Full-automatic placing and mixed packaging control system for horizontal pressing strips of photovoltaic glass stack

Abstract

The invention provides a full-automatic placing and mixed packaging control system for horizontal pressing strips of photovoltaic glass stacks, and relates to the field of glass stack packaging. The system is characterized in that a pose planning unit receives plane structure data of a glass stack, a target placement coordinate of a depression bar is obtained through pose calculation, scanning data of multiple layers of bins in a material storage unit is obtained, and material track data is obtained through a path optimization module; the material transfer unit analyzes the material track data to drive a three-degree-of-freedom precise linear motion module and a multi-claw parallel clamping mechanism to grab the materials in the material storage unit, packages a grabbing execution result into a grabbing signal and feeds the grabbing signal back to the material storage unit; and the material storage unit monitors the slot position state of each independent stock bin in real time, receives a grabbing signal fed back by the material transfer unit, and constructs a material feedback and updating mechanism in combination with a material supplementing mechanism. The technical problems that efficiency is low, positioning precision is poor, a stock bin has no data memory, and the stock bin and a grabbing unit are difficult to cooperate to meet mixed packaging are solved.

Description

Technical Field

[0001] This application relates to the field of glass stack packaging, and in particular to a fully automatic placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks. Background Technology

[0002] In traditional deep-processing packaging lines, horizontal pressure strip placement mainly employs three methods: manual, semi-automatic, and fully automatic. Manual placement relies entirely on manual labor to pick and place pressure strips according to glass dimensions and packaging specifications. It lacks a posture planning unit to calculate the glass stack's posture to determine precise placement coordinates, and a material storage unit for digital management and intelligent matching of the pressure strips. This results in extremely low efficiency (slow manual handling and high labor intensity), poor positioning accuracy (relying solely on visual judgment, making precise control impossible), and operators needing to work alongside the equipment for extended periods, leading to high safety risks and high labor costs. While semi-automatic placement utilizes simple mechanical or electric... Auxiliary equipment still requires manual intervention, resulting in limited efficiency improvements and difficulty in adapting to the multi-bar switching placement requirements of mixed packaging. Moreover, labor costs have not been effectively reduced. In addition, although fully automated placement methods (robotic arms, gantry structures, single-grip single-place or multi-grip single-place) do not require direct manual operation, the lack of collision-free path planning support from the pose planning unit in the single-grip single-place mode makes it impossible to integrate pose calculation and path optimization through a unified algorithm framework, leading to low efficiency and becoming a bottleneck in the packaging line. In the multi-grip single-place mode, the hopper unit lacks data memory function, making it impossible to achieve coordinated cooperation between multi-specification hoppers and gripping units, which is difficult to meet the needs of mixed packaging. Summary of the Invention

[0003] This application provides a fully automatic placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, which solves the technical problems of low efficiency, poor positioning accuracy, lack of data memory in the hopper, and difficulty in coordinating with the gripping unit to meet mixed packaging requirements in the prior art.

[0004] To achieve the above objectives, this application adopts the following technical solution: Firstly, a fully automatic placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks includes: a pose planning unit, which receives planar structural data of the glass stack, performs real-time pose calculation to obtain the target placement coordinates of the pressure strip, acquires scanning data of multiple layers of silos inside the material storage unit, identifies the position of an independent silo corresponding to the target placement coordinates through position conversion, and obtains material trajectory data through a path optimization module; a material transfer unit, whose input end is connected to the pose planning unit, which acquires the target placement coordinates and material transfer trajectory data, and parses the material trajectory data to drive a three-degree-of-freedom precision linear motion module and a multi-jaw parallel gripping mechanism to grasp the material inside the material storage unit, and encapsulates the grasping execution result into a grasping signal and feeds it back to the material storage unit; and a material storage unit, which is connected to a high-speed cyclic scanning unit, which monitors the slot status of each independent silo in real time, receives the grasping signal fed back by the material transfer unit, and constructs a material feedback and update mechanism in conjunction with a replenishment mechanism.

[0005] Based on the above technical solution, in the fully automatic placement and mixed packaging control system for photovoltaic glass stack horizontal pressure strips provided in this application, the combination of the pose planning unit and the material storage unit can effectively integrate glass stack pose calculation, intelligent silo matching, and collision-free path planning into a unified algorithm framework. This not only enables adaptive decision-making based on real-time data, but also completely "digitizes" the physical silo through the hierarchical architecture (perception layer, abstraction layer, collaboration layer, and visualization layer) of the material storage unit, achieving real-time transparent management and intelligent decision support for material status. Furthermore, by cooperating with the material transfer unit, the material transfer unit and the pose planning unit are deeply integrated, serving as a key information feedback source. This deeply integrates material flow and information flow, forming a continuously self-updating dynamic control closed loop of planning, execution, and feedback. This effectively solves the shortcomings of traditional manual, semi-automatic, and fully automatic packaging methods in terms of efficiency, accuracy, flexibility, continuity, and labor costs.

