A robot stacking planning method based on negative multinomial distribution

Abstract

The application provides a robot stacking planning method based on a negative multinomial distribution, comprising the following steps: extracting feature parameters of workpiece stacking, and constructing a workpiece arrival probability model; combining actual requirements to construct a target function, and forming an optimization model of workpiece arrival probability; solving the optimization model to determine an optimal solution of an inequality system; and adding a stacking basket and an RGV trolley according to the optimal solution of the inequality system. The application can economically realize robot stacking and basket loading of multiple workpieces under existing production conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automation production technology, in particular to a robot stacking planning method based on negative multinomial distribution. BACKGROUND

[0002] In many production and manufacturing enterprises, the products produced are mostly assembled by different specifications and different types of parts. In order to efficiently assemble the products, not only the various parts should be produced in proportion, but also the robot should intelligently identify and classify and stack according to the specifications and models.

[0003] Taking a project in the current coating industry as an example, the general process of robot grabbing workpieces to pack and stack includes: workpieces are conveyed at a conveying pace of 18s / piece; every 18s, a workpiece is placed on a plate chain conveyor; the robot moves along the slide rail, grabs the workpiece from the plate chain conveyor, and stacks it into baskets 1-5 (which can be increased to 6 baskets); the robot works at a pace of 18s / piece, and each packing basket can only pack and stack workpieces of the same specification and model; after the packing basket is filled with 20 workpieces, the RGV car (which can be increased) automatically drags the packing basket to the packing station, and completes the packing within 360s, and then drags the empty basket back to the stacking station by the RGV car.

[0004] Under the existing production conditions, the production line can only smoothly run the stacking and packing of two types of workpieces. However, in fact, there are thousands of types of workpieces, and if only two types of workpieces are produced, it will inevitably seriously restrict the production scheduling of other links, and thus affect the overall production efficiency.

[0005] The inventor has tried to run the stacking and packing of three types of workpieces, but the stacking basket turnover is not timely, resulting in system error and unable to run normally. Moreover, due to the limitation of the size of the site, the production space cannot be expanded; due to the limitation of the walking time of the robot, more stacking baskets cannot be added; considering these factors, there is an urgent need for a method that can utilize the existing production conditions to improve the stacking efficiency. SUMMARY

[0006] In order to solve the problems in the prior art, the present application provides a robot stacking planning method based on negative multinomial distribution, which can economically realize the robot stacking and packing of multiple types of workpieces using the existing production conditions.

[0007] To achieve the above-mentioned purpose, the present application provides a robot stacking planning method based on negative multinomial distribution, comprising the following steps:

[0008] Extracting the characteristic parameters of workpiece stacking, constructing a probability model of workpiece arrival;

[0009] Combining the actual demand to construct a target function to form an optimization model of workpiece arrival probability;

[0010] solving the optimization model to determine the optimal solution of the inequality system;

[0011] According to the optimal solution of the inequality system, the stacking basket and the RGV trolley are increased.

[0012] Further, the characteristic parameters of the workpiece stacking include workpiece name, rated load number of the stacking basket, and workpiece conveying beat, wherein,

[0013] The workpiece name is used to mark different workpieces to be stacked;

[0014] The rated load number of the stacking basket refers to the standard number of the same workpieces that can be loaded by the stacking basket;

[0015] The workpiece conveying beat refers to the time interval of the arrival of the workpieces.

[0016] Further, the probability model of the workpiece arrival refers to the prediction model of the arrival of different workpieces, and the probability model of the workpiece arrival is constructed by using a negative multinomial distribution according to the characteristic parameters of the workpiece stacking, and includes the probability model of the arrival of each of the plurality of workpieces.

[0017] Further, the actual demand in the step of constructing the objective function according to the actual demand to form the optimization model of the probability of the workpiece arrival refers to the relationship requirement between the probabilities of the arrival of the workpieces.

[0018] Further, the probability model of the arrival of each of the plurality of workpieces is as follows:

[0019]

[0020] Further, the step of solving the optimization model to determine the optimal solution of the inequality system includes minimizing the highest proportion of the workpieces as the target, and using a nonlinear programming algorithm to solve the optimization model to determine the satisfactory solution of the inequality system.

[0021] To achieve the above-mentioned purpose, the application further provides an electronic device including a memory and a processor, the memory stores a program running on the processor, and the processor executes the steps of the above-mentioned robot stacking planning method based on the negative multinomial distribution when running the program.

