An order sorting-based double-finger clamp stacking method and device and related medium

Abstract

The application discloses a double-finger clamp stacking method and device based on order sorting and related medium, and the method comprises the following steps: obtaining material parameters and performing structured modeling to obtain a material modeling object; performing double-grab grouping planning on the material modeling object to perform constraint checking and compatibility scoring pairing calculation, thereby obtaining a material grouping object; performing mixed stacking type planning on the material grouping object to generate stacking sequence and hierarchical pose set according to order sorting, thereby obtaining a stacking type object; performing grabbing and placing parameter solving by using the stacking type object to obtain an execution object; and performing instruction arrangement by using the execution object to generate a stacking instruction set for driving the double-finger clamp. The double-grab grouping planning and the mixed stacking type planning are performed on the material modeling object, and then the obtained stacking instruction set is used for double-grabbing and double-placing of the material, so that the operation efficiency of material transportation is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of logistics automation, and in particular to a double-finger clamp stacking method based on order sequencing, a double-finger clamp stacking device and related media. BACKGROUND

[0002] In the logistics automation stacking scenario, a mechanical arm cooperates with a clamp to complete the operation process of "grabbing-moving-stacking-returning". The mainstream existing scheme adopts a single-grabbing-single-stacking mode, that is, only one piece of material is grabbed and stacked in each cycle, and then returned to the grabbing position for the next cycle. In the application environment with multiple orders, this mode needs to frequently go back and forth between the grabbing position and the stacking position, and the path and action cannot be reused, resulting in low efficiency of the single-grabbing-single-stacking mode, which is difficult to meet the high-throughput stacking demand. SUMMARY

[0003] Embodiments of the present application provide a double-finger clamp stacking method based on order sequencing, a double-finger clamp stacking device and related media, aiming to solve the technical problem of low efficiency of the single-grabbing-single-stacking mode in the prior art.

[0004] In a first aspect, embodiments of the present application provide a double-finger clamp stacking method based on order sequencing, comprising:

[0005] obtaining material parameters and performing structured modeling to obtain a material modeling object;

[0006] performing double-grabbing grouping planning on the material modeling object to perform pairing calculation of constraint checking and compatibility scoring, to obtain a material grouping object;

[0007] performing mixed stacking type planning on the material grouping object to generate a stacking sequence and a hierarchical pose set according to order sequencing, to obtain a stacking type object;

[0008] solving grabbing and placing parameters by using the stacking type object to obtain an execution object;

[0009] performing instruction arrangement by using the execution object to generate a stacking instruction set for driving the double-finger clamp.

[0010] In a second aspect, embodiments of the present application provide a double-finger clamp stacking device based on order sequencing, comprising:

[0011] a data modeling unit configured to obtain material parameters and perform structured modeling to obtain a material modeling object;

[0012] a material grouping unit configured to perform double-grabbing grouping planning on the material modeling object to perform pairing calculation of constraint checking and compatibility scoring, to obtain a material grouping object;

[0013] A pile type planning unit is configured to perform mixed pile type planning on the material grouping object to generate a stacking sequence and a hierarchical pose set according to order sorting, and obtain a pile type object;

[0014] A parameter solving unit is configured to perform grabbing and placing parameter solving on the pile type object, and obtain an execution object;

[0015] An instruction arrangement unit is configured to perform instruction arrangement on the execution object, and generate a stacking instruction set for driving the double-finger clamp.

[0016] In a third aspect, an embodiment of the present application provides a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the order sorting based double-finger clamp stacking method of the first aspect when executing the computer program.

[0017] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is executable on a processor to implement the order sorting based double-finger clamp stacking method of the first aspect.

[0018] An embodiment of the present application provides an order sorting based double-finger clamp stacking method, which comprises obtaining material parameters and performing structured modeling to obtain a material modeling object; performing double-grab grouping planning on the material modeling object to perform pairing calculation of constraint checking and compatibility scoring, and obtain a material grouping object; performing mixed pile type planning on the material grouping object to generate a stacking sequence and a hierarchical pose set according to order sorting, and obtain a pile type object; performing grabbing and placing parameter solving on the pile type object, and obtain an execution object; and performing instruction arrangement on the execution object, and generate a stacking instruction set for driving the double-finger clamp. The double-grab grouping planning and the mixed pile type planning are performed on the material modeling object, and then the obtained stacking instruction set is used to perform double-grab double-place on the material, so that the operation efficiency of material transportation is improved.

[0019] An embodiment of the present application also provides an order sorting based double-finger clamp stacking device, a computer device, and a storage medium, which also have the above beneficial effects. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0021] Figure 1A flowchart of a double-finger clamp stacking method based on order sorting provided by the embodiment of the present application is shown in the figure.

[0022] Figure 2 A schematic block diagram of a double-finger clamp stacking device based on order sorting provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0024] It should be understood that when used in the specification and the appended claims, the terms "comprise" and "comprising" indicate the presence of the described features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0025] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0026] It should be further understood that the term "and / or" used in the present application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes these combinations.

[0027] Please see the following Figure 1 , Figure 1 A flowchart of a double-finger clamp stacking method based on order sorting provided by the embodiment of the present application is shown in the figure, which specifically includes steps S101-S105.

[0028] S101, obtaining material parameters and performing structured modeling to obtain a material modeling object;

[0029] S102, performing double-grab grouping planning on the material modeling object to perform pairing calculation of constraint checking and compatibility scoring, to obtain a material grouping object;

[0030] S103, performing mixed stacking type planning on the material grouping object to generate stacking order and hierarchical pose set according to order sorting, to obtain a stacking type object;

[0031] S104, solve the grabbing and placing parameters by using the pile-shaped object, and obtain an execution object;

[0032] S105, arrange instructions by using the execution object, and generate a stacking instruction set for driving the double-finger clamp.