[0006] In conjunction with the first aspect mentioned above, in one possible implementation, the pose planning unit is used to receive the planar structural data of the glass stack, perform real-time pose calculation, and obtain the target placement coordinates of the pressure strip. Specifically, this includes: receiving the planar structural data of the glass stack, extracting the length, width, height, number of pillars N, and the relative position offset (Δx) of each pillar. i ,Δy i） Where i∈[1,N]; Based on the planar structure data, a two-dimensional planar coordinate system is established with the center of the glass stack as the origin, the length direction as the X-axis, and the width direction as the Y-axis; and a position calibration processing algorithm is used. Calculate the actual coordinates of each column base, where d k Δx is the preset gap between the k-th column and the (k+1)-th column. iand Δy i Real-time acquisition and calibration are performed via a laser scanning unit; a standard offset reference value δ is pre-set according to the standard bonding specifications between the pressure strip and the glass stack. xbase ,δ ybase The offset of the pressure bar (δ) is obtained by real-time correction of the reference offset based on the gripper parameters of the material transfer unit. x ,δ y） The offset of the pressure strip (δ) is superimposed on the column base coordinates. x ,δ y） Generate the final target placement coordinates.

[0007] In conjunction with the first aspect mentioned above, in one possible implementation, the process of acquiring scan data of multiple layers of silos within a material storage unit and identifying the location of an independent silo corresponding to the target placement coordinates through position transformation specifically includes: acquiring structured scan data of the multiple layers of silos; uniquely numbering each layer and slot within each layer of the multiple layers of silos; extracting the layer number and slot number of the slot to form a unique two-dimensional array [layer number, slot number]; mapping the two-dimensional array set to a set of virtual coordinate points corresponding one-to-one with the physical silos; and converting the physical dimensions into actual displacements in the coordinate system through a calibration algorithm; acquiring the two-dimensional array (x 层， y 槽 ) and the corresponding target placement coordinates (x 目标， y 目标 ), through the nearest neighbor matching algorithm Calculate the Euclidean distance between the target placement coordinates and the two-dimensional array points. Record the two-dimensional array points that satisfy the minimum distance threshold θ as the target silo positions, and output the three-dimensional array [silo position, number of layers, number of slots] of the independent silos corresponding to the target placement coordinates.

[0008] In conjunction with the first aspect mentioned above, in one possible implementation, the process of obtaining material trajectory data through the path optimization module specifically includes: obtaining the target placement coordinates P of the j-th material in the pose planning unit. j target和 The coordinates B of the corresponding three-dimensional array j window And the initial position pos of the material i(t） Using the spatiotemporal occupancy grid method through the conflict cost function Predict potential motion path intersections and conflict costs C collision pos i(t） pos represents the position of cell i at time t. k(t） This represents the position of element k at time t, where T represents the total time, and d safeAs a preset safety distance, I is an indicator function; with minimizing the total movement time and conflict cost as multiple objectives, a constrained mixed integer linear programming algorithm is used to solve the problem, generating collision-free material transfer trajectory data and sending it to the corresponding material transfer unit.

[0009] In conjunction with the first aspect mentioned above, in one possible implementation, the process of parsing material trajectory data to drive a three-degree-of-freedom precision linear motion module and a multi-jaw parallel gripping mechanism to grasp materials inside the material storage unit, and encapsulating the grasping execution result as a grasping signal to feed back to the material storage unit, specifically includes: the material transfer unit receiving trajectory data and a three-dimensional array of the target silo, combining the required quantity and specification type of the pressure strip in the glass stack structure data, and generating a grasping scheme through a grasping strategy decision function; the grasping scheme being passed to a multi-task switching control logic based on a state machine model; the multi-task switching control logic parsing the grasping scheme to obtain the gripper configuration parameters, and switching to the corresponding running state, including idle state, initialization state, probe state, grasping state, placement state, and abnormal handling state; adjusting the actual parameters of the gripper through the gripper configuration parameters to drive the coordinated action of three sets of mechanical side-pulling grippers, wherein the gripper configuration parameters include clamping force, opening and closing stroke, and response time; and using a gripper status monitoring algorithm to detect the grasping success rate and abnormal state of each gripper in real time, and encapsulating the grasping execution result as a grasping signal to feed back to the material storage unit.