[0022] To achieve the above-mentioned purpose, the application further provides a computer readable storage medium, which stores computer instructions, and the computer instructions execute the steps of the above-mentioned robot stacking planning method based on the negative multinomial distribution when running.

[0023] To achieve the above object, the application further provides a stacking and stacking device applying a robot stacking planning method based on negative multinomial distribution, comprising a plate chain conveyor, a stacking robot, a stacking robot track, a stacking basket, an RGV trolley, an RGV track, a packing work station,

[0024] The number of the stacking baskets and the RGV trolleys is determined according to the optimal solution determined by the robot stacking planning method based on negative multinomial distribution.

[0025] The robot stacking planning method based on negative multinomial distribution has the following beneficial effects compared with the prior art:

[0026] (1) The robot stacking of four kinds of workpieces and the basket loading can be realized, which is higher than the previous two kinds of workpieces, and the efficiency is improved;

[0027] (2) A new algorithm based on negative multinomial distribution is provided for the robot stacking planning of the type of stacking workpieces and the number of stacking baskets, and the calculation problem of enterprises is solved;

[0028] (3) The nonlinear inequality system involved in the solution of the problem is solved by using the mathematical optimization method, and the workpiece type ratio ensuring the stability of the system is obtained;

[0029] (4) A program for simulating actual operation by using Python is developed, and the accuracy of the algorithm is verified (attached later).

[0030] Other features and advantages of the application will be set forth in the following description, and in part will become apparent to those skilled in the art, or will be learned by practice of the application. BRIEF DESCRIPTION OF DRAWINGS

[0031] The accompanying drawings are used to provide a further understanding of the application, and constitute a part of the specification, and together with the embodiments of the application, are used to explain the application, and do not constitute a limitation of the application. In the drawings:

[0032] Figure 1 A flowchart of the robot stacking planning method based on negative multinomial distribution according to the application;

[0033] Figure 2 A simulation code diagram in embodiment 2 of the robot stacking planning method based on negative multinomial distribution according to the application;

[0034] Figure 3 A stacking and stacking device applying the robot stacking planning method based on negative multinomial distribution of the application. DETAILED DESCRIPTION

[0035] Embodiments of the present application will be described herein below with reference to the accompanying drawings. While certain embodiments of the present application are shown in the drawings, it is understood that the present application can be embodied in various forms and should not be construed as being limited to the embodiments set forth herein, but rather, these embodiments are provided so that the present application can be more thoroughly and completely understood. It is understood that the drawings of the present application and the embodiments are for illustrative purposes only and are not intended to limit the scope of the present application.

[0036] It is understood that the various steps recited in the method embodiments of the present application can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit performing the steps shown. The scope of the present application is not limited in this respect.

[0037] The term "comprising" and variations thereof as used herein are open-ended, that is "including, but not limited to". The term "based on" is "based, at least in part, on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Related terms are defined in the following description.

[0038] It is noted that the terms "a" or "an", as used herein, are defined as "one or more" and can be used interchangeably with the phrase "one or more". The term "or" as used herein is defined as "and / or", unless otherwise indicated.

[0039] In traditional probability, the workpiece is generally compared to a customer, and the interval of the customer's arrival generally obeys an exponential distribution, and the interval time is any real number greater than zero. In a real production line, in order to ensure that the robot can successfully grasp the workpiece, the time interval of the workpiece arrival is fixed (such as 18s in the above production line).

[0040] The workpiece is generally regarded as a customer who has arrived and accepted service, but in a production line, the workpiece must be immediately served and cannot wait, otherwise the system will report an error, so the traditional queuing theory cannot be used to solve the problem.

[0041] Therefore, a new idea is needed to consider how to ensure the smooth operation of the system.

[0042] We assume that the conveyor delivers k kinds of workpieces, and according to the production of 3000 workpieces per day, each workpiece delivered is one of the k kinds, and can only be one kind, and the kind is unknown before arrival. Because each kind is packed when there are 20 workpieces, the essence of the problem is that the 21st, 41st, 61st, 81st, … workpiece of the same kind arrives and needs an empty basket to wait for it to be loaded. This involves a negative multinomial distribution.

[0043] The technical scheme of the application is based on negative multinomial distribution, and gives a code stacking and basket loading method for multiple workpieces, and simulation test is carried out, and the result is feasible.

[0044] Hereinafter, an embodiment of the application will be described in detail with reference to the drawings.

[0045] Embodiment 1

[0046] Figure 1 For the robot code stacking planning method flowchart based on negative multinomial distribution according to the application, hereinafter, the robot code stacking planning method based on negative multinomial distribution will be described in detail. Figure 1 The robot code stacking planning method based on negative multinomial distribution according to the application will be described in detail.

[0047] In step S1, the characteristic parameters of workpiece code stacking are extracted, and a probability model of workpiece arrival is constructed.

[0048] In the embodiment of the application, the characteristic parameters of workpiece code stacking include workpiece name, code basket rated load number, and workpiece conveying cycle, wherein the workpiece name is used to mark different workpieces to be stacked, for example, workpiece A, workpiece B, workpiece C, workpiece D, etc.; the code basket rated load number refers to the standard number of the same workpieces that can be loaded in the code basket, for example, the code basket rated load number can be marked as SN, and SN is an integer; the workpiece conveying cycle refers to the time required for conveying a single workpiece, and is recorded as m / piece, for example, the conveying cycle is 18s / piece.

[0049] The probability model of workpiece arrival refers to the probability model of arrival of different workpieces.

[0050] In the flow line, since the arrival of each workpiece is random, the arrival of workpieces cannot be accurately controlled, and it is necessary to adjust the proportion of workpiece arrival by using negative multinomial distribution according to the characteristic parameters of workpiece code stacking, so as to ensure the probability model of workpiece arrival for smooth operation of the system as follows:

[0051]

[0052] For convenience of description, it is assumed that there are four kinds of existing workpieces, and the workpiece names are A, B, C, and D respectively, and the workpiece produced by each workpiece production cycle is one of A, B, C, and D, and the arrival probabilities thereof are respectively P A , P B , P C , and P D . Considering that the code basket is full of SN, it will be dragged by the RGV trolley to be packed, therefore, the arrival time of the SNnth A, B, C, and D is mainly considered, and in probability theory, the arrival time is subject to negative multinomial distribution.

[0053] Since the marginal distribution of negative multinomial distribution is negative binomial distribution, the arrival of each kind of workpiece obeys negative binomial distribution.

[0054] Take A kind of workpiece as an example: suppose the SNth A arrives, and there are N A workpieces in total, then N A is a random number, but the probability of its taking value n a is:

[0055]

[0056] The point prediction of N A is The point prediction of the actual arrival time is

[0057] The interval prediction of N A is

[0058]

[0059] The interval prediction of the actual arrival time is

[0060] Since each stacking basket can only stack workpieces of the same specification and model, it can be assumed that from the 1st workpiece, A, B, C, and D workpieces are placed into baskets 1, 2, 3, and 4 respectively, and P A >P B >P C >P D , then the process of filling the stacking baskets is:

[0061] 1. A workpiece first fills basket 1, and the next A workpiece is placed into basket 5;

[0062] 2. B workpiece fills basket 2, and to ensure normal operation of the system, the SN+1th B needs to be able to be placed into the returned basket 1, that is:

[0063]

[0064] 3. C workpiece fills basket 3, and similarly, to ensure normal operation of the system, the SN+1th C needs to be able to be placed into the returned basket 2, that is:

[0065]

[0066] 4. D workpiece fills basket 4, and similarly, to ensure normal operation of the system, the SN+1th D needs to be able to be placed into the returned basket 3, that is:

[0067]

[0068] 5. If workpiece A fills basket 5, similarly, for the system to operate normally, it needs to satisfy the condition that the 2nd*SN+1th workpiece A can be put back into basket 4, that is:

[0069]

[0070] Similarly, the further operating conditions are covered by the above four constraints.

[0071] When there are more types of workpieces, the above process can be followed to add the corresponding number of constraints.

[0072] In step S2, an objective function is constructed based on actual needs to form an optimization model for the probability of workpiece arrival.

[0073] In embodiments of the present invention, in order to determine the optimal ratio of various product models, it is necessary to solve the nonlinear inequality system composed of the above inequalities.

[0074] Since there is no solution to this system of inequalities in actual calculations, attempts to find exact analytical solutions or approximate numerical solutions using mathematical software such as Maple and Matlab are futile.

[0075] This invention solves the problem by using the system of inequalities as constraints and combining them with practical requirements. For example, P A >P B >P C >P D >0,P A +P B +P C +P D =1, where the difference between the two sides of the system of inequalities is as large as possible, construct the objective function, and form an optimization model.

[0076] In step S3, the optimization model is solved to determine the optimal solution of the system of inequalities.

[0077] In an embodiment of the present invention, p A Minimize or p D With the goal of maximizing, a nonlinear programming algorithm is used to solve the optimization model and determine the satisfactory solution of the system of inequalities.

[0078] In step S4, based on the optimal solution of the system of inequalities, additional palletizing baskets and RGV trolleys are added.

[0079] In an embodiment of the present invention, since the mathematical expectation is calculated based on probability, in order to "hedge" the deviation caused by probability, the company can increase the number of palletizing baskets and RGV carts, taking into account the condition that the company can increase the number of baskets and RGV carts.

[0080] Example 2

[0081] Below, combined with a specific production condition, the robot stacking planning method based on negative multinomial distribution of the present application is further described and explained.

[0082] In the present embodiment, the workpieces of the flow line production include A, B, C and D, the rated load of the stacking basket is 20, and the production cycle is 18s / piece.

[0083] According to step S1, the probability models of the workpiece arrivals are established as follows:

[0084]

[0085]

[0086]

[0087]

[0088] According to step S2, the objective function is constructed in combination with the actual requirements to form an optimization model of the workpiece arrival probabilities.

[0089] The actual requirements are: P A > P B > P C > P D > 0, P A + P B + P C + P D = 1, wherein the difference between both sides of the inequality group is as large as possible, forming an optimization model with the above four inequalities as the constraint conditions.

[0090] According to step S3, the optimization model is solved to determine the optimal solution of the inequality group.

[0091] Taking the minimization of p A or the maximization of p D as the goal, the optimization model is solved by using a nonlinear programming algorithm to determine the satisfactory solution of the inequality group, and the optimal solution in the case of the present embodiment is:

[0092] p A ≈ 0.42, p B ≈ 0.26, p C ≈ 0.18, and p D ≈ 0.14.

[0093] According to step S4, the stacking baskets and RGV trolleys are increased according to the optimal solution of the inequality group.

[0094] According to the optimal solution obtained in step S3, the stacking baskets are increased to 6 and the RGV trolleys are increased to 2, and after re-simulating the operation of the system under this working condition, the system can smoothly perform the stacking operation.

[0095] Embodiment 3

[0096] The embodiment of the present application also provides an electronic device, including a memory and a processor, the memory has a program running on the processor, and the processor executes the steps of the robot stacking planning method based on the negative multinomial distribution when running the program. The robot stacking planning method based on the negative multinomial distribution is introduced in the foregoing part, and will not be repeated here.

[0097] Embodiment 4

[0098] The embodiment of the present application also provides a computer readable storage medium, which has computer instructions stored thereon, and the computer instructions execute the steps of the robot stacking planning method based on the negative multinomial distribution when running, and the robot stacking planning method based on the negative multinomial distribution is introduced in the foregoing part, and will not be repeated here.

[0099] The present application also provides a stacking and stacking device applying the robot stacking planning method based on the negative multinomial distribution, as shown in the figure, including a plate chain conveyor 11, a stacking robot 12, a stacking robot track 13, stacking baskets 14A-14F, RGV trolleys 15A-15B, RGV tracks 16, and packing workstations 17A-17B. Figure 2

[0100] The plate chain conveyor 11 includes a motor, a plate chain, a frame, and a tensioning device, and the circular motion of the plate chain conveys the workpieces to run forward.

[0101] The stacking robot track 13 is arranged in parallel on one side of the plate chain conveyor 11, and a robot automatic sliding table is arranged thereon. The stacking robot 12 is fixed on the robot automatic sliding table, and the robot automatic sliding table drives the stacking robot 12 to move left and right on the stacking robot track 13 to stack workpieces of different specifications and models into corresponding stacking baskets.

[0102] The stacking robot 12 moves along the stacking robot track 13, and an automatic gripper device installed at the end of the arm of the stacking robot 12 is used to grab the workpieces to be stacked on the plate chain conveyor 11 and move the workpieces to the corresponding stacking baskets.

[0103] Each of the stacking baskets 14A-14F includes a frame, walking wheels, and a locking and fixing device. The stacking baskets are placed on the tracks of the frame base, and the locking and fixing device locks and fixes the stacking baskets on the frame base to prevent left and right movement during robot stacking.

[0104] ​RGV trolley 15A, 15B is used to carry workpieces to move on the RGV trolley track, and the RGV trolley 15A, 15B is provided with an automatic dragging device and a locking and fixing device. When any one of the stacking baskets 14A-14F is full of workpieces, the RGV trolley 15A or 15B automatically runs to the stacking basket, the automatic dragging device of the RGV trolley automatically runs to the stacking basket, the hook on the dragging device automatically hooks the stacking basket, the locking and fixing device on the stacking basket automatically opens, and the automatic dragging device drags the stacking basket to the RGV trolley. After the stacking basket is in place, the locking and fixing device on the RGV trolley locks and fixes the stacking basket full of workpieces.

[0105] The RGV trolley track 16 is arranged between the stacking and packing stations, and the RGV trolleys 15A, 15B can move freely along the track.

[0106] The packing stations 17A, 17B include a packing platform, a packing locking and fixing device, and an automatic rotating device.

[0107] In work, the RGV trolley carries a full load of stacking baskets and runs to the packing station 17A or 17B. The locking and fixing device on the RGV trolley automatically opens, and the automatic dragging device of the RGV trolley drags the full load of stacking baskets to the packing platform of the packing station for workpiece packing. The packing locking and fixing device locks and fixes the stacking basket, and after the packing is completed, the packing platform automatically rotates 90 degrees to unload the packed workpieces from the stacking basket. The packing platform reversely automatically rotates 90 degrees to return to the packing state, and the RGV trolley drags the empty stacking basket on the packing platform to the stacking station.

[0108] The number of the above-mentioned RGV trolleys and stacking baskets is determined according to the optimal solution calculated by the above-mentioned robot stacking planning method based on the negative multinomial distribution combined with specific production conditions.

[0109] Those skilled in the art can understand that the above description is only the preferred embodiments of the present application and is not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments or make equivalent replacements to some technical features. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A robot palletizing planning method based on negative multinomial distribution, comprising the following steps: Extract the characteristic parameters of workpiece palletizing and construct a probability model of workpiece arrival; By combining actual needs, an objective function is constructed to form an optimization model for the probability of workpiece arrival; Solve the optimization model to determine the optimal solution to the system of inequalities; Based on the optimal solution of the system of inequalities, add palletizing baskets and RGV trolleys; The characteristic parameters of the workpiece palletizing include: workpiece name, rated load capacity of palletizing baskets, and workpiece production cycle time. The workpiece name is used to identify different workpieces to be stacked; The rated load capacity of a palletizing basket refers to the standard number of the same type of workpiece that the palletizing basket can hold. The production cycle time of a workpiece refers to the time interval between the arrival of the workpiece. The probability model for the arrival of each of the multiple workpieces is as follows:

2. The robot palletizing planning method based on negative multinomial distribution according to claim 1, characterized in that, The probability model for the arrival of a workpiece refers to a prediction model for the arrival of different workpieces. Based on the characteristic parameters of workpiece stacking, the probability model for the arrival of a workpiece is constructed using a negative multinomial distribution, including the probability model for the arrival of each of the multiple workpieces.

3. The robot palletizing planning method based on negative multinomial distribution according to claim 2, characterized in that, The actual requirement in the step of constructing an objective function based on actual needs to form an optimal model of the probability of workpiece arrival is the requirement for the relationship between the probabilities of arrival of each workpiece.

4. The robot palletizing planning method based on negative multinomial distribution according to claim 1, characterized in that, The steps of solving the optimization model and determining the optimal solution of the system of inequalities include: using a nonlinear programming algorithm to solve the optimization model with the goal of minimizing the highest proportion of workpieces, and determining the optimal solution of the system of inequalities.

5. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a program that runs on the processor, and the processor executes the steps of the robot palletizing planning method based on the negative multinomial distribution as described in any one of claims 1-4 when running the program.

6. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed, they perform the steps of the robot palletizing planning method based on negative multinomial distribution as described in any one of claims 1-4.

7. A palletizing and stacking device applying a robot palletizing planning method based on negative multinomial distribution, comprising: Plate chain conveyor, palletizing robot, palletizing robot track, palletizing basket, RGV trolley, RGV track, packaging workstation, characterized in that, The number of palletizing baskets and RGV trolleys is determined by the optimal solution based on the robot palletizing planning method according to any one of claims 1-4.

Images