[0033] In step S101, material parameter information for stacking is acquired, which includes data for representing the shape, mass, and position and posture of the material, and sorting information related to an order. Based on the information, the material is structured modeled in a unified format, and a material modeling object that can be directly called by subsequent planning and solving modules is obtained.

[0034] In an embodiment, the step S101 includes:

[0035] Original parameters of the material are collected, and an original parameter set is obtained;

[0036] A mapping relationship of attribute fields, constraint fields, and positioning fields is established by using the original parameter set, and a structured data set is obtained;

[0037] The structured data set is subjected to clamp capability checking, and a compatibility checking result is obtained;

[0038] The compatibility checking result is subjected to stacking constraint fusion, and an identification code and three-dimensional coordinates are associated as positioning parameters, and a regularized object is obtained;

[0039] The regularized object is objectized and packaged, and a material modeling object is obtained.

[0040] In this embodiment, original parameters of the material are collected, and an original parameter set is obtained. The original parameters include length, width, height, and weight data for describing the basic shape and mass of the material, and constraint data such as pallet length, width, and maximum stacking height for limiting the operation boundary; and pose information output by a visual recognition system is also acquired, including three-dimensional coordinates and attitude angles in the camera coordinate system. The above original parameters are preferentially converted by order data to make data from different sources have consistent measurement units and field naming. Then, three types of mapping relationships of attribute fields, constraint fields, and positioning fields are established based on the original parameter set, and a structured data set is obtained. The attribute fields are used to store physical attributes (length, width, height, and weight) for subsequent compatibility determination; the constraint fields are used to store pallet length, width, and maximum stacking height as planning boundaries; and the positioning fields are used to store visual recognition coordinates as the pose reference for grabbing and placing. Through the mapping, a unified data representation that can be directly accessed by downstream modules is obtained.

[0041] Further, a clamp capability check is performed on the structured data set, and a compatibility check result is output. The check rules include: a width bearing rule, the material width does not exceed the maximum support length W max of the clamp; a weight bearing rule, if the material is a single-grab object, the total weight M of the material is within the range of the maximum weight bearing of the single fork to twice the maximum weight bearing (M_single_max≤M≤2×M_single_max); if the material is a double-grab candidate, the weight of any material does not exceed the maximum weight bearing of the single fork (M1≤M_single_max and M2≤M_single_max); and an opening and closing distance rule, for subsequent possible pair grabbing, the length half of any two materials is less than the maximum opening and closing distance of the double finger (L1 / 2+L2 / 2<D_max). The rules are stored in the form of check marks and threshold values in the structured data set to support subsequent planning calls. The compatibility check result is fused with the stacking constraints to obtain a regularized object. In this process, the available space boundary is limited by the length and width of the pallet and the maximum stacking height; at the same time, the identification code of each material is associated with its three-dimensional coordinates (X, Y, Z) and attitude angle for positioning parameters, to ensure that the data association of the same material in the three dimensions of attribute, constraint and positioning is complete and the retrieval path is clear. Finally, the regularized object is objectized and encapsulated according to the regularized object, and a material modeling object is output. The material modeling object includes a physical attribute subset, a job constraint subset, a positioning parameter subset, and a clamp compatibility subset, and provides an access interface to the outside in a unified data structure, which is directly called in the subsequent grouping planning, stacking planning and parameter solving stages.

[0042] In step S102, based on the material modeling object, a grouping planning is performed for the cooperative grabbing demand of the double-finger clamp. This step performs pair selection and pairing determination on the materials according to the clamp capability and the job constraint, taking into account the job compatibility and the grouping effectiveness, and outputs a material grouping object representing the grouping result.

[0043] In an embodiment, the step S102 includes:

[0044] The material modeling object is extracted to obtain a candidate material set;

[0045] The weight screening set is obtained by performing weight screening processing on the candidate material set;

[0046] The size screening set is generated based on the length determination rule using the weight screening set to generate a size screening label, and the size screening set is obtained;

[0047] The queue object is obtained by performing queue classification on the size screening set; wherein the queue object includes a double-grab queue to be grouped and a stacking queue to be stacked;

[0048] The queue object is subjected to constraint checking to generate candidate pairs for the materials in the double-gripper grouping queue, while outputting compatibility scores, and the material grouping object is obtained by integration.

[0049] In this embodiment, based on the obtained material modeling object, records with complete attribute fields, constraint fields, and positioning fields are extracted to obtain a candidate material set. The candidate material set is used as an input set for subsequent grouping calculation, and its elements include physical attributes such as length, width, height, and weight, operation constraints such as pallet length and width and maximum stacking height, and three-dimensional coordinates and attitude angles represented in a visual recognition coordinate system. The candidate material set is subjected to weight screening processing to generate a weight screening set. The screening rule is based on the maximum single-fork load-bearing parameter M_single_max: when a material is used as a single-gripper object, its weight M falls within the interval M_single_max≤M≤2×M_single_max, and it is determined as not meeting the double-gripper condition; when a material is used as a double-gripper paired object, its single-piece weight must meet the requirements of M1≤M_single_max and M2≤M_single_max, and it can enter subsequent size checking. After the above determination, materials that meet the double-gripper weight condition are retained in the weight screening set.

[0050] Further, the length determination rule is used to generate a size screening label for the weight screening set to obtain a size screening set. The length determination is based on the maximum opening and closing distance D_max of the double-finger gripper and sets a safety margin of 50 mm: if the length L of the material satisfies L / 2<D_max-50 mm, it is labeled as meeting the double-gripper size condition; if L / 2≥D_max-50 mm, it is labeled as not meeting the double-gripper size condition. This labeling realizes the rapid diversion of candidate materials in the size dimension. The size screening set is then subjected to queue classification to obtain a queue object. Materials that meet both the weight screening condition and the size screening condition are classified into the double-gripper grouping queue, and the remaining materials are classified into the stacking queue. The queue object maintains the association between the material identification code and the positioning parameters in the data structure, ensuring traceability and callability in subsequent pairing and planning processes.

[0051] On this basis, constraint checking and compatibility scoring are performed on the queue object to generate candidate pairs and optimize the pairing for the materials in the to-be-grouped double-pick queue. The constraint checking adopts a hard-then-soft strategy: the hard constraint takes the opening and closing geometric conditions as the core, and determines whether the length half of any two materials A and B satisfies L_A / 2+L_B / 2≤D_max; those that do not satisfy are not entered into the scoring link. The candidate pairs that pass the hard constraint enter the compatibility scoring, and the scoring indicators include weight matching degree, height matching degree, width matching degree, and consistency of special stacking requirements, wherein the smaller the weight, width, and height differences are, the higher the score is, and combinations with the same special stacking restrictions are given additional scores; when the combination score is lower than a preset threshold, it is removed, and those higher than the threshold are sorted in descending order of score to determine the priority pairing order. Finally, the paired materials that satisfy the constraints and pass the scoring screening are integrated to obtain the material grouping object. The material grouping object records the member identifiers of each pair of paired materials, the merge parameter reference for subsequent planning, and the corresponding compatibility score, and removes the successfully paired materials from the to-be-grouped double-pick queue, and updates them in the to-be-stacked queue to be used for double-pick stacking, providing grouped input for subsequent mixed stacking type planning.

[0052] In an embodiment, the queue object is subjected to constraint checking to generate candidate pairs for the materials in the to-be-grouped double-pick queue, while outputting compatibility scores, and the material grouping object is obtained by integration, including:

[0053] specification analysis is performed on the to-be-grouped double-pick queue in the queue object to generate a specification analysis object;

[0054] same-specification priority grouping is performed using the specification analysis object to obtain a same-specification grouping object;

[0055] two-by-two combination enumeration and difference characteristic scoring are performed on the to-be-grouped set in the same-specification grouping object to generate a candidate pair set;

[0056] constraint checking and compatibility scoring are performed on the candidate pair set respectively to obtain a compatibility scoring object;

[0057] pairing selection is performed using the compatibility scoring object to update and output the material grouping object.

[0058] In the embodiment, the to-be-grouped double-grabbing queue in the queue object is analyzed to obtain a specification analysis object. The specification analysis object takes the length, width, height, and weight of the material as basic specification dimensions, and combines the maximum opening and closing distance D_max of the double-finger clamp and the maximum load-bearing parameter of the single fork to label the potential pairing paths of the same specification and different specifications, which serves as the basis for subsequent priority and constraint determination. In the same-specification preferential grouping stage, the specification analysis object is used to group the materials with a large number of the same length, width, height, and weight first. If the length L of the same-specification material is less than or equal to the maximum opening and closing distance D_max, the material is preferentially paired within the same specification to minimize the height difference when stacking on the same layer. When the pairing of the same-specification material is completed, if there is still a single material left, the single material is temporarily stored back to the to-be-grouped double-grabbing queue and waits for subsequent pairing with other-specification materials. If the length L of the same-specification material is greater than the maximum opening and closing distance D_max, the material is not paired within the same specification, and all materials of the same specification are retained in the to-be-grouped double-grabbing queue to be paired with other-specification materials according to the different-specification rule. In the different-specification grouping stage, a hard constraint check is first performed, that is, the geometric accessibility of the lengths of any two materials A and B in the to-be-grouped double-grabbing queue is determined. Only when the sum of the lengths of the two materials is less than or equal to the maximum opening and closing distance D_max, that is, the length of material A / 2 + the length of material B / 2 ≤ D_max, the combination of the two materials enters the compatibility scoring link. The combination that does not satisfy the condition is directly rejected. All combinations that satisfy the hard constraint form a candidate pairing initial set.

[0059] Further, the candidate pairing initial set is scored according to the soft constraint difference characteristics to obtain a candidate pairing set. The scoring includes four items: first, weight matching degree (full score 30 points), the smaller the weight difference between the two materials, the higher the score, calculated in the manner of “1 minus the ratio of the weight difference to the larger one, multiplied by 30 points”; second, height matching degree (full score 30 points), the smaller the height difference between the two materials, the higher the score, calculated in the manner of “1 minus the ratio of the height difference to the larger one, multiplied by 30 points”; third, width matching degree (full score 20 points), the smaller the width difference between the two materials, the higher the score, calculated in the manner of “1 minus the ratio of the width difference to the larger one, multiplied by 20 points”; and fourth, consistency of special stacking requirements (additional 20 points), when the two materials have the same special stacking restrictions (such as prohibited from being placed on the bottom layer, prohibited from being stacked on a certain type of material, etc.), additional points are given. The four scores of each combination are summed up to obtain a compatibility total score; when the total score is less than a preset threshold of 64 points, the combination is determined as not meeting the grouping conditions and is rejected.

[0060] In the combination of association constraint checking and compatibility score aggregation stage, the candidate pairs whose scores reach the threshold are sorted in descending order of total score, combined with queue deduplication and uniqueness constraints (each material is only allowed to participate in one final pairing in the same round), and selected to obtain the compatibility score object. The compatibility score object records the member identification of each pair of paired materials, the matching degree score, the total score, and the cited geometric accessibility and job boundary constraint information, which is used to support parameter calculation for subsequent stack type planning. Finally, according to the compatibility score object, the pairing selection and queue update are performed in sequence, and the material grouping object is output. For each pair of materials selected, it is removed from the double-grab queue and transferred to the palletizing queue; if there are still single-piece materials left after the end of the screening and pairing in a round, they are put into the single-grab queue for subsequent processing. The above process is executed in a loop until the double-grab queue is empty or only single-piece materials are left. Through this process, the material grouping object including multiple paired material entries is obtained, providing grouped, available and weighted input data for subsequent order-based mixed stack type planning.

[0061] In step S103, based on the material grouping object, mixed stack type planning is performed according to the order sorting requirements. This step takes the grouped materials as the basic stacking unit, determines the stacking order and hierarchical relationship of each unit, and determines the placement posture and position parameters for each unit in the layer to obtain the stack type object that can be used to call the solution.