[0010] In conjunction with the first aspect mentioned above, in one possible implementation, the process of transmitting the gripping scheme to the multi-task switching control logic of the three-degree-of-freedom precision linear motion module and the multi-jaw parallel clamping mechanism specifically includes: receiving the gripping scheme through the real-time data interaction channel between the pose planning unit and the material storage unit; the gripping scheme including target placement coordinates, a three-dimensional array, and gripping parameters; adjusting the clamping parameters of the three sets of mechanical side-pulling grippers by querying the gripper configuration mapping table according to the pressure bar specification type in the gripping scheme, and ensuring that the motion synchronization error of the three sets of grippers is within the allowable range through a parallel clamping synchronization algorithm; and using displacement sensors and pressure sensors... The system collects the gripper position and clamping force in real time, compares them with the corresponding clamping parameters in the gripper configuration mapping table to generate a deviation signal, compares the deviation signal with the deviation threshold to generate a gripping signal. If the deviation signal exceeds the deviation threshold, a gripping failure signal is sent to the material storage unit through the communication interface layer; otherwise, a gripping success signal is sent. The gripping signal includes a timestamp, target silo number, and deviation amount. Based on the target placement coordinates and three-dimensional array in the gripping scheme, the system calculates the joint angular displacement of the three-degree-of-freedom precision linear motion module using an inverse kinematics algorithm. The joint angular displacement, gripper clamping parameters, and gripping signal are summarized to generate multi-task switching control logic.

[0011] In conjunction with the first aspect mentioned above, in one possible implementation, the material storage unit specifically includes: a material status perception layer, used to collect the slot status data of each independent silo in real time through a high-speed cyclic scanning unit. The slot status data includes the material presence status, material type code, and slot physical coordinates; a data abstraction and mapping layer, connected to the material status perception layer, used to abstract the physical silo space into a two-dimensional array model and convert the slot physical coordinates into virtual coordinates in the system coordinate system through a calibration conversion algorithm; a collaborative control layer, connected to the data abstraction and mapping layer and the material transfer unit, used to generate a dynamic grasping strategy based on the grasping signal of the material transfer unit and the pressure bar demand data of the pose planning unit, and trigger a replenishment warning based on the material inventory prediction model; and a visualization management and control layer, connected to the collaborative control layer, used to render the silo status view in real time on the HMI interface. The view includes the material inventory, type identification, and warning status of each slot, and provides a manual intervention interface.

[0012] In conjunction with the first aspect mentioned above, in one possible implementation, the material feedback mechanism process specifically includes: real-time acquisition of slot status data of each independent silo via a high-speed cyclic scanning unit. The slot status data includes the material presence status indicating whether material is present in each slot, the material type code identifying the material specification type, and the physical coordinates of the slot's three-dimensional coordinates. The material storage unit receives grab signals sent by the material transfer unit, parses the information contained in the grab signals, and updates the silo status based on the grab signals. A material inventory estimate is obtained by calculating replenishment demand based on a material inventory prediction model. When the material inventory estimate is lower than the minimum inventory requirement, a replenishment warning signal is generated, and a time series analysis method is used to predict future material demand in conjunction with historical grab data. The grab strategy is adjusted based on the real-time silo status and the replenishment warning signal. The grab strategy includes automatically switching to a backup silo when the main silo is short of material, and optimizing the grab sequence and path based on the material inventory ratio of each silo.

[0013] In conjunction with the first aspect mentioned above, one possible implementation also includes a digital definition unit: A high-speed cyclic scanning unit identifies and assigns a unique material type code to each specification of pressure strip in real time, establishing a dynamic mapping relationship between the material type code and the physical silo slot; the target pressure strip type code contained in the specification switching instruction of the glass stack is parsed, and the set of available slots matching the target type code in each silo is queried to generate a slot activation strategy; based on the slot activation strategy and the glass stack planar structure data output by the pose planning unit, the path and gripper parameters of the material transfer unit are reconstructed to support the sequential grabbing of pressure strips of different specifications from different silos; after the specification switching is completed, the high-speed cyclic scanning unit verifies the consistency between the type code of the actually grabbed pressure strip and the target type code. If they are inconsistent, an abnormal handling state is triggered, and a type change failure message is displayed through the visual control layer.

[0014] Secondly, a fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks is provided, comprising: a communication unit and a processing unit; the communication unit is used to receive planar structure data of the glass stack; the processing unit is used to perform real-time pose calculation on the planar structure data of the glass stack to obtain the target placement coordinates of the pressure strip, acquire scanning data of multi-layer silos inside the material storage unit, identify the position of the independent silo corresponding to the target placement coordinates through position conversion, obtain material trajectory data through a path optimization module, the input end of the material transfer unit is connected to the pose planning unit to acquire the target placement coordinates and the material transfer trajectory data, and parse the material trajectory data to drive a three-degree-of-freedom precision linear motion module and a multi-jaw parallel clamping mechanism to grasp the material inside the material storage unit, and encapsulate the grasping execution result into a grasping signal to feed back to the material storage unit, the material storage unit is connected to a high-speed cyclic scanning unit to monitor the slot status of each independent silo in real time, and receive the grasping signal fed back by the material transfer unit, and construct a material feedback and update mechanism in conjunction with a replenishment mechanism.

[0015] