Intelligent stacking and typesetting method and system based on self-adaptive algorithm

Abstract

The invention relates to the technical field of stacking and typesetting, in particular to an intelligent stacking and typesetting method and system based on a self-adaptive algorithm, and the method comprises the steps: respectively obtaining a product parameter set and a tray parameter set, and extracting the maximum length, the maximum width and the maximum height from the product parameter set to determine the division step length of a three-dimensional discrete space set; pre-screening coordinate points meeting a threshold value by the system, and inputting the coordinate points into a target function to generate an arrangement scheme set; introducing a support consistency index and a mass weighted average value to generate a unit stability mark set and an overall structure judgment mark; the system generates an operable product set in the arrangement scheme set by traversing the structure dependency relationship, and the user performs adjustment according to the operable product set; and the system screens the arrangement points conforming to the threshold value to form an execution task set and sorts the execution task set, corresponding robot execution vectors are constructed, and an instruction sequence formed by all the execution vectors is coded into a standard action file issued to a robot controller.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic stacking and arranging, and in particular to an intelligent automatic stacking and arranging method and system based on an adaptive algorithm. BACKGROUND

[0002] In modern industrial logistics and manufacturing scenarios, automatic stacking and arranging is directly related to the efficiency and safety of product packaging, warehousing and transportation. Traditional intelligent automatic stacking systems mostly rely on manually preset fixed templates or models trained based on large-scale historical data to generate arrangement schemes. Template-based schemes are often only suitable for single pallets and single product scenarios. Once the product size, quantity or pallet specifications change, the template cannot be adapted and manual intervention or even redesign is required, which seriously reduces the flexibility of the system.

[0003] While systems relying on historical data have some degree of adaptability in some scenarios, they are highly dependent on data size and quality, have high model training costs, and are slow to react to new materials or temporary tasks, making them unable to meet the real-time requirements of industrial sites. At the same time, existing systems mostly focus on maximizing space utilization and do not adequately consider structural stability, lack of judgment of physical constraints such as center of gravity and support, leading to problems such as slippage and collapse during robot execution, affecting job safety.

[0004] Further, most systems lack interactive functions and instruction conversion capabilities, users cannot make targeted local adjustments based on system-generated schemes, and it is also difficult to directly generate robot-executable control instructions. These problems result in the existing technology generally lacking in adaptability, stability and execution, especially in high-level stacking, irregular products, non-ideal pallets and complex tasks requiring human-machine collaboration, making it difficult for existing solutions to meet industrial needs. SUMMARY

[0005] The present application aims to provide an intelligent automatic stacking and arranging method and system based on an adaptive algorithm to solve the problems raised in the background.

[0006] To achieve the above-mentioned purpose, the technical solution adopted by the present application is: An intelligent automatic stacking and arranging method based on an adaptive algorithm, comprising: Respectively acquiring a product parameter set and a pallet parameter set, extracting the maximum length, maximum width and maximum height from all product parameter sets to determine the division step of a three-dimensional discrete space set for placing products, and defining the three-dimensional discrete space set as a position set for placing products; The system pre-screens all the placeable coordinate points in the position set to obtain coordinate points meeting the threshold value, inputs the coordinate points into a target function, and the target function performs a discrete search in the position set, combines heuristic strategy arrangement iteration to generate an arrangement scheme set; Based on the position set and the product size parameter set, a support consistency index and a quality weighted average value are introduced to respectively determine the physical support relationship of each product unit and the relationship between the position of the overall quality center of gravity and the tray support boundary, and output a unit stability index set and an overall structure judgment index; The system calculates the dependent state of each unit in the arrangement scheme set by traversing the structure dependency relationship, generates an operable product set, and the user adjusts according to the operable product set. The system reconstructs the local structure legality index of the support area where the adjusted unit is located. If the legality index is greater than the threshold value, the system calculates the influence of the adjustment operation on the global center. When the influence of the global center meets the threshold value, the system accepts the adjustment and outputs the adjusted arrangement scheme and the stability index set; The system selects arrangement points meeting the threshold value from the stability index set to form an execution task set. The system sorts the tasks in the execution task set according to the stacking number and spatial distribution of the arrangement structure, constructs a corresponding robot execution vector based on the execution task set, and encodes the instruction sequence composed of all execution vectors into a standard action file to be sent to the robot controller.

[0007] As preferred, the tray parameter set includes the tray length, width, height, and maximum allowed bearing mass, and the product parameter set includes the product length, width, height, mass, and number.

[0008] As preferred, after defining the three-dimensional discrete space set as the position set of placing products, it further includes: The system calculates the maximum stacking number allowed by the current task according to the proportional relationship between the product height and the maximum stacking height of the tray.

[0009] As preferred, after outputting the unit stability index set and the overall structure judgment index support consistency index, it further includes: When the unit stability index is greater than the threshold value, it indicates that the product structure is in a potentially unstable state; When the unit stability index is less than the threshold value, it indicates that the product is unstable and needs to be rearranged.

[0010] As preferred, after the system reconstructs the local structure legality index of the support area where the adjusted unit is located, it further includes: If the legality index is less than the threshold value, the system determines that the operation makes the structure unstable, automatically rolls back, and prompts the user that the operation is out of the allowed range.

[0011] Preferably, the set of execution tasks includes placement coordinates, orientation angles, and target item information.

[0012] Preferably, the standard action file can be directly converted into RAPI, KRL, or URScript control language format.

[0013] Preferably, the set of product parameters and the set of pallet parameters are input by a user when configuring a task scheduling interface or automatically transmitted by an enterprise resource planning system through a standard data interface, and the system automatically loads and performs data consistency verification in an initialization stage.

[0014] Preferably, the set of arrangement schemes includes product center point coordinates and orientation angles.

[0015] An intelligent palletizing layout system based on an adaptive algorithm, comprising: A space modeling module is configured to obtain sets of product and pallet parameters, extract maximum length, maximum width, and maximum height from all sets of product parameters to determine a division step of a set of three-dimensional discrete spaces for placing products, and define the set of three-dimensional discrete spaces as a set of positions for placing products. A layout optimization module is configured to perform pre-screening on all placeable coordinate points in the set of positions to obtain coordinate points meeting a threshold value, input the coordinate points into a target function, perform discrete search on the set of positions, combine heuristic strategy arrangement iteration, and generate a set of arrangement schemes. A structure verification module is configured to introduce support consistency indicators and quality weighted average values based on the set of positions and the set of product size parameters to respectively determine physical support relationships of each product unit and a relationship between a position of a global quality center of gravity and a support boundary of a pallet, and output a set of unit stability indicators and a global structure determination indicator. A structure adjustment module is configured to calculate a dependent state of each unit in the set of arrangement schemes by traversing structure dependency relationships, generate a set of operable products, adjust the set of operable products according to a user, reconstruct a local structure legality indicator of a support area where the adjusted unit is located, calculate an influence of the adjustment operation on a global center when the legality indicator is greater than a threshold value, and accept the adjustment and output an adjusted arrangement scheme and a set of stability indicators when the influence of the global center meets a threshold value. An action execution module is configured to screen arrangement points meeting a threshold value from the set of stability indicators, compose a set of execution tasks, sort tasks in the set of execution tasks according to a number of stacked layers and spatial distribution of an arrangement structure, construct corresponding robot execution vectors based on the set of execution tasks, and encode an instruction sequence composed of all execution vectors as a standard action file to be issued to a robot controller.

[0016] Compared with the prior art, the present application has the following advantages: The present application realizes the whole process closed loop from input product and pallet parameters to final execution action through the links of step-by-step modeling, layout optimization, stability verification, local fine-tuning and robot instruction generation. The system first establishes a three-dimensional space model according to the task parameters to ensure that the layout calculation has clear boundary conditions, and then introduces a comprehensive optimization strategy of height penalty, support regularity and barycenter control in the arrangement process to generate a preliminary arrangement scheme that meets the structural stability requirements. Subsequently, lightweight physical rules are used for structural legality verification to determine whether the arrangement is executable from the aspects of local support relationship and overall barycenter distribution. On this basis, the system allows users to perform translation or rotation operation on part of the adjustable units in the stable structure, and automatically completes local legality verification and global barycenter constraint judgment after fine-tuning to ensure that the human-machine collaborative adjustment is still within the safe range. Finally, the system converts the verified arrangement scheme into robot control instructions containing grabbing points, placing points, angles and action sequences, which are directly called and executed by the industrial controller.

[0017] The present application does not need to rely on fixed templates or large-scale data training, the algorithm structure is lightweight, suitable for fast running under ordinary computing resource conditions, widely applicable to industrial scenes such as palletizing, boxing and loading, and can significantly improve the structural stability, adaptability and actual deployability while ensuring efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 The figure is a flow chart of the intelligent palletizing layout method based on the adaptive algorithm in the embodiment of the present application. Figure 2 The figure is a block diagram of the intelligent palletizing layout system based on the adaptive algorithm in the embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0020] Please refer to Figure 1 The present application proposes an intelligent palletizing layout method based on an adaptive algorithm, which includes: Step one: respectively acquire product parameter set and pallet parameter set, extract maximum length, maximum width and maximum height from all product parameter sets to determine the division step of three-dimensional discrete space set for placing products, and define the three-dimensional discrete space set as the position set for placing products, which specifically includes: The goal of this step is to build a three-dimensional space modeling structure and product property set for subsequent layout optimization calculation according to the product parameters and pallet parameters input in the task configuration stage. The data used in this step has a clear source and clear structure, and can be imported by the system front-end input interface or the task scheduling module in the industrial control system in a standard format, with good operability and engineering feasibility.

[0021] The geometric and physical parameters of the product mainly include: The length of the product , width , height , mass and quantity . The pallet parameters include the length , width , height and maximum allowable carrying mass of the pallet. All of the above parameters are input by the user when configuring the task scheduling interface or automatically transmitted by the enterprise resource planning system through a standard data interface. The system automatically loads and performs data consistency verification in the initialization stage. All length parameters will be rounded to the system's specified minimum length unit (e.g. integer multiples of 10) to ensure the accuracy of space division and layout consistency.

[0022] In the space modeling process, the system divides the pallet size into a three-dimensional discrete space set that can be used for product placement. To accommodate various product specifications rather than relying on fixed templates, this step proposes an adaptive space discretization strategy based on the maximum product size. The system first extracts the maximum length, width and height from all products, denoted as , , respectively, and determines the division step size of the three-dimensional discrete unit accordingly. The resulting space coordinate set after discretization is defined as: where , , are the index variables in the pallet length, width and height directions, respectively, with the value range calculated based on the pallet size and the corresponding step size; , , represent the maximum length, width and height of all product sizes in the current task, rounded to the system's supported minimum unit; represents the set of all theoretically possible product placement positions, constituting the spatial boundary for subsequent layout search and optimization.

[0023] Meanwhile, to prevent the generated layout structure from exceeding the allowed stacking height of the tray, the system needs to calculate the maximum number of stacking layers allowed for the current task according to the proportional relationship between the product height and the maximum stacking height of the tray. The number of layers is represented in integer form, and the calculation method is as follows: wherein, represents the maximum number of stacking layers allowed for the product of the category in the current task; represents the maximum stacking height allowed by the system, which is obtained by subtracting the additional safety height reserved from the maximum stacking height parameter of the tray design; is the height of the product of the category; the symbol represents the floor operation.

[0024] Each coordinate point of the spatial point set represents a theoretical placement position. Whether a specific product can be placed in this position needs to be matched and verified according to the remaining size at this position and the size of the current product. Those that do not meet the conditions will be removed from the layout candidate set in subsequent steps. This step only defines the spatial boundary and does not perform structure judgment. The structure stability will be handled in the subsequent layout and simulation steps.

[0025] Step two: the system performs pre-screening on all placeable coordinate points in the position set to obtain coordinate points that meet the threshold value. The coordinate points are input into the objective function, which performs discrete search in the position set, combines heuristic strategy layout iteration, and generates a layout scheme set, which specifically includes: This step takes the spatial point set constructed in step one and the product parameter set as input to calculate a set of spatial layout schemes to guide subsequent robot stacking or boxing operations. Unlike traditional algorithms that only pursue volume utilization, this step is specially designed for the "structure stability priority" scenario requirement proposed in the present invention, ensuring that the system not only calculates a reasonable spatial layout combination, but also meets the preliminary stacking stability and center of gravity balance in structure, providing a reliable foundation for subsequent physical simulation and robot execution.

[0026] First, the system performs pre-screening on all placeable coordinate points in the point set , requiring each candidate point to meet the condition that the product size can be completely accommodated without exceeding the boundary. The judgment rules are as follows: wherein, respectively represent the continuous effective space length in three directions starting from point , which is obtained by The geometric structure is directly calculated. The above constraints ensure that no product will exceed the limits or overlap due to insufficient space.

[0027] After screening, the layout optimization stage begins. This invention proposes an objective function with a "structural stability regularization term" to prioritize locations with a low center of gravity and sufficient support area during the layout process, thereby constructing a preliminary steady-state layout strategy. The objective function is designed as follows: The meanings of the variables are as follows: Indicates the first The height of each product after placement, i.e. This reflects the vertical position of the product within the stack; The bottom support ratio of this product is defined as the ratio of the actual contact area of ​​the product to the theoretical bottom area. The ratio is determined by the lower structure; This indicates the quality of the product. Local centroid coordinates of the location , used to calculate the overall center of gravity offset of the current structure; The reference coordinates for weighting the quality center of all products in the current layer; coefficients These represent the weights of the spatial height penalty, support stability penalty, and center of gravity offset penalty, respectively; constant term. To prevent positive numbers with a denominator of zero, they are usually set to 0. .

[0028] This objective function possesses the following innovative characteristics: 1. The first penalty is height, which encourages products to be arranged in a lower position, which helps to lower the overall center of gravity; 2. The second term is a support regularization term, which applies when the support area is insufficient ( (Small) Increased penalty to prevent suspension or "heavy object falling" structures; 3. The third method controls the overall mechanical center of the structure to converge towards the center by penalizing local center of gravity shifts, which is suitable for center of gravity-sensitive scenarios such as robot loading or shelf stacking.

[0029] This objective function can be obtained through a discrete search method. Concentrated execution at limited points, combined with heuristic strategies (such as prioritizing support ratios) The system iterates through the placement of products (at higher positions). Each time a product is placed, the system updates the occupied area and local mass distribution information of the current structure, and continues selecting the placement of the next product until all products are placed or no more products can be placed.

[0030] From the perspective of each product Two angles are allowed by default: and The system automatically switches based on spatial adaptability. The rotation strategy is considered during the layout search and is not controlled by a separate variable.

[0031] Step 3: Based on the location set and product size parameter set, introduce the support consistency index and the weighted average value of mass to determine the physical support relationship of each product unit and the relationship between the position of the overall mass center of gravity and the pallet support boundary, and output the unit stability flag set and the overall structure judgment flag, specifically including: The goal of this step is to verify the layout scheme generated in step two. Does it possess structural stability in a physical sense and can it be safely executed by subsequent automated equipment such as robots? In the previous step, we have initially constructed a reasonable arrangement combination of structures through spatial nesting rules and support priority mechanisms. However, since this process did not introduce actual force analysis, the following problems may still exist: partial product suspension, abnormal stacking force transmission, and outward deviation of the overall center of gravity of the structure. These problems can easily cause engineering risks such as stack slippage, overturning, and robot placement failure during palletizing.

[0032] To achieve structural verification without introducing a large-scale physics simulation engine, this step designs a lightweight, hierarchical, multi-constraint engineering stability verification mechanism. This mechanism is based on physical support relationships and focuses on mass distribution to jointly evaluate element stability and overall stability. All input data are derived from the arrangement results of the previous step. With the product parameter set provided in step one No additional information needs to be introduced.

[0033] First, the system constructs a support mapping matrix for each layer, that is, it determines whether each product unit has a valid physical support relationship. To this end, we define a support consistency index for each product unit. This is used to measure whether the object is in a placement state where it is "sufficiently supported by the underlying structure and the supporting structure is stable." The specific calculation method is as follows: in, This indicates the actual supporting area between the bottom of the current unit and the object in contact with it on the lower layer. This is the theoretical area of ​​the bottom surface of the product; It is the two-dimensional coordinate of the product's center of mass projected onto the lower layer. These are the centroid coordinates of the actual support area; It is half the length of the product's diagonal; Weighting coefficient set for the system. The index has the following physical meaning: the first term measures whether the support area is sufficient, and the second term measures whether the center of force is near the support center. The value range is When , it indicates that the product structure is in a potentially unstable state; when , it can be considered that the local placement structure meets the most basic mechanical stability.

[0034] Compared with the traditional support area judgment (such as only judging ), the index introduces the mass center offset factor, and has stronger robustness and judgment accuracy for irregular stacking, special-shaped or center of gravity offset products. At the same time, since it only depends on the geometric structure and mass center position, it does not need any complex modeling and can be calculated online in real time.

[0035] Next, to ensure the macroscopic structural stability of the entire stack, the system needs to evaluate the relationship between the position of the overall mass center of gravity and the tray support boundary to prevent the risk of overturning due to the offset of the center of gravity. We define the two-dimensional center of gravity coordinates of the entire stack structure as the weighted average of the mass of all products, which is calculated as follows: where is the center coordinate of each product, which is directly provided by the arrangement scheme ; is the mass of the corresponding product, which is provided by the product parameter set . If the Euclidean distance between the center of gravity coordinates and the tray center position exceeds a certain proportion of the short side length of the tray (for example ), it is considered that there is a risk of overturning as a whole.

[0036] The above two judgment mechanisms respectively judge whether the arrangement scheme is physically meaningful from the aspects of "local support structure" and "global mass distribution", and are all explainable, implementable, and lightweight judgment models that can be embedded into the control system. Unlike the traditional "rendering + simulation" scheme, this step can output stability analysis results within milliseconds, adapt to the production rhythm, and the results are traceable.

[0037] Finally, the system outputs two result variables: the first is the unit stability flag set , where indicates that the product unit is stable in support structure, indicates instability and needs to be rearranged; the second is the overall structure judgment flag . When all and the center of gravity offset (where is the tray center, When the maximum center of gravity offset allowed by the system is reached, , otherwise .

[0038] If or , the system returns to step two, and only locally optimizes the unstable area, avoiding the whole structure to be rebuilt. This step plays a key role in bridging the gap between "geometric arrangement" and "structurally executable" in the whole invention system, and has high engineering practicality and innovative structural design significance.

[0039] Step four: the system calculates the dependent state of each unit in the arrangement scheme set through the traversal of the structure dependency relationship, generates the set of operable products, and the user adjusts according to the set of operable products. The system reconstructs the local structure legality index of the unit in the adjusted support area, and if the legality index is greater than the threshold value, calculates the influence of the adjustment operation on the global center. When the influence of the global center meets the threshold value, the system accepts the adjustment and outputs the adjusted arrangement scheme and the stability flag set, which specifically includes: This step provides a controllable range of human-computer interaction mechanism based on the structure stable arrangement scheme generated by the system, allowing the user to perform translation or rotation operations on the specified product unit, and updating the overall arrangement structure under the premise of local structure legality. This step neither regenerates the arrangement nor builds a physical simulation model, but only limits the modification within the structure and local recalculation of a part of the unit in the verified structure, serving the human-computer co-creation needs in real-world scenarios, such as: The customer wants to stack a certain type of product on the top layer for priority delivery; The label surface of a certain type needs to face outward to cooperate with code scanning or visual detection; Some areas need to be left empty for additional manual restocking and other special operations.

[0040] First, the system identifies the "operable product set" according to the flag set , which only contains product units that meet the following conditions: 1. , that is, the current structure of the product is stable; 2. No other product is supported by this unit; 3. The product is the uppermost layer, or the unit above it does not directly contact it.

[0041] The system calculates the "dependent state" of each unit in the set through the traversal of the structure dependency relationship, generates the set , and marks it visually in the user interface.

[0042] The user selects a unit in the interface ​and adjust its position from to and its orientation from to . Upon confirmation, the system immediately reconstructs the local structure validity index for its support area, which is calculated as follows: where: : the contact area between the product and the support surface after adjustment, calculated from the underlying structure in ; : the theoretical base area of the product, calculated from the product dimensions ; : the projection of the product mass center, calculated using ; : the geometric center of the adjusted underlying support area; : half of the diagonal of the product base; : a system-fixed parameter used to weigh the contribution of the two terms.

[0043] If , the system determines that the operation destabilizes the structure, automatically rolls back, and prompts the user with "Operation out of allowed range." If is met, the system further calculates the impact of this operation on the global center of gravity: where: is the mass-weighted center of gravity of the original arrangement structure (defined in step three); is the updated center of gravity, recalculated from and the product mass ; is the maximum allowed center of gravity drift set by the system, typically left and right of the short side of the pallet.

[0044] If both indicators are met, the system accepts this fine-tuning and updates the adjustment results to the new arrangement structure and regenerates the stability flag . Otherwise, the original structure is restored, and the operation is invalid.

[0045] Example: Suppose in a certain task, product A is a carton with dimensions , mass , and original position . The user wants to move it to and rotate it to . The system first determines whether it belongs to the operable set , and if so, calculates and If any of the conditions are not met, the operation is revoked.

[0046] Finally, this step outputs two variables: the new arrangement scheme and the new set of structure legality flags . Only when the adjustment does not destroy the structural stability is the result adopted by the system. Otherwise, the original scheme is maintained.

[0047] Step five: The system selects arrangement points that meet the threshold from the stability index set to form an execution task set. The system sorts the tasks in the execution task set according to the stacking number and spatial distribution of the arrangement structure, builds a corresponding robot execution vector based on the execution task set, and all execution vectors constitute an instruction sequence that will be encoded as a standard action file issued to the robot controller, which specifically includes: This step aims to convert the arrangement scheme output by step four that is structurally stable and logically legal and the stability flag set into a set of operation instructions that can be directly executed by industrial robots. This step no longer optimizes, models, or verifies the structure, but only completes the instruction generation from arrangement data to physical task actions, which is the execution landing link of the entire system.

[0048] The system first selects all stable units, i.e., arrangement points that meet to form an executable task set . Each task corresponds to a specific placement operation and contains placement coordinates , orientation angles , and information about the target object.

[0049] Subsequently, to reduce robot repeated paths and invalid switching, the system sorts the tasks in according to the stacking number and spatial distribution of the arrangement structure. The sorting priority is first based on height (z-axis) from low to high, and then arranged using the shortest spatial distance principle within the same layer. The path cost objective function is: This formula represents the total displacement path during robot execution, which can be quickly calculated in each layer execution order through greedy sorting or local optimization algorithms, without the need for global optimization.

[0050] For each placement task sorted, the system builds a corresponding robot execution vector , which represents a complete pick-and-place operation: Where: ​: The picking point of the product, usually from the unified feeding point, can be set in the task configuration; : The placement point; : The placement angle; : The action label, which is automatically generated by the system according to the product type and stacking method (such as clamping, rotating, and slow placement, etc.), without user input.

[0051] For example, if the product is a type of plastic box with dimensions , the picking point is , the system determines the placement point to be , the angle is , and the action is clamping and placing, then the instruction vector is: All instruction sequences will be encoded into standard action files issued to the robot controller, which can be directly converted into RAPID (ABB), KRL (KUKA), or URScript (UR) control language formats.

[0052] In addition, to prevent collisions or misplacement caused by too fast operation, the system adds a minimum safety interval and a minimum movement distance (such as 50ms and 50mm) between adjacent instructions during instruction generation. The specific values are provided by the robot model and the workbench initialization configuration, and are not recalculated in this step.

[0053] Please refer to Figure 2 , in the second aspect of the present application, an intelligent stacking and layout system based on an adaptive algorithm is proposed, comprising: a space modeling module for acquiring parameter sets of products and pallets, respectively, extracting the maximum length, maximum width, and maximum height from all product parameter sets to determine the division step of a three-dimensional discrete space set for placing products, and defining the three-dimensional discrete space set as a position set for placing products; a layout optimization module for the system to pre-screen all placeable coordinate points in the position set to obtain coordinate points meeting the threshold, input the coordinate points into a target function, and perform discrete search on the position set, combined with heuristic strategy arrangement iteration, to generate a layout scheme set; a structure verification module for introducing support consistency indicators and quality weighted average values based on the position set and product size parameter set to respectively judge the physical support relationship of each product unit and the relationship between the position of the overall mass center and the support boundary of the pallet, and output unit stability flag set and overall structure judgment flag; The structure adjustment module is configured to calculate the dependent state of each unit in the arrangement scheme set by traversing the structure dependency relationship, generate an operable product set, and adjust the arrangement scheme according to the operable product set; the system reconstructs the local structure legality index of the support area where the adjusted unit is located, and if the legality index is greater than a threshold value, calculates the influence of the adjustment operation on the global center; when the influence of the global center meets the threshold value, the system accepts the adjustment and outputs the adjusted arrangement scheme and the stability index set. The action execution module is configured to filter arrangement points meeting the threshold value from the stability index set to form an execution task set, sort the tasks in the execution task set according to the stacking number and spatial distribution of the arrangement structure, construct a corresponding robot execution vector based on the execution task set, and encode the instruction sequence formed by all the execution vectors into a standard action file to be sent to the robot controller.

[0054] The above embodiments only describe the preferred embodiments of the present application, and do not limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by ordinary engineering technicians in the art without departing from the design spirit of the present application shall fall within the protection scope determined by the claims of the present application.

Claims

1. A smart palletizing and layout method based on an adaptive algorithm, characterized in that, include: Obtain the product parameter set and pallet parameter set respectively. Extract the maximum length, maximum width and maximum height from all product parameter sets to determine the partitioning step size of the three-dimensional discrete space set used to place the products, and define the three-dimensional discrete space set as the set of positions to place the products. The system pre-screens all placeable coordinate points in the location set to obtain coordinate points that meet the threshold. The coordinate points are then input into the objective function, which performs a discrete search on the location set. Combined with heuristic strategies, the system iterates through the arrangement to generate a set of arrangement schemes. Based on the location set and product size parameter set, a support consistency index and a mass-weighted average value are introduced to determine the physical support relationship of each product unit and the relationship between the position of the overall mass center of gravity and the pallet support boundary, and output the unit stability flag set and the overall structure judgment flag. The system traverses structural dependencies, calculates the dependent state of each unit in the set of layout schemes, and generates a set of operable products. Users make adjustments based on the set of operable products. The system reconstructs the local structural legality index of the support area where the adjusted unit is located. If the legality index is greater than the threshold, the system calculates the impact of the adjustment operation on the global center. When the impact on the global center meets the threshold, the system accepts the adjustment and outputs the adjusted layout scheme and stability flag set. The system selects arrangement points that meet the threshold from the stability index set to form an execution task set. The system sorts the tasks in the execution task set according to the stacking layer and spatial distribution of the arrangement structure. Based on the execution task set, the system constructs the corresponding robot execution vectors. The instruction sequence composed of all execution vectors will be encoded into a standard action file to be issued to the robot controller.

2. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, The pallet parameter set includes pallet length, width, height, and maximum allowable load capacity, while the product parameter set includes product length, width, height, mass, and quantity.

3. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, After defining the three-dimensional discrete space set as the set of locations where products are placed, it also includes: The system calculates the maximum number of stacking layers allowed for the current task based on the ratio between the product height and the maximum stacking height of the pallet.

4. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, After the output unit stability flag set and the overall structure judgment flag support the consistency index, it also includes: When the unit stability index is greater than the threshold, it indicates that the product structure is in a potentially unstable state. When the unit stability index is less than the threshold, it indicates that the product is unstable and needs to be rearranged.

5. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, After the system reconstructs the local structural legality index of the support area where the adjusted unit is located, it also includes: If the legality index is less than the threshold, the system determines that the operation has caused structural instability, automatically rolls back, and prompts the user that the operation has exceeded the allowed range.

6. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, The set of tasks to be performed includes placement coordinates, orientation angle, and target item information.

7. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, The standard action file can be directly converted into RAPI, KRL, or URScript control language formats.

8. The intelligent palletizing and layout method based on adaptive algorithm according to claim 1, characterized in that, The product parameter set and tray parameter set are entered by the user when configuring the task scheduling interface, or automatically transmitted by the enterprise resource management system through a standard data interface. The system automatically loads and performs data consistency verification during the initialization phase.

9. The intelligent palletizing and layout method based on an adaptive algorithm according to claim 1, characterized in that, The set of layout schemes includes the product center point coordinates and orientation angle.

10. An intelligent palletizing and layout system based on an adaptive algorithm, characterized in that, include: The spatial modeling module is used to obtain the parameter sets of products and pallets respectively, extract the maximum length, maximum width and maximum height from all product parameter sets, determine the partitioning step size of the three-dimensional discrete space set for placing products, and define the three-dimensional discrete space set as the set of positions for placing products; The layout optimization module is used by the system to pre-screen all placeable coordinate points in the location set to obtain coordinate points that meet the threshold. The coordinate points are input into the objective function, and the objective function performs a discrete search in the location set. Combined with heuristic strategies, the layout is iterated to generate a set of layout schemes. The structural verification module is used to determine the physical support relationship of each product unit and the relationship between the position of the overall mass center of gravity and the pallet support boundary based on the location set and product size parameter set, by introducing support consistency index and mass weighted average value, and outputs unit stability flag set and overall structure judgment flag. The structure adjustment module is used by the system to traverse structural dependencies, calculate the dependent state of each unit in the layout scheme set, generate an operable product set, and the user makes adjustments based on the operable product set. The system reconstructs the local structural legality index of the support area where the adjusted unit is located. If the legality index is greater than the threshold, the system calculates the impact of the adjustment operation on the global center. When the impact on the global center meets the threshold, the system accepts the adjustment and outputs the adjusted layout scheme and stability flag set. The action execution module is used by the system to select arrangement points that meet the threshold from the stability index set to form an execution task set. The system sorts the tasks in the execution task set according to the stacking layer and spatial distribution of the arrangement structure, and constructs the corresponding robot execution vectors based on the execution task set. The instruction sequence composed of all execution vectors will be encoded into a standard action file to be sent to the robot controller.

Images