多尺寸箱子的混合码垛方法中将等高线图转化为坐标系的示意图；
图13是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图14是本申请某些实施方式实际垛型的高度及摆放情况的示意图；
图15是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图16是本申请某些实施方式的多尺寸箱子的混合码垛方法执行将模拟相差90°的两个向量围绕各角落进行摆放的示意图；
图17至图19是本申请某些实施方式的多尺寸箱子的混合码垛方法的流程示意图；
图20是本申请某些实施方式的非易失性计算机可读存储介质和处理器的连接状态示意图。
具体实施方式
下面详细描述本申请的实施方式，所述实施方式的示例在附图中示出，其中，相同或类似的标号自始至终表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施方式是示例性的，仅用于解释本申请的实施方式，而不能理解为对本申请的实施方式的限制。
智能混合尺寸箱子码垛中，可参照本申请实施例提供的混合码垛方法进行。具体可以由处理器结合相通信的机械手执行本申请实施例提供的方法，处理器获得计算结果后控制机械手进行操作。
请参阅图1，本申请实施方式提供一种多尺寸箱子的混合码垛方法。该多尺寸箱子的混合码垛方法包括：
01：获取已有垛型各角落的角落坐标及角落朝向，沿角落朝向存在平面能够放置待码箱子；
02：将各角落的角落坐标及角落朝向与待码箱子的至少两种放置方式进行组合以得到多个控 制，多个控制的集合形成原始控制空间；
03：模拟放置待码箱子至原始控制空间的各控制中；
04：根据模拟移除原始控制空间中不可行的控制以得到有效控制空间；
05：根据模拟获取有效控制空间中的最优控制；及
06：将待码箱子放置在最优控制中。
部分实施例中，角落坐标可以为一维坐标、二维坐标，一维坐标应用于沿直线码放一排箱子的场景，二维坐标用于在托盘空间、码放平台、笼车等放置空间内码放箱子的场景。可选地，角落朝向包括以角落坐标为起始点，向空间中某一点延伸的方向。若角落坐标为一维坐标，则角落朝向可以由一维坐标轴的正向、负向中选取。若角落坐标为二维坐标，则角落朝向可以由二维坐标中两坐标轴的正向、负向中选取，可选地还可以从与两坐标轴中任一轴形成某一夹角的方向中选取，例如角落坐标与x轴呈现45°角，与x轴呈现135°角等。当然角落方向的选取范围不局限于以上。较佳的，角落朝向的选取，需取决于参照此方向能否摆放箱子。举例而言，若将某尺寸箱子的某一边沿角落坐标的某一方向放置时，不存在障碍物影响箱子的放置，则可将该方向确定为角落坐标。再例如若角落坐标的以某一方向为边界的区域内存在平面可以容纳任意尺寸的箱子，则将该方向确定为角落朝向。
部分实施例中，可以通过光学传感器建立垛型模型在计算机中模拟真实空间，进而识别是否存在平面能够放置待码放箱子。
部分实施中，若确定角落坐标及相应的角落朝向，待码放箱子的放置方式包括：如图2下图所示，以箱子长边为参照，即将箱子长边与角落朝向重合；或者，如图2上图所示，以箱子短边为基准，即将箱子短边与角落朝向重合。
部分实施例中，不可行的控制可以包括箱子码放至该控制后，不利于垛型稳定的控制；或者，在码放箱子至该控制的移动路径上存在障碍物导致箱子无法被码放等。
部分实施例中，最优控制可以根据箱子码放至该控制中是否有利于垛型稳定、是否有利于垛型平整等判定标准进行确定。
请参阅图3，本申请实施方式提供一种多尺寸箱子的混合码垛设备100。多尺寸箱子的混合码垛设备100包括一个或多个处理器10。本申请实施方式多尺寸箱子的混合码垛方法可应用于本申请实施方式的多尺寸箱子的混合码垛设备100。其中，一个或多个处理器10用于执行01至06中的方法。即，一个或多个处理器10可用于：获取已有垛型各角落的角落坐标及角落朝向，沿角落朝向存在平面能够放置待码箱子；将各角落的角落坐标及角落朝向与待码箱子的至少两种放置方式进行组合以得到多个控制，多个控制的集合形成原始控制空间；模拟放置待码箱子至原始控制空间的各控制中；根据模拟移除原始控制空间中不可行的控制以得到有效控制空间；根据模拟获取有效控制空间中的最优控制；及将待码箱子放置在最优控制中。
具体地，一个或多个处理器10通过已有垛型各角落的坐标及朝向与待码箱子的至少两种放置方式进行组合以得到多个控制，多个控制的集合形成原始控制空间。如图2所示，对于在一个角落坐标上待码箱子可以是以长边为参照的放置方式，也可以是以短边为参照的放置方式。对于一个角落坐标，该角落坐标的角落朝向可以为东、西、南、北、东南、东北、西南、西北八个朝向中的一种。如图4所示，对一个角落坐标的八个角落朝向及两种放置方式进行组合，则可以得到十六个控制。例如，当角落朝向为东时，请结合图4中的图(a)和图(c)，根据以不同边为基准的放置，可以得到序号1和6的控制。又例如，当角落朝向为北时，请继续结合图4中的图(a)和图(c)，根据以不同边为基准的放置，可以得到序号2和7的控制。再例如，当角落朝向为东北时，请结合图4中的图(b)和图(d)，根据以不同边为基准的放置，可以得到序号9和14的控制。还例如，当角落朝向为西北时，请结合图4中的图(b)和图(d)，根据以不同边为基准的放置，可以得到序号10和15的控制。以此类推，即可得到十六个控制。
请参阅图5，本申请实施方式提供一种多尺寸箱子的混合码垛装置200。多尺寸箱子的混合码垛装置200包括第一获取模块211、第二获取模块212、模拟模块213、第三获取模块214、第四获取模块215和控制模块216。本申请实施方式的多尺寸箱子的混合码垛方法可应用于本申请实施方式的多尺寸箱子的混合码垛装置200。其中，第一获取模块211、第二获取模块212、模拟模块213、第三获取模块214、第四获取模块215和控制模块216可分别用于执行01、02、03、04、05和06中的方法。即第一获取模块211可用于获取已有垛型各角落的角落坐标及角落朝向，沿角落朝向存在平面能够放置待码箱子。第二获取模块212可用于将各角落的角落坐标及角落朝向与待码箱子的至少两种放置方式进行组合以得到多个控制，多个控制的集合形成原始控制空间。模拟模块213用于模拟放置待码箱子至原始控制空间的各控制中。第三获取模块214用于根据模拟移除原始控制空间中不可行的控制以得到有效控制空间。第四获取模块215用于根据模拟获取有效控制空间中的最优控制。控制模块216用于将待码箱子放置在最优控制中。
本申请实施方式的多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备100、多尺寸箱子的混合码垛装置200利用已有垛型各角落的角落坐标及角落朝向来实现有效控制空间的规划，并通过对现实情况的模拟仿真获取有效控制空间中的最优控制，及将待码箱子放置在最优控制中，无需如MPLP算法一样对空间分层规划后对单空间直线等分，也无需如DPLP算法需要基于大量同尺寸的箱子进行优化，从而能够适用于多尺寸且每种尺寸数量少的箱子的码垛场景。
请参阅图6，本申请实施方式提供的多尺寸箱子的混合码垛方法，还包括：
07：对码垛进行拍照并建立模型以获取已有垛型。
部分实施例中具体可以驱动拍照设备对码垛(即垛体当前的情况)进行拍照，而后获取拍照设备反馈的拍照数据，根据拍照数据建立模型，而后对模型进行识别，完成已有垛型的获取。
请结合图3，在某些实施方式中，一个或多个处理器10还用于执行07中的方法。即一个或 多个处理器10用于对码垛进行拍照并建立模型以获取已有垛型。
请结合图5，在某些实施方式中，多尺寸箱子的混合码垛装置200还包括第五获取模块217，第五获取模块217还用于执行07中的方法。即，第五获取模块217用于对码垛进行拍照并建立模型以获取已有垛型。
具体在一个实施例中，对码垛进行拍照的设备可以是工业相机、3D相机、网络相机等具有拍照功能的设备。多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200可以与拍照设备通过线缆建立通信连接，拍照设备通过线缆将拍摄码垛后的信息传输给多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200，一个或多个处理器10或第五获取模块217通过该信息建立模型，从而获取已有垛型。又例如，拍照设备可以与多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200通过蓝牙、无线局域网(Wireless Fidelity，WiFi)、近场通信(Near Field Communication，NFC)、无线光通信(LiFi--Light Fidelity，LiFi)等建立通信连接，拍照设备通过蓝牙、WIFI、NFC、LiFi等将拍摄垛型信息传输给多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200，一个或多个处理器10或第五获取模块217通过该信息建立模型，从而获取已有垛型。在另一个实施例中，对码垛进行拍照的设备可以是安装在多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200内的具有拍照功能的设备。该设备通过线路将拍摄码垛后的信息直接传输给一个或多个处理器10或第五获取模块217，从而建立模型，获取已有垛型。
部分实施例中，拍照设备可以安装在实际垛体的某一侧，从垛体某一视角拍照，帮助建模识别垛型。较佳的，上述视角可以为实际进行码垛操作的机械手的操作视角，可选地机械手可如图3示意。例如，若对垛型进行从下至上顺序的码垛，需要参照垛体的最高层情况进行码垛参考，则垛型最高层为机械手的操作视角，进而拍照设备安装在垛体的上方；若对垛型进行在某一容纳空间中(例如厢式货车)的从内至外的码垛，需要参照垛体的最外层情况进行码垛，则垛型最外层为机械手的操作视角，进而拍照设备安装在垛体相对容纳空间中的外侧。
可选地，部分实施例中的模型种类可以为点云模型、也可以为包围盒模型；可以基于码垛垛型整体的、裸露在外的、距离拍照设备最近的各面建立模型，也可以基于组成码垛垛型的最外围各面建立模型，当然也可以基于码垛垛型整体建立模型。
请参阅图7，在某些实施方式中，07：对码垛进行拍照并建立模型以获取已有垛型，可包括：
071：每一次码垛均拍摄码垛的当前垛型并建立整体模型以获取已有垛型；或
072：拍摄首次码垛的首次垛型，基于首次垛型及后续码垛过程中所码放箱子的位置、高度以获取已有垛型；或
073：每一次码垛均拍摄码垛的当前垛型，并只建立垛型的最高层箱子的平面的模型，以获取已有垛型；或
074：拍摄首次码垛的首次垛型，后续码垛每码放一次，在首次垛型的相应位置上增加所码放的箱子，并基于所拍摄的最高层箱子的平面模型对所增加箱子进行矫正，以获取已有垛型。
请结合图3，在某些实施方式中，一个或多个处理器可用于执行071、072、073和074中的方法，即一个或多个处理器10可用于每一次码垛均拍摄码垛的当前垛型并建立整体模型以获取已有垛型；或拍摄首次码垛的首次垛型，基于首次垛型及后续码垛过程中所码放箱子的位置、高度以获取已有垛型；或每一次码垛均拍摄码垛的当前垛型，并只建立垛型的最高层箱子的平面的模型，以获取已有垛型；或拍摄首次码垛的首次垛型，后续码垛每码放一次，在首次垛型的相应位置上增加所码放的箱子，并基于所拍摄的最高层箱子的平面模型对所增加箱子进行矫正，以获取已有垛型。
请结合图5，在某些实施方式中，第五获取模块217可包括获取组件2170，获取组件2170可用于执行071、072、073和074中的方法。即，获取组件2170用于每一次码垛均拍摄码垛的当前垛型并建立整体模型以获取已有垛型；或用于拍摄首次码垛的首次垛型，基于首次垛型及后续码垛过程中所码放箱子的位置、高度以获取已有垛型；或每一次码垛均拍摄码垛的当前垛型，并只建立垛型的最高层箱子的平面的模型，以获取已有垛型；或拍摄首次码垛的首次垛型，后续码垛每码放一次，在首次垛型的相应位置上增加所码放的箱子，并基于所拍摄的最高层箱子的平面模型对所增加箱子进行矫正，以获取已有垛型。
具体地，通过拍摄垛体的当前情况，处理器10或获取组件2170进行垛型的建模，具体地，一个或多个处理器10或获取组件2170根据拍照设备拍摄的垛型情况建立模型以获取已有垛型的方式具有多种。其中，对码垛进行拍照的设备可以是3D智能相机、工业相机、网络相机等具有拍照功能的设备。
在一个实施例中，对一个待码箱子执行码垛后，拍照设备则会对当前垛型进行拍摄，以使得处理器10进行建模。如此可以保证模型与实际垛型情况完全吻合，有较高的还原度，进而保证混合码垛操作的高精准度。例如，拍照设备先对最初的垛型进行拍摄后处理器10或获取组件2170建立最初的整体模型获取已有垛型，所建立模型可如图11左侧立体效果部分所示，当多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200对一个待码箱子进行码垛后，拍照设备则会再次对新的垛型进行拍摄，一个或多个处理器10或获取组件2170则根据新拍摄的图像建立新的整体模型以更新已有垛型。其中，拍照设备会与多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200保持同步，就是说，多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200进行一次码垛操作后，则拍照设备便会执行拍摄操作，使得一个或多个处理器10或获取组件2170进行垛型的更新。
在另一个实施例中，拍照设备只拍摄多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200进行首次码垛前所基于的首次垛型，使得处理器10或获取组件2170进行建模进而获 得已有垛型，一个或多个处理器10或获取组件2170通过已有垛型及后续码垛过程中的每次被码箱子的位置、高度估计各次码放操作后的已有垛型，而不触发拍照设备进行拍照，也就是说，在码垛过程中拍照设备只进行一次拍照，处理器10或获取组件2170只进行一次建模。如此节省了建模过程中所占用的时间、计算资源，提高混合码垛整体效率。其中首次所建立的模型可以是垛型的整体模型，也可以是垛型的最高层箱子的平面模型。举例而言，若首次垛型如图8，为垛型最高层箱子的平面模型，图中托盘中实线部分为已码放的箱子，其中的数字为箱子高度，分别包括：高50的箱子、以及高为70的箱子，若基于首次垛型，决定将高度为60的箱子填充至虚线所示区域，则在该区域当前高度0的基础上增加填充该区域的箱子的高度60，作为首次操作后所估计的、第二次操作所基于的二次垛型；若决定在第二次垛型基础上在高为60的箱子上填充高度为50的箱子，则所填充区域的高度为当前的60+50＝110，作为第二次操作后所估计的、第三次操作所基于的三次垛型。需要说明的是，每一次所填充区域的尺寸与填充该区域的箱子的长宽尺寸相一致。
在再一个实施例中，拍照设备会拍摄多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200每次码垛后的当前垛型，进而一个或多个处理器10或获取组件2170基于每次的拍摄结果建立每次码垛后的当前垛型模型，且所建立垛型为当前跺型最高层箱子的平面模型，如此可以减少建模数据运算量增加建模效率，且能够一定程度上保证建模的精确度进而一定程度上提高混合码垛的精确度。举例而言，如图9，基于当前垛型的拍照数据进行的建模可如(a)所示，其可示意为在托盘中分别包括高70、高50的箱子，若决定将高60的箱子码放至高70的箱子与托盘边缘之间，则在码放后拍照建模所获得结果如图中(b)所示；基于图中(b)所示模型，若经过判定决定将高50的箱子码放在图示(b)中高70的箱子上，则码放操作后所拍照建模结果可如图中(c)所示；基于图中(c)所示模型，若经过判定决定将高50的箱子码放在图示(c)中高50的箱子上，则在拍照后当前垛型模型更新为图示(d)。
在又一个实施例中，拍照设备首先会拍摄多尺寸箱子的混合码垛设备100进行首次码垛前所基于的首次垛型建模，当码垛操作每进行一次，则多尺寸箱子的混合码垛设备100会在首次垛型的相应位置上增加码放的箱子，预估当前垛型模型。同时拍照设备还会拍摄最高层箱子，进而一个或多个处理器10建立对应的平面模型，并基于平面模型对所加箱子进行校正，以获取已有垛型。如此，本实施例在建立模型的基础上，通过拍照探查实际码垛情况，通过最高面建模的方式调整在先建立的模型，最高程度保证了混合码垛的精准度。