[0062] In an embodiment, the step S103 comprises:

[0063] Merging each pair of grouped materials in the material grouping object into a single material and calculating the merging parameters to obtain a merged material set;

[0064] According to the order sequence, the merged materials in the merged material set are sequentially arranged to obtain an order sequence set;

[0065] Performing hierarchical rule calculation on the order sequence set to obtain a hierarchical order set;

[0066] Assigning materials with similar heights in the hierarchical order set to the same layer and calculating the placement orientation and position of the merged materials in each layer to obtain a hierarchical pose set;

[0067] Objectifying and packaging the hierarchical pose set to obtain a stack type object.

[0068] In the present embodiment, a heuristic rule algorithm can be used to plan the mixed stacking of grouped materials, ensuring that the stacking meets the overall principle of "heavy below and light above, large below and small above, and center of gravity in the middle". First, each pair of grouped materials in the material grouping object is merged into a single material and the merging parameters are calculated to obtain a set of merged materials. The merging parameters are determined as follows: the merging length is the sum of the lengths of the two materials plus a safety gap, specifically "merging length = material 1 length + material 2 length + 5mm (safety gap)"; the merging weight is the larger of the two weights "merging weight = max (material 1 weight, material 2 weight)"; the merging width is the larger of the two widths "merging width = max (material 1 width, material 2 width)"; and the merging height is the larger of the two heights "merging height = max (material 1 height, material 2 height)". The above calculations are performed for each pair of grouped materials to obtain merged material entries for subsequent planning.

[0069] Further, the merged materials in the set of merged materials are sequentially arranged according to the order sequence to obtain a set of order sequences. The sequential arrangement follows a priority rule: the merging weight is the highest priority, followed by the merging base area (merging length x merging width), and then the stability index is the third priority. According to this multi-level priority strategy, the merged materials are generated into linear order sequences as input sequences for hierarchical calculation. Then, hierarchical rule calculation is performed on the set of order sequences to obtain a set of hierarchical order sequences. The hierarchical rule revolves around the overall principle: first, the weight and base area are used to assign the level from bottom to top, so that the merged materials with heavier weight and larger base area are given priority to occupy the lower layer position; then, the tray boundary and single-layer capacity are used to check the capacity of each layer to ensure that the layer arrangement meets the requirements in terms of geometry and load bearing; and the center of gravity of each layer is calculated simultaneously when assigning the layers, so that the overall center of gravity is kept near the geometric center of the tray, thereby achieving the arrangement requirement of center of gravity in the middle.

[0070] Further, the same-layer aggregation distribution is performed on the merged materials with similar heights in the hierarchical sorting set, and the placement orientation and position of each layer of the merged materials are calculated to obtain a hierarchical pose set. Specifically, in each layer, the entries with smaller merged height difference values are preferentially placed in the same layer unit to unify the height reference plane of the layer; the candidate orientation set of each merged material is calculated (based on the merged length, merged width, and available laying direction), and the orientation that meets the layer arrangement gap and tray boundary is selected; after the orientation is determined, the plane placement coordinates (X, Y) and attitude angle (θ) of each merged material are calculated based on the remaining space in the layer, the safety gap with adjacent materials, and the geometric center of the tray, and the placement height (Z) is obtained by accumulating the layer height, and finally the pose parameter entry of each merged material is obtained. The pose entries of the same layer are summarized to form the pose sub-set of the layer, and the complete hierarchical pose set is obtained by layer-by-layer accumulation. Finally, the hierarchical pose set is objectized and packaged to obtain a stack object. The stack object at least includes: order sequence index, hierarchical number, pose parameters of each merged material, and associated merged parameter references (merged length, merged width, merged height, and merged weight).

[0071] In step S104, the grabbing and placing parameters are solved based on the stack object. This step takes the hierarchical pose set as input, calculates the key poses, opening and closing amounts, travel and safety related parameters of the double-finger gripper in the grabbing and placing process, and merges the control elements related to execution to obtain an execution object for driving the equipment.

[0072] In an embodiment, the step S104 includes:

[0073] The grabbing detection object is obtained by performing grabbing point positioning calculation based on the stack object.

[0074] The grabbing parameter object is obtained by performing distance and grabbing pose calculation using the grabbing detection object.

[0075] The safety parameter object is obtained by performing safety opening and closing distance calculation using the grabbing parameter object.

[0076] The grabbing control object is obtained by performing grabbing process control planning on the safety parameter object.

[0077] The execution object is generated by performing lifting and parameterized packaging on the grabbing control object.

[0078] In the embodiment, the obtained pile type object is used to calculate the positioning of the grabbing points according to the relationship between the on-site calibration, and a grabbing detection object is generated. The reference point selected by the stacking station is taken as the origin of the camera coordinate system, and the coordinates P1=(X0, Y0, Z0) of the origin of the camera coordinate system relative to the origin of the manipulator coordinate system are obtained through multiple teaching. When the material is delivered to the stacking station, the camera is triggered to detect and identify, and the grabbing point coordinates P2=(X1, Y1, Z1, θ1) and P3=(X2, Y2, Z2, θ2) of the two pieces of double-grabbed material in the camera coordinate system are obtained. P1, P2, P3 and their source identifiers are recorded together to obtain a grabbing detection object as an input for subsequent calculation. Based on the grabbing detection object, distance and grabbing pose calculation is performed to obtain a grabbing parameter object. The target opening and closing amount D1 of the double-finger clamp when taking goods is calculated, where D1 is the center distance of the two materials in the grabbing transverse direction, and the absolute value of |X1-X2| is taken. The detection result in the camera coordinate system is converted to the manipulator coordinate system to obtain the grabbing pose point P4, where the translation component is determined as follows: the X coordinate is X0+(X1+X2) / 2, the Y coordinate is Y0+Y2-W1 / 2, and the Z coordinate is Z0+Hf; where W1 is the width of material 1, and Hf is the height offset from the plane of the clamp bottom mechanism to the clamp mounting surface. P4 and the detection attitude angle information are taken as the parameter set of the grabbing pose, and D1 is archived together as the grabbing parameter object.