请参阅图10，在某些实施方式中，获取已有垛型各角落的角落坐标，包括：
011：将已有垛型转化为等高线图；及
012：采用基于等高线图的计算几何学方法获取等高线图的各拐点坐标，以作为已有垛型各角落的角落坐标。
请结合图3，一个或多个处理器10可用于执行011和012中的方法，即一个或多个处理器10可用于将已有垛型转化为等高线图；采用基于等高线图的计算几何学方法获取等高线图的各拐点坐标，以作为已有垛型各角落的角落坐标。本领域技术人员可以理解到的是，本申请实施例中的等高线图反应垛型的最高层箱子的上表面距离垛型所码放平面的高度，可选地可以根据深度图像计算而得，或者可以根据所建立的点云模型获得。
请结合图5，第一获取模块211可包括转换单元2110和第一获取单元2111。转换单元2110可用于执行011中的方法，第一获取单元2111可用于执行012中的方法。即转换单元2110用于将已有垛型转化为等高线图，第一获取单元2111用于采用基于等高线图的计算几何学方法获取等高线图的各拐点坐标，以作为已有垛型各角落的角落坐标。
在一个实施例，如图11所示，左图左侧三个箱子尺寸为55厘米，剩余两个箱子从左到右尺寸依次为边长为55厘米*55厘米*75厘米的箱子(长宽高分别为55厘米、55厘米和75厘米)和110厘米*75厘米*100厘米的箱子(长宽高分别为110厘米、75厘米和100厘米)。右图为根据已有垛型转化的等高线图，并标识有对应区域内的高度以表达该区域内码垛箱子的堆积情况。其中，未标识高度的区域表示为未码放箱子的区域，其高度为0。
部分实施例中，等高线图中的拐点坐标即为等高线延伸方向发生转变的一点的坐标。请结合图12，在一个实施例中，一个或多个处理器10或第一获取单元2111根据如图12左图所示的等高线图，利用计算机几何学方法获取等高线图的各拐点坐标，即对等高线图建立坐标系，从而获取各个拐点的坐标。例如，如图12所示，以等高线图左下角为参考点建立坐标系，即可获得等高线图各拐点坐标，如图12右图所示，用虚线标识的点即为等高线图的部分拐点。从左往右，从低到高各拐点的坐标分别为(55，55)，(55，110)，(110，35)，(110，55)，(110，110)，(220，35)，(220，110)。将获取的拐点坐标作为已有垛型各角落的角落坐标。
请参阅图13，在某些实施方式中，获取已有垛型各角落的角落朝向，包括：
013：将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；及
014：将各角落相邻的预定区域内的高度不高于对应角落的角落高度的方向作为角落的角落朝向。
较佳的，步骤013、014可以帮助从角落坐标的各个方向中筛选掉部分方向而确定角落朝向，进而帮助减小模拟放置的运算量。
请结合图3，一个或多个处理器10可以用于执行013和014中的方法，即一个或多个处理器10可用于将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；及将各角落相邻的预定区域内的高度不高于对应角落的角落高度的方向作为角落的角落朝向。
请结合图5，转化单元2110还可用于执行013中的方法，第一获取单元2111还用于执行014 中的方法。即，转化单元2110还可用于将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度。第一获取单元2111还可用于将各角落相邻的预定区域内的高度不高于对应角落的角落高度的方向作为角落的角落朝向。
可选地，预定区域的尺寸可随机确定；较佳地可以依据待码放箱子的尺寸确定；具体可以将待码放箱子尺寸的平均数值作为预定区域的尺寸；或者，如果待码放箱子只有一个，可以将该箱子尺寸作为预定区域尺寸。较佳实施例中，预定区域的形状为矩形。
具体地，如图11所示，一个或多个处理器10或转化单元2110将已有垛型转化为等高线图并标识有对应区域的高度以表达该区域内码垛箱子的堆积情况。一个或多个处理器10或第一获取单元2111通过对应区域标识的高度即可得到等高线图各个拐点的高度，并将其作为已有垛型各角落的角落高度。一个或多个处理器10或第一获取单元2111通过对比角落高度与该角落相邻的预定区域高度。即可将不高于该角落高度的方向作为该角落的角落朝向。如图11右图所示，A点的角落高度包括0厘米、55厘米、75厘米和110厘米。在一个实施例中，当A点的角落高度为0厘米时，该角落坐标的东向至南向内各方向均满足预定区域内不高于角落高度，即预定区域内朝向0厘米区域内的各方向不高于0厘米，则角落朝向可以从东至南向区域中选取，具体可以包括东、东南、南。当A点的角落高度为55厘米时，则0厘米区域、55厘米区域不高于55厘米，则A点的角落朝向可以为东、东南、南、西南、南。当A点的角落高度为75厘米时，则0厘米区域不高于75厘米，55厘米区域不高于75厘米以及，75厘米区域不高于75厘米，则A点的角落朝向可以从上述区域中任何一区域中选取，可以为北、东北、东、东南、南、西南、西。当A点的角落高度为110厘米时，则0厘米、55厘米、75厘米、110厘米四个区域都不高于于110厘米，则A点的角落朝向可以为任一朝向。
示例性地如图14所示，P点的角落高度包括40厘米和0厘米，其中0厘米区域包括已有垛型中码放好的箱子之间的缝隙、以及40厘米箱子右侧未放置箱子区域。对于40厘米的角落高度，可以判断0厘米区域以及、40厘米区域两个区域内的方向均满足预定区域内不高于P点的角落高度40厘米，即P点的角落朝向可以在0厘米区域以及、40厘米区域内选取，部分实施例中可以包括西、西北、北、东北、东、东南、南，由于P点西南方向的邻域(预定区域)内存在高于角落高度40厘米的50厘米箱子，进而西南方向不能被确定为角落朝向。对于0厘米的角落高度，可以判断北、东北、东、东南、南为角落朝向，而西南、西北两个朝向由于存在高度分别为40厘米、50厘米的箱子，而不能被选为角落朝向；西向为两箱子的缝隙，尽管不高于角落高度0厘米，但该方向不符合部分较佳实施例中“预定区域内”的相关要求，进而不能被选为角落朝向。
示例性地如图14所示，Q点的角落高度包括0厘米和50厘米。对于0厘米的高度，0厘米区域不高于该数值，但由于0厘米的区域为两箱子间的窄小缝隙不能达到上述预定区域的要求，因此该角落高度不存在角落朝向。对于50厘米的角落高度，50厘米区域内方向满足预定区域内 不低于该高度的要求，即Q点的角落朝向可以包括50厘米区域内的任一朝向，可选地为南、东南、东、东北、北。
需要说明的是，在部分实施中，垛体码放区域的边界所在方向不作为角落朝向，在该方向无法码放箱子。示例性地如图14所示，M点的角落高度包括0厘米和30厘米。对于0厘米的角落高度，其西南、西、西北、北、东北、东、东南朝向均满足预定区域内高度不高于角落高度的要求。对于30厘米的角落高度，其各方向均满足角落朝向的要求。
请参阅图15，在某些实施方式中，获取已有垛型各角落的角落朝向，还包括：
015：将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；
016：模拟相差90°的两个向量围绕各角落进行摆放；及
017：当两个向量本身所处高度与所对应角落的角落高度相同，且两个向量所相差的90°之间的预设区域的高度不高于对应角落的角落高度时，将两个所述向量的摆放姿态作为对应角落的角落朝向。
较佳的，步骤015、016、017可以帮助从角落坐标的各个方向中选取具有较大可能性能稳定放置箱子的方向，进而帮助提高原始控制空间中的有效控制空间。
可选地，预设区域的尺寸可随机确定；较佳地可以依据待码放箱子的尺寸确定；具体可以将待码放箱子尺寸的平均数值作为预设区域的尺寸；或者，如果待码放箱子只有一个，可以将该箱子尺寸作为预设区域尺寸。较佳实施例中，预设区域的形状为矩形。
请结合图3，一个或多个处理器10可以用于执行015、016和017中的方法，即一个或多个处理器10可用于将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；模拟相差90°的两个向量围绕各角落进行摆放；当两个向量本身所处高度与所对应角落的角落高度相同，且两个向量所相差的90°之间的预设区域的高度不高于对应角落的角落高度时，将两个所述向量的摆放姿态作为对应角落的角落朝向。
请结合图5，第一获取模块211还可包括模拟单元2112。其中，转化单元2110还用于执行015中的方法，模拟单元2112还用于执行016中的方法，第一获取单元2111还用于执行017中的方法。即转化单元2110用于将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度，模拟单元2112用于模拟相差90°的两个向量围绕各角落进行摆放，第一获取单元2111用于当两个向量本身所处高度与所对应角落的角落高度相同，且两个向量所相差的90°之间的预设区域的高度不高于对应角落的角落高度时，将两个所述向量的摆放姿态作为对应角落的角落朝向。
请结合图16，B点为所确定的角落坐标，其角落高度分别为0、55、75、110，其中的xy轴为相差90°的两个向量。可选的，可将x轴分别沿东、西、南、北、东南、东北、西南、西北八 个方向放置。示例性地，x轴朝向东相应地y轴朝向北：若xy轴均落在75的高度区域，则xy本身所处高度均为75，所对应的角落高度均为75，且，xy之间预设区域内的高度均不高于75，进而当前xy方向为角落朝向；若由于细微偏差y轴落在110高度区域的边缘、x轴仍在75高度区域，则当前x本身所处高度为75对应角落高度为75，而y本身所处高度为110其对应角落高度与x轴相异，进而当前xy方向不能被作为角落朝向；若由于细微差别x轴落在0高度区域边缘、y轴仍落在75高度区域内，则当前x本身所处高度为0对应角落高度为0，而y轴本身所处高度为75对应角落高度为75与x所对应的角落高度相异，进而当前xy方向不能被作为角落朝向。再一次示例性地，若x轴朝向东北相应地y轴朝向西北，此时x轴必然落在75高度区域而y轴必然落在110高度区域，x轴所对应的角落高度为75而y轴所对应的角落高度为110，两者所对应角落高度不同，进而当前xy方向不能被作为角落朝向。又一示例性地，若此时75高度区域内存在缝隙，当x轴朝向东相应地y轴朝向北且xy轴均落在75的高度区域时，xy间预设区域内的高度未高出75，则可将当前xy方向作为角落朝向。另一示例性地，若此时75高度区域内存在突起障碍物，当x轴朝向东相应地y轴朝向北且xy轴均落在75的高度区域时，如果前述障碍物使得xy间预设区域内的高度高出75，当前xy方向不能作为角落朝向；如果前述障碍物未在xy间预设区域内，进而xy间预设区域内的高度未高出75，则可将当前xy方向作为角落朝向。
本申请实施方式的多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备100、多尺寸箱子的混合码垛装置200通过对待码箱子在同一角落坐标进行模拟摆放，进一步确定该角落坐标可以进行摆放的角落朝向。从而完善了待码箱子可以摆放的多种姿态，从而提高了对多尺寸箱子的混合码垛设备100或多尺寸箱子的混合码垛装置200的码垛的准确性。
请参阅图17，在某些实施方式中，04：根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间，包括041：移除掉模拟中无法码放的控制、码放不平稳的控制、及不存在有效填充轨迹的控制，以得到有效控制空间。
请结合图3，一个或多个处理器10可用于执行041中的方法，即一个或多个处理器10还用于移除掉模拟中码放不平稳的控制、及不存在有效填充轨迹的控制，以得到有效控制空间。
请结合图5，第三获取模块214还包括第二获取单元2140。第二获取单元2140用于执行041中的方法，即第二获取单元2140用于移除掉模拟中无法码放的控制、码放不平稳的控制、及不存在有效填充轨迹的控制，以得到有效控制空间。
在一个实施例中，一个或多个处理器10或第二获取单元2140基于模型对待码箱子的多种控制集合形成的原始控制空间的各控制进行模拟放置。通过对模拟放置的结果的筛选从而得到有效控制空间。假设原始控制空间具有5个控制，依次模拟放置待码箱子在这5个控制中，在一个例子中，若模拟放置待码箱子到第1-4个控制中时，该待码箱子都能填充进该第1-4个控制中，而模拟放置待码箱子到第5个控制中时，该待码箱子不能填充进该第5个控制中，例如该控制尺寸 不符合而不能码放相应箱子，则一个或多个处理器10或第二获取单元2140将会移除原始控制空间中的该第5个控制。同时，在模拟过程中，待码箱子放置于原始控制空间后，却不能平稳放置，产生倾斜角度，则一个或多个处理器10或第二获取单元2140将会移除原始控制空间中的该控制，例如，若模拟放置待码箱子到第1-4个控制中时，尽管该待码箱子都能填充进该第1-4个控制中，但该待码箱子填充进该第1个控制中后，待码箱子却不能平稳放置，具有一定倾斜角度，容易造成垮塌，则一个或多个处理器10或第二获取单元2140也会移除原始控制空间中的该第1个控制。若一个实施例中，在模拟放置过程中，待码箱子可以放置于该原始空间内，且待码箱子可以平稳放置，此时，需考虑实际放置过程中，是否能够将箱子由当前位置移动至该控制空间所在位置，若机器人在移动过程中存在运动学奇异点、或者若在箱子当前位置与控制空间所在位置之间存在不可跨越的障碍，则代表该原始控制空间不存在有效填充轨迹，则一个或多个处理器10或第二获取单元2140也会移除原始控制空间中的该控制，例如，若模拟放置待码箱子到第2-4个控制中时，尽管该待码箱子都能按照预定姿态填充进该第2-4个控制中，且该待码箱子填充进该第2-4个控制中后，待码箱子均能平稳放置，没有倾斜角度，不容易造成垮塌，但若不存放置该待码箱子填充进该第2个控制的路线时，一个或多个处理器10或第二获取单元2140也会移除原始控制空间中的该第2个控制。由此，经过移除掉模拟中不能被填充而导致箱子无法码放的控制、码放不平稳的控制、及不存在有效填充轨迹的控制后，得到有效控制空间仅包括第2及第3个控制。
本申请实施方式的多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备100、多尺寸箱子的混合码垛装置200通过一个或多个处理器10或第二获取单元2140模拟移除不可行的控制以得到有效控制空间。提前排除了实际码垛过程中的多种不适合码放的可能性，从而提高了对多尺寸箱子的混合码垛设备100的码垛效率。
请参阅图18，在某些实施方式中，05：根据模拟获取有效控制空间中的最优控制，包括：
051：获取执行模拟后待码箱子的第一参数集与第一预设参数集之间的第一匹配度，第一参数集包括待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、距离边缘或指定容器角的距离中的至少一个；
053：获取执行模拟后形成的新垛型的第二参数集与第二预设参数集之间的第二匹配度，第二参数集包括所述新垛型的高度、容积率、顶面不平整度中的至少一个；及
055：根据第一匹配度、第二匹配度、第一匹配度预设的第一权重、及第二匹配度预设的第二权重获取各控制的总匹配度；
057：将多个控制中总匹配度最高的控制作为最优控制。
请结合图3，一个或多个处理器10可用于执行051、053、055和057中的方法。即一个或多个处理器10可用于获取执行模拟后待码箱子的第一参数集与第一预设参数集之间的第一匹配度，第一参数集包括待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、 距离边缘或指定容器角的距离中的至少一个；获取执行模拟后形成的新垛型的第二参数集与第二预设参数集之间的第二匹配度，第二参数集包括所述新垛型的高度、容积率、顶面不平整度中的至少一个；及根据第一匹配度、第二匹配度、第一匹配度预设的第一权重、及第二匹配度预设的第二权重获取各控制的总匹配度；将多个控制中总匹配度最高的控制作为最优控制。
请结合图5，第四获取模块215还包括第三获取单元2150和第一控制单元2151。第三获取单元2140用于执行051、053和055中的方法，第一控制单元2151用于执行057中的方法。即第三获取单元2150用于获取执行模拟后待码箱子的第一参数集与第一预设参数集之间的第一匹配度，第一参数集包括待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、距离边缘或指定容器角的距离中的至少一个；获取执行模拟后形成的新垛型的第二参数集与第二预设参数集之间的第二匹配度，第二参数集包括所述新垛型的高度、容积率、顶面不平整度中的至少一个；及根据第一匹配度、第二匹配度、第一匹配度预设的第一权重、及第二匹配度预设的第二权重获取各控制的总匹配度。第一控制单元2151用于将多个控制中总匹配度最高的控制作为最优控制。
具体地，一个或多个处理器100或第三获取单元2150在对待码箱子进行模拟放置前，会对该待码箱子设有第一预设参数集，以限定该待码箱子在已有垛型中的理想放置情况。再对待码箱子进行模拟放置后，则获取对待码箱子的模拟放置情况的第一参数集。第一参数集包括待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、距离边缘或指定容器角的距离等。一个或多个处理器10或第三获取单元2150通过第一预设参数集与第一参数集的对比得到第一匹配度。
一个或多个处理器100或第三获取单元2150在待码箱子进行模拟放置前，还会对放置后的新垛型设有第二预设参数集，以限定新垛型的理性放置情况，再对所有待码箱子进行模拟放置完成后，则获取当前垛型的第二参数集。第二参数集包括所述新垛型的高度、容积率、顶面不平整度等。一个或多个处理器10或第三获取单元2150通过第二预设参数集与第二参数集的对比得到第二匹配度。
一个或多个处理器10或第三获取单元2150还会根据第一匹配度、第二匹配度、第一匹配度预设的第一权重、及第二匹配度预设的第二权重获取各控制的总匹配度。例如，一个或多个处理器10对待码箱子的多种控制集合形成的原始控制空间的各控制进行模拟放置，在对第2个控制进行模拟放置后，第一匹配度为90％，第二匹配度为95％，第一匹配度预设的第一权重为70％，第二匹配度预设的第二权重为30％，则可得到该控制的总匹配度为91.5％。在同一控制空间中的第3控制进行模拟码放后，第一匹配度为95％，第二匹配度为90％，第一匹配度预设的第一权重为70％，第二匹配度预设的第二权重为30％，则可得到该控制的总匹配度为93.5％。则一个或多个处理器1或第一控制单元2151会选取该控制的总匹配度为93.5％的，即总匹配度最高的控制(第 3控制)作为最优控制。
本申请实施方式的多尺寸箱子的混合码垛方法、多尺寸箱子的混合码垛设备100和多尺寸箱子的混合码垛装置200通过一个或多个处理器10或第三获取单元2150模拟有效控制空间获取最优控制空间。提前过滤了实际码垛过程中匹配度较低的码垛方式，从而提高了对多尺寸箱子的混合码垛设备100的码垛效率。
请参阅图19，本申请实施方式提供的多尺寸箱子的混合码垛方法，还包括：
08：当有效控制空间为空集时，停止执行码垛。
请结合图3，一个或多个处理器10可用于执行08中的方法，即一个或多个处理器10可用于当有效控制空间为空集时，停止执行码垛。
请结合图5，控制模块216还包括第二控制单元2160。第二控制单元2160用于执行08中的方法，即第二控制单元2160用于当有效控制空间为空集时，停止执行码垛。
具体地，在一个实施例中，所有待码箱子都被放置成功，则代表有效控制空间为空集，即码垛任务完成，此时停止执行码垛。在另一个实施例中，待码箱子没有被完全放置，有剩余，而当前垛型没有放置空间时，则代表有效控制空间为空集，此时，停止执行码垛。
请参阅图20，本申请实施方式还提供一种包含计算机程序301的非易失性计算机可读存储介质300。当计算机程序301被一个或多个处理器10执行时，使得一个或多个处理器10执行上述任一实施方式的多尺寸箱子的混合码垛方法。
例如，计算机程序301被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：01：获取已有垛型各角落的角落坐标及角落朝向，沿角落朝向存在平面能够放置待码箱子；02：将各角落的角落坐标及角落朝向与待码箱子的至少两种放置方式进行组合以得到多个控制，多个控制的集合形成原始控制空间；03：模拟放置待码箱子至原始控制空间的各控制中；04：根据模拟移除原始控制空间中不可行的控制以得到有效控制空间；05：根据模拟获取有效控制空间中的最优控制；及06：将待码箱子放置在最优控制中。
例如，计算机程序301被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：07：对码垛进行拍照并建立模型以获取已有垛型。071：每一次码垛均拍摄码垛的当前垛型并建立整体模型以获取已有垛型；或072：拍摄首次码垛的首次垛型，基于首次垛型及后续码垛过程中所码放箱子的位置、高度以获取已有垛型；或073：每一次码垛均拍摄码垛的当前垛型，并只建立垛型的最高层箱子的平面的模型，以获取已有垛型；或074：拍摄首次码垛的首次垛型，后续码垛每码放一次，在首次垛型的相应位置上增加所码放的箱子，并基于所拍摄的最高层箱子的平面模型对所增加箱子进行矫正，以获取已有垛型。
例如，计算机程序301被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：
011：将已有垛型转化为等高线图；及012：采用基于等高线图的计算几何学方法获取等高线图的各拐点坐标，以作为已有垛型各角落的角落坐标。013：将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；及014：将各角落相邻的预定区域内的高度不高于对应角落的角落高度的方向作为角落的角落朝向。015：将已有垛型转化为等高线图，并将等高线图的各拐点高度作为已有垛型各角落的角落高度；016：模拟相差90°的两个向量围绕各角落进行摆放；及017：当两个向量本身所处高度与所对应角落的角落高度相同，且两个向量所相差的90°之间的预设区域的高度不高于对应角落的角落高度时，将两个所述向量的摆放姿态作为对应角落的角落朝向。
例如，计算机程序201被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：041：移除掉模拟中无法码放的控制、码放不平稳的控制、及不存在有效填充轨迹的控制，以得到有效控制空间。
再例如，计算机程序301被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：
051：获取执行模拟后待码箱子的第一参数集与第一预设参数集之间的第一匹配度，第一参数集包括待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、距离边缘或指定容器角的距离中的至少一个；053：获取执行模拟后形成的新垛型的第二参数集与第二预设参数集之间的第二匹配度，第二参数集包括所述新垛型的高度、容积率、顶面不平整度中的至少一个；及055：根据第一匹配度、第二匹配度、第一匹配度预设的第一权重、及第二匹配度预设的第二权重获取各控制的总匹配度；057：将多个控制中总匹配度最高的控制作为最优控制。
还例如，计算机程序301被一个或多个处理器10执行时，使得处理器10执行以下多尺寸箱子的混合码垛方法：08：当有效控制空间为空集时，停止执行码垛。
在本说明书的描述中，参考术语“某些实施方式”、“一个例子中”、“示例地”等的描述意指结合所述实施方式或示例描述的具体特征、结构、材料或者特点包含于本申请的至少一个实施方式或示例中。在本说明书中，对上述术语的示意性表述不一定指的是相同的实施方式或示例。而且，描述的具体特征、结构、材料或者特点可以在任何的一个或多个实施方式或示例中以合适的方式结合。此外，在不相互矛盾的情况下，本领域的技术人员可以将本说明书中描述的不同实施例或示例以及不同实施例或示例的特征进行结合和组合。
流程图中或在此以其他方式描述的任何过程或方法描述可以被理解为，表示包括一个或更多个用于实现特定逻辑功能或过程的步骤的可执行指令的待码的模块、片段或部分，并且本申请的优选实施方式的范围包括另外的实现，其中可以不按所示出或讨论的顺序，包括根据所涉及的功能按基本同时的方式或按相反的顺序，来执行功能，这应被本申请的实施例所属技术领域的技术人员所理解。
尽管上面已经示出和描述了本申请的实施方式，可以理解的是，上述实施方式是示例性的，不能理解为对本申请的限制，本领域的普通技术人员在本申请的范围内可以对上述实施方式进行变化、修改、替换和变型。

Claims (12)

  1. 一种多尺寸箱子的混合码垛方法，其特征在于，包括：
    获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；
    将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；
    模拟放置所述待码箱子至所述原始控制空间的各控制中；
    根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；
    根据模拟获取所述有效控制空间中的最优控制；及
    将所述待码箱子放置在所述最优控制中。
  2. 根据权利要求1所述混合码垛方法，其特征在于，还包括：
    对码垛进行拍照并建立模型以获取所述已有垛型。
  3. 根据权利要求2所述混合码垛方法，其特征在于，所述对码垛进行拍照并建立模型以获取所述已有垛型，包括：
    每一次码垛均拍摄码垛的当前垛型并建立整体模型以获取所述已有垛型；或
    拍摄首次码垛的首次垛型，基于所述首次垛型及后续码垛过程中所码放箱子的位置、高度以获取所述已有垛型；或
    每一次码垛均拍摄码垛的当前垛型，并只建立垛型的最高层箱子的平面的模型，以获取所述已有垛型；或
    拍摄首次码垛的首次垛型，后续码垛每码放一次，在首次垛型的相应位置上增加所码放的箱子，并基于所拍摄的最高层箱子的平面模型对所增加箱子进行矫正，以获取所述已有垛型。
  4. 根据权利要求1所述混合码垛方法，其特征在于，所述获取已有垛型各角落的角落坐标，包括：
    将所述已有垛型转化为等高线图；及
    采用基于所述等高线图的计算几何学方法获取所述等高线图的各拐点坐标，以作为所述已有垛型各角落的角落坐标。
  5. 根据权利要求1所述混合码垛方法，其特征在于，所述获取已有垛型各角落的角落朝向，包括：
    将所述已有垛型转化为等高线图，并将所述等高线图的各拐点高度作为所述已有垛型各角落的角落高度；及
    将各角落相邻的预定区域内的高度不高于对应角落的所述角落高度的方向作为所述角落的角落朝向。
  6. 根据权利要求1所述混合码垛方法，其特征在于，所述获取已有垛型各角落的角落朝向，包括：
    将所述已有垛型转化为等高线图，并将所述等高线图的各拐点高度作为所述已有垛型各角落的角落高度；
    模拟相差90°的两个向量围绕各角落进行摆放；及
    当两个所述向量本身所处高度与所对应角落的所述角落高度相同，且两个所述向量所相差的90°之间的预设区域的高度不高于对应角落的所述角落高度时，将两个所述向量的摆放姿态作为对应角落的角落朝向。
  7. 根据权利要求1所述混合码垛方法，其特征在于，所述根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间，包括：
    移除掉模拟中无法码放的所述控制、码放不平稳的所述控制、及不存在有效填充轨迹的所述控制，以得到所述有效控制空间。
  8. 根据权利要求1至7任意一项所述混合码垛方法，其特征在于，所述根据模拟获取所述有效控制空间中的最优控制，包括：
    获取执行所述模拟后所述待码箱子的第一参数集与第一预设参数集之间的第一匹配度，所述第一参数集包括所述待码箱子与临近箱子的接触面积、支撑面面积、支撑面百分比、支撑箱子个数、距离边缘或指定容器角的距离中的至少一个；
    获取执行所述模拟后形成的新垛型的第二参数集与第二预设参数集之间的第二匹配度，所述第二参数集包括所述新垛型的高度、容积率、顶面不平整度中的至少一个；
    根据所述第一匹配度、所述第二匹配度、所述第一匹配度预设的第一权重、及所述第二匹配度预设的第二权重获取各控制的总匹配度；及
    将多个所述控制中所述总匹配度最高的所述控制作为最优控制。
  9. 根据权利要求1所述混合码垛方法，其特征在于，还包括：
    当所述有效控制空间为空集时，停止执行码垛。
  10. 一种多尺寸箱子的混合码垛设备，其特征在于，包括一个或多个处理器，一个或多个所述处理器用于：
    获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；
    将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；
    模拟放置所述待码箱子至所述原始控制空间的各控制中；
    根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；
    根据模拟获取所述有效控制空间中的最优控制；及
    控制将所述待码箱子放置在所述最优控制中。
  11. 一种多尺寸箱子的混合码垛装置，其特征在于，包括：
    第一获取模块，所述第一获取模块用于获取已有垛型各角落的角落坐标及角落朝向，沿所述角落朝向存在平面能够放置待码箱子；
    第二获取模块，所述第二获取模块用于将各角落的所述角落坐标及所述角落朝向与所述待码箱子的至少两种放置方式进行组合以得到多个控制，多个所述控制的集合形成原始控制空间；
    模拟模块，所述模拟模块用于模拟放置所述待码箱子至所述原始控制空间的各控制中；
    第三获取模块，所述第三获取模块用于根据模拟移除所述原始控制空间中不可行的控制以得到有效控制空间；
    第四获取模块，所述第四获取模块用于根据模拟获取所述有效控制空间中的最优控制；及
    控制模块，所述控制模块用于控制将所述待码箱子放置在所述最优控制中。
  12. 一个或多个存储有计算机程序的非易失性计算机可读存储介质，当所述计算机程序被一个或多个处理器执行时，实现权利要求1至9任意一项所述的混合码垛方法。
PCT/CN2020/127255 2020-11-03 2020-11-06 多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质 WO2022094968A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011210486.3 2020-11-03
CN202011210486.3A CN112499276B (zh) 2020-11-03 2020-11-03 多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质

Publications (1)

Publication Number Publication Date
WO2022094968A1 true WO2022094968A1 (zh) 2022-05-12

Family

ID=74955470

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/127255 WO2022094968A1 (zh) 2020-11-03 2020-11-06 多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质

Country Status (2)

Country Link
CN (2) CN112499276B (zh)
WO (1) WO2022094968A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115180416A (zh) * 2022-07-08 2022-10-14 深圳蓝胖子机器智能有限公司 一种码垛方法、码垛设备及存储介质

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113284178B (zh) * 2021-06-11 2024-04-09 梅卡曼德（北京）机器人科技有限公司 物体码垛方法、装置、计算设备及计算机存储介质
CN113351522B (zh) * 2021-06-11 2023-01-31 梅卡曼德（北京）机器人科技有限公司 物品分拣方法、装置及系统
CN113895728B (zh) * 2021-09-30 2023-03-21 合肥辰视机器人科技有限公司 贪心码垛方法、装置及计算机可读存储介质
CN114529155A (zh) * 2022-01-17 2022-05-24 湖南视比特机器人有限公司 工件动态码垛装框方法及系统
CN114313892A (zh) * 2022-01-26 2022-04-12 武汉科技大学 多种类直型耐火砖自动化码垛方法及装置
CN114692237B (zh) * 2022-06-02 2022-08-26 杭州灵西机器人智能科技有限公司 基于多模态评估指标的混合码垛方法、装置、设备及介质
CN114862063B (zh) * 2022-07-04 2022-09-30 杭州灵西机器人智能科技有限公司 一种在线混合码垛的方法、系统、电子装置和存储介质
CN116835334A (zh) * 2023-07-17 2023-10-03 湖北普罗格科技股份有限公司 基于3d视觉的无序码垛方法、装置、介质及设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4641271A (en) * 1983-11-09 1987-02-03 Hitachi, Ltd. Piling planning method and piling system of cargoes by palletizing robot
CN108750685A (zh) * 2018-04-28 2018-11-06 武汉库柏特科技有限公司 一种离线混合码垛方法及系统
CN109230580A (zh) * 2018-10-11 2019-01-18 西安中科光电精密工程有限公司 一种基于混放物料信息采集的拆垛机器人系统及方法
US20190352107A1 (en) * 2016-12-30 2019-11-21 Gebo Cermex Canada Inc. Robotic palletizing
CN110723553A (zh) * 2019-10-23 2020-01-24 安吉智能物联技术有限公司 一种基于搬运机器人用的垛型生成方法

Family Cites Families (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04303328A (ja) * 1991-03-30 1992-10-27 Toshiba Corp カートン配置決定方法
JP3610592B2 (ja) * 1993-11-30 2005-01-12 凸版印刷株式会社 刷本積載装置制御方法およびそのシステム
KR100382266B1 (ko) * 2000-08-18 2003-05-01 이용중 로봇 자세 보정 알고리즘을 이용한 크기 가변작업대상물의 자동 상표 부착방법
JP4275679B2 (ja) * 2006-04-28 2009-06-10 株式会社インテージ 荷積みプラン作成方法及びそのプログラム
FR2959731A1 (fr) * 2010-05-04 2011-11-11 Sidel Participations Dispositif d'orientation d'emballages a l'entree d'un palettiseur
JP2013082518A (ja) * 2011-10-07 2013-05-09 Toyota Motor Corp 積み付け支援装置
CN108275294B (zh) * 2017-01-06 2020-06-05 北京京东尚科信息技术有限公司 码垛方法、码垛装置、计算机可读存储介质及电子设备
CN108275293B (zh) * 2017-01-06 2020-06-05 北京京东尚科信息技术有限公司 码垛方法、码垛装置、计算机可读存储介质及电子设备
CN106829469A (zh) * 2017-03-30 2017-06-13 武汉库柏特科技有限公司 一种基于双相机的机器人无序抓取装置及方法
CN107671896B (zh) * 2017-05-19 2020-11-06 重庆誉鸣科技有限公司 基于scara机器人的快速视觉定位方法及系统
CN107720237B (zh) * 2017-08-31 2020-02-07 广州泰行智能科技有限公司 一种基于空间坐标系的垛件排样方法和装置
CN107479581B (zh) * 2017-09-21 2021-01-26 北京京东尚科信息技术有限公司 货箱码放控制装置、系统、方法和存储介质
CN110222862A (zh) * 2018-03-02 2019-09-10 北京京东尚科信息技术有限公司 码垛方法和装置
CN110371694B (zh) * 2018-07-24 2021-07-30 北京京东乾石科技有限公司 用于将箱体混合码垛的方法和系统
CN109353833B (zh) * 2018-11-27 2021-06-29 深圳市汇川技术股份有限公司 机器人码垛点位生成方法、设备及计算机可读存储器
CN109592433B (zh) * 2018-11-29 2021-08-10 合肥泰禾智能科技集团股份有限公司 一种货物拆垛方法、装置及拆垛系统
CN109384042B (zh) * 2018-11-29 2020-07-21 湖南视比特机器人有限公司 序列增量式码垛方法及其装置和计算机可读存储介质
CN111410045A (zh) * 2019-01-08 2020-07-14 北京京东尚科信息技术有限公司 货箱搬运方法和装置
CN111788134B (zh) * 2019-02-14 2022-02-18 深圳配天智能技术研究院有限公司 自动码垛的方法、设备以及存储装置
CN110348792A (zh) * 2019-07-15 2019-10-18 珠海格力智能装备有限公司 码垛的配置方法及装置、码垛的方法及装置
CN110422521B (zh) * 2019-07-17 2021-06-01 上海新时达机器人有限公司 不规则随机物料的机器人侧边拆垛方法及其装置
CN110815230B (zh) * 2019-11-21 2022-02-22 广东力生智能有限公司 一种基于虚拟采样的六轴码垛机器人位姿适配方法
CN111099363B (zh) * 2020-01-09 2021-10-22 湖南视比特机器人有限公司 码垛方法、码垛系统和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4641271A (en) * 1983-11-09 1987-02-03 Hitachi, Ltd. Piling planning method and piling system of cargoes by palletizing robot
US20190352107A1 (en) * 2016-12-30 2019-11-21 Gebo Cermex Canada Inc. Robotic palletizing
CN108750685A (zh) * 2018-04-28 2018-11-06 武汉库柏特科技有限公司 一种离线混合码垛方法及系统
CN109230580A (zh) * 2018-10-11 2019-01-18 西安中科光电精密工程有限公司 一种基于混放物料信息采集的拆垛机器人系统及方法
CN110723553A (zh) * 2019-10-23 2020-01-24 安吉智能物联技术有限公司 一种基于搬运机器人用的垛型生成方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115180416A (zh) * 2022-07-08 2022-10-14 深圳蓝胖子机器智能有限公司 一种码垛方法、码垛设备及存储介质
CN115180416B (zh) * 2022-07-08 2024-05-28 深圳蓝胖子机器智能有限公司 一种码垛方法、码垛设备及存储介质

Also Published As

Publication number Publication date
CN112499276A (zh) 2021-03-16
CN113651118B (zh) 2023-02-10
CN112499276B (zh) 2021-10-29
CN113651118A (zh) 2021-11-16

Similar Documents

Publication Publication Date Title
WO2022094968A1 (zh) 多尺寸箱子的混合码垛方法、设备、装置及计算机可读存储介质
JP7429384B2 (ja) リアルタイム載置シミュレーションを使用してパッケージをパレタイズするためのロボットシステム
JP7454148B2 (ja) パッキング機構を有するロボットシステム
US11383380B2 (en) Object pickup strategies for a robotic device
JP6704157B1 (ja) 動的パッキング機構を有するロボットシステム
JP6729953B1 (ja) エラー検出および動的パッキング機構を有するロボットシステム
JP6771799B1 (ja) 壁に基づくパッキング機構を有するロボットシステム及びその動作方法
JP6813229B1 (ja) 自動物体検出機構を備えたロボットシステム、および、その動作方法
JP7023450B2 (ja) カメラによって生成される画像情報に基づいて、動作計画を実行する方法および計算システム
JP7433609B2 (ja) 物体識別のための方法および計算システム
WO2021243924A1 (zh) 托盘空间规划方法、码垛方法及其装置、及电子设备
JP2017520418A (ja) 軌道計画のための物体測定基準のリアルタイム確定
Thangavelu et al. Dry stacking for automated construction with irregular objects
CN109901627B (zh) 一种无人机的着陆位姿调整方法、系统及相关组件
JP6678984B1 (ja) カメラキャリブレーションの姿勢を決定するための方法及びシステム
JP5544464B2 (ja) 対象物の３次元位置・姿勢認識装置及びその方法
CN111131813B (zh) 用于确定相机校准的姿态的方法、系统和介质
Yun et al. Adaptive coordinating construction of truss structures using distributed equal-mass partitioning
CN117011362A (zh) 货物体积的计算方法、容积率的动态计算方法
CN109697757A (zh) 一种三维模型贴地方法、装置、系统及可读存储介质
TWI757826B (zh) 自動置物方法及系統
CN116342858B (zh) 物体检测方法、装置、电子设备及存储介质
RU2681471C1 (ru) Способ перемещения тарно-штучных грузов на складе
CN116423525A (zh) 一种自动手眼标定方法及系统、存储介质
Ma et al. Using planar features for fast localization in indoor environments

Legal Events

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

Ref document number: 20960429

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 11/08/2023)

122 Ep: pct application non-entry in european phase

Ref document number: 20960429

Country of ref document: EP

Kind code of ref document: A1