[0079] Further, based on the grabbing parameter object, safety opening and closing distance calculation is performed to obtain a safety parameter object. For the double-finger target opening and closing amount in the placement stage, the placement safety opening and closing amount D2=L1 / 2+L2 / 2+α is calculated, where L1 and L2 are the lengths of the two materials, and α is a preset safety gap (e.g., 2 mm). The values of D2 and α and their association with the material size field are recorded together to obtain a safety parameter object, which is used to ensure that the opening and closing of the clamp in the placement stage meets the gap requirement. The safety parameter object is further subjected to grabbing process control planning to generate a grabbing control object. The control planning includes continuous action segments and condition quantities: first, the manipulator moves along the planned path to the grabbing pose P4 while opening the double fingers to the target opening and closing amount; second, the clamp fork approaches and supports the material from the direction of the lifting mechanism at the bottom of the material; third, the pressing cylinder is pressed down until the pressure sensor reaches the set threshold; and fourth, the material is lifted to a safe height. The entry condition, maintenance condition and termination condition of each action segment and the corresponding sensor threshold are stored in the grabbing control object.

[0080] Finally, the execution object is generated by lifting and parameterizing the grasping control object. The execution object includes at least: a grasping pose P4, a grasping opening and closing amount D1, a placing safe opening and closing amount D2, a bias parameter Hf and a safety gap a, and a time sequence and threshold set corresponding to the action segment; and provides an access interface in a unified field format for direct reading and translation into device control instructions by the instruction arrangement of the following step S105.

[0081] In step S105, the execution object is used for instruction arrangement. This step converts the execution parameter set into a control instruction sequence that can be issued to the mechanical arm and the double-finger clamp, organizes and connects the starting, motion, stacking and off-site action segments, generates a stacking instruction set for driving the double-finger clamp to complete the stacking work, and outputs the stacking instruction set as a final output for the on-site control system to call.

[0082] In an embodiment, the step S105 includes:

[0083] The execution object is read for a placing parameter set;

[0084] The placing parameter set is used for starting action planning to output an initial path segment;

[0085] The initial path segment is used for parabolic path generation and obstacle avoidance solving, and the path is optimized for obstacle avoidance to obtain a placing path set;

[0086] The placing path set is used for deceleration zone and placing action solving to output a placing execution sequence;

[0087] The placing execution sequence is used for off-site and continuation process planning to obtain a stacking instruction set for driving the double-finger clamp.

[0088] In this embodiment, the execution object is read for a placing parameter set. The placing parameter set includes at least: target placing pose parameters X5, Y5, Z5 (coordinates of the sixth axis of the robot relative to the origin of the robot coordinate system), a safety distance parameter β (used for pre-approaching before placing, typically 100 mm), a target opening and closing amount D2 in the placing phase (calculated from the length parameter and the safety gap), and control amounts such as vertical lifting amount and action time sequence threshold. The above parameters are determined by the hierarchical pose set of the stacking object and the safety and pose fields in the execution object. The placing parameter set is used for starting action planning to output an initial path segment. Specifically, after the clamp feedback takes the completed signal, the robot is lifted vertically upward by 300 mm in the joint space to ensure that the clamp and the material are completely separated from the conveying line and the jacking mechanism, and a starting transition pose is obtained. The path start between the transition pose and the target placing pose is recorded as the initial path segment, which is used to connect the subsequent spatial motion planning.

[0089] Further, the initial path segment is used for parabolic path generation and obstacle avoidance solving, and the path is optimized for obstacle avoidance to obtain a placement path set. That is, starting from the transition pose after vertical lifting, a parabolic space trajectory is generated to make the manipulator move along a smooth curve to the pre-approaching point P5=(X5-β, Y5-β, Z5); then, based on the trajectory, the path is discretized and cost evaluated in combination with the obstacle distribution in the work area, and the path segment with potential collision risk is replaced and resampled to ensure that the material does not collide with any obstacle during stacking, and the placement path set is obtained. The placement path set is used to solve the deceleration zone and placement action, and the placement execution sequence is output. When the end effector reaches P5 along the path, it enters the deceleration zone, reduces the end effector speed, and gradually approaches the target placement position according to the pose interpolation; when the in-place threshold is met, the action sequence of “fork retraction-pose fixing-placement” is executed: the clamp maintains the supporting attitude of the material, and the fixing mechanism supports and limits the material at the stacking position; then, the double-finger clamp controls the opening and closing according to the target opening and closing amount D2 of the placement stage to complete the material release; after the retraction action is completed, a short stay is maintained to confirm the stable force of the fixing mechanism, and a placement completion flag is recorded. The entry conditions, speed / acceleration settings, and corresponding relationships with sensor thresholds of the above action segments are solidified into the placement execution sequence.

[0090] Finally, the placement execution sequence is planned for off-site and continuation, and a stacking instruction set for driving the double-finger clamp is obtained. The off-site planning makes the manipulator move away from the stacking position in a safe direction for a distance, and a continuation path is generated again using a parabolic trajectory to the next grasping pose; during the continuation process, the travel speed and end effector pose are restored to the grasping preparation state. The instructions are organized in time and logical order, translated into control instructions and parameter tables executable by the device, and constitute the final stacking instruction set for driving the manipulator and double-finger clamp to complete placement and enter the next cycle.

[0091] In a specific implementation scenario, a batch of shipment orders is faced, and the orders include three paper boxes to be processed and a set of double-finger clamp capability parameters. The system completes data modeling, grouping planning, mixed stacking type planning, and execution control according to the preset process, and finally completes single-grabbing and double-grabbing operations at the stacking station.

[0092] First, the system obtains the original parameters of the materials from the shipment order to obtain the original parameter set: the size of paper box A is L A ×W A ×H A = 530 mm × 430 mm × 300 mm, and the weight M A = 20 kg; the size of paper box B is L B ×W B ×H B= 460mm x 400mm x 300mm, weight M B = 11kg; the size of the carton C is L C x W C x H C = 440mm x 370mm x 290mm, weight M C = 9kg. The capacity parameters of the double-fingered gripper are M_single_max = 15kg, D_max = 460mm. The above parameters are objectified and encapsulated as the input of subsequent planning and solving.

[0093] Further, the double-grabbing grouping planning is performed. In terms of weight screening, the weight of the carton A is 20kg > M_single_max, which is determined not to meet the double-grabbing weight condition; the weights of the cartons B and C are 11kg and 9kg respectively, both of which are ≤ M_single_max, and are entered into the next step as double-grabbing candidates. In terms of size screening, according to the criterion of L A / 2 = 265mm < (D_max - 50mm) = 410mm, L B / 2 = 230mm < 410mm, L C / 2 = 220mm < 410mm, all of which meet the double-grabbing size condition. The soft constraint score is performed on B and C which meet the weight and size conditions: weight matching degree (1 - |M B - M C | / max(M B , M C )) x 30 = 24.54 points, height matching degree (1 - |H B - H C | / max(H B , H C )) x 30 = 28.99 points, width matching degree (1 - |W B - W C | / max(W B , W C )) x 20 = 18.2 points, and the total score of the three is 71.73 points, which exceeds the threshold value of 64 points, and it is determined that the double-grabbing pairing is available. Thus, the grouping result is determined as follows: the carton A adopts single-grabbing single-placing; the cartons B and C adopt double-grabbing double-placing.

[0094] Further, the mixed stack type planning is performed and the pose and sequence are generated. According to the principle of "heavy down and light up, large down and small up, and gravity center in the middle", the system solves the layout and orientation of the bottom layer: carton A is placed at the bottom layer coordinate (600, 300, 0, 0°); carton B is placed at the bottom layer coordinate (200, 300, 0, 90°); and carton C is placed at the bottom layer coordinate (200, 760, 0, 90°). Accordingly, the stack type object is obtained and the stacking sequence is output, wherein carton A is placed on the bottom layer first, and then the double-grabbing and double-placing of carton B and carton C are performed.

[0095] The execution phase is entered: the corresponding materials are transported to the stacking station according to the stacking sequence; the vision camera triggers detection and outputs the grabbing point and pose information, the system obtains the double-finger gripper grabbing pose and the taking opening and closing amount D1 (the absolute value of the horizontal center distance of the two target materials) under the "camera coordinate system - robot coordinate system" calibration relationship, and calculates the grabbing pose point combined with the gripper offset parameter to guide the gripper to the position. The grabbing action is performed according to the preset sequence: the gripper moves to the target pose and opens to D1, the forks support the materials from the direction of the bottom lifting mechanism, the material is lifted to a safe height as a whole after the pressure cylinder is pressed to the pressure threshold, and the execution object is formed.

[0096] Finally, the placement and closed-loop verification are completed according to the execution object. The system first plans the starting action, makes the robot vertically lift 300mm to completely separate from the conveying line and the lifting mechanism; then generates the placement path based on the parabolic main path obstacle avoidance solution, first reaches the pre-approaching point P5 = (X5 - β, Y5 - β, Z5), where β is 100mm, then slowly approaches the target pose, enters the deceleration zone to perform the fork withdrawal and release action, and the double fingers complete the material release according to the placement target opening and closing amount D2 (obtained from L1 / 2 + L2 / 2 + safety gap α). After the placement is completed, the vision detection confirms that the actual position deviation is ≤2mm, and the placement completion flag is recorded; the system immediately plans the off-site and continuation path to return to the next grabbing pose in a parabolic manner, and enters the next cycle. Through the above process, carton A completes single grabbing and single placing, carton B and carton C complete double grabbing and double placing, and the whole batch meets the order sorting and precision requirements, and completes the specific implementation scenario.

[0097] The application is based on order data to generate a mixed stacking type in advance, and then executes according to the stacking sequence of the stacking type. Through the structured modeling of material attributes, operation constraints and positioning parameters, combined with constraint checking and compatibility scoring of double-grab grouping, the size and weight matched grouped materials are preferentially selected; in the hierarchical planning, the rules of "heavy down and light up, large down and small up, and gravity center in the middle" are followed, and the orientation and placement pose of the combined materials are uniformly solved and packaged. The above process assigns each material to the most suitable position, the layer height reference is consistent, the stress path is clear, the overall stability of the stacking is improved, and at the same time the remaining space in the tray plane and height direction is effectively filled, so that the space utilization rate is improved. The application also adopts a "double-grab double-place" operation mode to complete the grabbing and placing of two materials in a single action cycle; combined with the parabolic main path and obstacle avoidance optimized instruction arrangement, the number of round trips and empty stroke time between the grabbing position and the stacking position is reduced, and the number of path switching and end reversing is also reduced. Compared with the traditional single-grab single-place mode, under the same working conditions and equipment capacity parameters, the overall beat is significantly shortened, and the system throughput is significantly improved; in a typical scenario, the comprehensive efficiency can be improved by about 70%, which can meet the high-throughput stacking demand based on order sorting.

[0098] In combination Figure 2 As shown in the figure, Figure 2 A schematic block diagram of a double-finger clamp stacking device based on order sorting provided by an embodiment of the application, the double-finger clamp stacking device based on order sorting 200, comprising:

[0099] A data modeling unit 201 for obtaining material parameters and performing structured modeling to obtain a material modeling object;

[0100] A material grouping unit 202 for double-grab grouping planning of the material modeling object to perform constraint checking and compatibility scoring pairing calculation to obtain a material grouping object;

[0101] A stacking type planning unit 203 for mixed stacking type planning of the material grouping object to generate a stacking sequence and hierarchical pose set according to order sorting to obtain a stacking type object;

[0102] A parameter solving unit 204 for solving grabbing and placing parameters using the stacking type object to obtain an execution object;

[0103] An instruction arrangement unit 205 for instruction arrangement using the execution object to generate a stacking instruction set for driving the double-finger clamp.

[0104] In the embodiment, the data modeling unit 201 acquires material parameters and performs structured modeling to obtain a material modeling object; the material grouping unit 202 performs double-grab grouping planning on the material modeling object to perform constraint checking and compatibility scoring pairing calculation, to obtain a material grouping object; the stack type planning unit 203 performs mixed stack type planning on the material grouping object to generate a stacking sequence and a hierarchical pose set according to order sorting, to obtain a stack type object; the parameter solving unit 204 performs grabbing and placing parameter solving using the stack type object to obtain an execution object; and the instruction arrangement unit 205 performs instruction arrangement using the execution object to generate a stacking instruction set for driving a double-finger clamp.

[0105] In an embodiment, the data modeling unit 201 is specifically configured to:

[0106] acquire original parameters of the material to obtain an original parameter set;

[0107] establish a mapping relationship of attribute fields, constraint fields, and positioning fields using the original parameter set to obtain a structured data set;

[0108] perform clamp capability checking on the structured data set to obtain a compatibility checking result;

[0109] fuse stacking constraints on the compatibility checking result, and associate an identification code with a three-dimensional coordinate as a positioning parameter to obtain a regularized object;

[0110] objectize and encapsulate the regularized object to obtain a material modeling object.

[0111] In an embodiment, the material grouping unit 202 is specifically configured to:

[0112] perform candidate item extraction on the material modeling object to obtain a candidate material set;

[0113] perform weight screening processing on the candidate material set to obtain a weight screening set;

[0114] generate size screening annotations using the weight screening set based on length determination rules to obtain a size screening set;

[0115] perform queue classification on the size screening set to obtain a queue object; wherein the queue object includes a double-grab queue to be grouped and a stacking queue to be stacked;

[0116] perform constraint checking on the queue object to generate candidate pairs for materials in the double-grab queue to be grouped in pairs, while outputting compatibility scores, and to integrate to obtain a material grouping object.

[0117] In an embodiment, the material grouping unit 202 is further specifically configured to:

[0118] specification analysis on the double-grabbing queue in the queue object to be grouped to generate a specification analysis object;

[0119] grouping with the same specification using the specification analysis object to obtain a same-specification grouping object;

[0120] pairing candidate set generation by two-by-two combination enumeration and difference characteristic scoring on the same-specification grouping object to be grouped;

[0121] combination association constraint checking and compatibility score summarization on the pairing candidate set respectively to obtain a compatibility score object;

[0122] pairing selection using the compatibility score object to update the output material grouping object.

[0123] In an embodiment, the stacking planning unit 203 is specifically configured to:

[0124] merge each pair of grouped materials in the material grouping object into a single material and calculate the merging parameters to obtain a merged material set;

[0125] order arrangement of the merged materials in the merged material set according to the order sequence to obtain an order sequence set;

[0126] level rule calculation on the order sequence set to obtain a level sorting set;

[0127] assigning materials with similar heights in the level sorting set to the same layer and calculating the placement orientation and position of the merged materials in each layer to obtain a level pose set;

[0128] objectification packaging of the level pose set to obtain a stacking object.

[0129] In an embodiment, the parameter solving unit 204 is specifically configured to:

[0130] grasp point positioning calculation according to the stacking object to obtain a grasp detection object;

[0131] distance and grasp pose calculation using the grasp detection object to obtain a grasp parameter object;

[0132] safety opening and closing distance calculation using the grasp parameter object to obtain a safety parameter object;

[0133] grasp process control planning on the safety parameter object to obtain a grasp control object;

[0134] lifting and parameterization packaging of the grasp control object to generate an execution object.

[0135] In an embodiment, the instruction arrangement unit 205 is configured to:

[0136] read the placement parameters of the execution object to obtain a set of placement parameters;

[0137] plan the initial action of the set of placement parameters to output an initial path segment;

[0138] generate a parabolic path and solve the obstacle avoidance using the initial path segment, and simultaneously optimize the path to avoid obstacles to obtain a set of placement paths;

[0139] solve the deceleration zone and placement action using the set of placement paths to output a placement execution sequence;

[0140] plan the off-site and follow-up process of the placement execution sequence to obtain a set of stacking instructions for driving the double-finger clamp.

[0141] Since the embodiments of the device part correspond to the embodiments of the method part, the embodiments of the device part are described in the description of the embodiments of the method part, and will not be described here.

[0142] The embodiments of the application also provide a computer readable storage medium, which has a computer program stored thereon, and the computer program can implement the steps provided by the above embodiments when executed. The storage medium can include: a U disk, a mobile hard disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a magnetic disk or an optical disk, and various storage program codes.

[0143] The embodiments of the application also provide a computer device, which can include a memory and a processor, the memory has a computer program stored therein, and the processor can implement the steps provided by the above embodiments when calling the computer program in the memory. Of course, the computer device can also include various network interfaces, power supplies, graphics card devices, etc., and the performance of the graphics card can be used to operate the model, such as inference and training.

[0144] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts of each embodiment can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the related parts can be referred to the method part. It should be pointed out that, for ordinary skilled in the art, without departing from the principles of the present application, the present application can be improved and modified, and these improvements and modifications also fall within the protection scope of the claims of the present application.

[0145] It also needs to be explained that in the present specification, the relational terms such as first and second and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the a stated elements.

Claims

1. A two-finger clamp palletizing method based on order sequencing, characterized by, The method comprises the following steps: acquiring material parameters and performing structured modeling to obtain a material modeling object; performing double-grab grouping planning on the material modeling object to perform constraint checking and compatibility scoring pairing calculation to obtain a material grouping object, comprising: performing candidate item extraction on the material modeling object to obtain a candidate material set; performing weight screening processing on the candidate material set to obtain a weight screening set; generating size screening labels based on length determination rules using the weight screening set to obtain a size screening set; performing queue classification on the size screening set to obtain a queue object; wherein the queue object comprises a double-grab queue to be grouped and a stacking queue to be grouped; performing constraint checking on the queue object to generate candidate pairs for the materials in the double-grab queue to be grouped in pairs, while outputting compatibility scores, and integrating to obtain a material grouping object; performing mixed stacking type planning on the material grouping object to generate stacking order and hierarchical pose sets according to order sequences to obtain a stacking type object, comprising: merging each pair of grouped materials in the material grouping object into a single material and calculating the merging parameters to obtain a merged material set; sequentially arranging the merged materials in the merged material set according to order sequences to obtain an order sequence set; performing hierarchical rule calculation on the order sequence set to obtain a hierarchical order set; assigning materials with similar heights in the hierarchical order set to the same layer and calculating the placement orientation and position of the merged materials in each layer to obtain a hierarchical pose set; objectifying and packaging the hierarchical pose set to obtain a stacking type object; solving grabbing and placing parameters using the stacking type object to obtain an execution object; performing instruction arrangement using the execution object to generate a stacking instruction set for driving a double-finger gripper.

2. The order sequencing based dual finger gripper palletizing method of claim 1, wherein, The method of acquiring material parameters and performing structured modeling to obtain a material modeling object comprises: collecting original parameters of the material to obtain an original parameter set; establishing a mapping relationship of attribute fields, constraint fields, and positioning fields using the original parameter set to obtain a structured data set; performing gripper capability checking on the structured data set to obtain a compatibility checking result; fusing stacking constraints and associating identification codes and three-dimensional coordinates as positioning parameters based on the compatibility checking result to obtain a regularized object; objectifying and packaging the regularized object to obtain a material modeling object.

3. The order sequencing based dual finger gripper palletizing method of claim 1, wherein, The method of performing constraint checking on the queue object to generate candidate pairs for the materials in the double-grab queue to be grouped in pairs, while outputting compatibility scores, and integrating to obtain a material grouping object comprises: performing specification analysis on the double-grab queue to be grouped in the queue object to generate a specification analysis object; performing same-specification priority grouping using the specification analysis object to obtain a same-specification grouping object; performing two-by-two combination enumeration and difference characteristic scoring on the to-be-grouped set in the same-specification grouping object to generate a candidate pair set; respectively performing combination relevance constraint checking and compatibility score summarization on the candidate pair set to obtain a compatibility score object; performing pair selection using the compatibility score object to update and output a material grouping object.

4. The order sequencing based dual finger gripper palletizing method of claim 1, wherein, The grabbing and placing parameter solving using the pile type object obtains an execution object, including: Grabbing point positioning calculation is performed according to the pile type object to obtain a grabbing detection object; Distance and grabbing pose calculation is performed using the grabbing detection object to obtain a grabbing parameter object; Safety opening and closing distance calculation is performed using the grabbing parameter object to obtain a safety parameter object; Grabbing process control planning is performed on the safety parameter object to obtain a grabbing control object; Lifting and parameterized packaging are performed on the grabbing control object to generate an execution object.

5. The order sequencing based dual finger gripper palletizing method of claim 1, wherein, The instruction arrangement using the execution object generates a stacking instruction set for driving the double-finger clamp, including: Placing parameter reading is performed on the execution object to obtain a placing parameter set; Initial path segment output is obtained by performing initial action planning on the placing parameter set; The parabolic path generation and obstacle avoidance solving are performed using the initial path segment, and the path is simultaneously optimized to avoid obstacles to obtain a placing path set; The deceleration zone and placing action solving are performed using the placing path set to output a placing execution sequence; The off-site and continuation process planning are performed on the placing execution sequence to obtain a stacking instruction set for driving the double-finger clamp.

6. An order sequencing based dual finger gripper palletizing device, characterized in that, It includes: A data modeling unit is configured to obtain material parameters and perform structured modeling to obtain a material modeling object; A material grouping unit is configured to perform double-grabbing grouping planning on the material modeling object to perform constraint checking and compatibility scoring pairing calculation to obtain a material grouping object; A pile type planning unit is configured to perform mixed pile type planning on the material grouping object to generate a stacking sequence and hierarchical pose set according to order sorting to obtain a pile type object; A parameter solving unit is configured to perform grabbing and placing parameter solving using the pile type object to obtain an execution object; An instruction arrangement unit is configured to perform instruction arrangement using the execution object to generate a stacking instruction set for driving the double-finger clamp; The material grouping unit is specifically configured to extract candidate items from the material modeling object to obtain a candidate material set; perform weight screening processing on the candidate material set to obtain a weight screening set; and generate size screening labels based on length determination rules using the weight screening set to obtain a size screening set; Queue classification is performed on the size screening set to obtain a queue object; wherein the queue object includes a double-grabbing queue to be grouped and a stacking queue to be stacked; constraint checking is performed on the queue object to generate candidate pairs for the materials in the double-grabbing queue to be grouped, and compatibility scores are output simultaneously, and a material grouping object is integrated; The stacking planning unit is specifically configured to combine each pair of grouped materials in the material grouping object into a single material and calculate a combination parameter to obtain a combination material set; arrange the combination materials in the combination material set in sequence according to an order sequence to obtain an order sequence set; perform hierarchical rule calculation on the order sequence set to obtain a hierarchical sorting set; distribute materials with similar heights in the hierarchical sorting set to the same layer and calculate the placement orientation and position of the combination materials in each layer to obtain a hierarchical pose set; and perform objectification packaging on the hierarchical pose set to obtain a stacking object.

7. A computer device, comprising: The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the double-finger clamp stacking method based on order sorting.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to implement the double-finger clamp stacking method based on order sorting.

Images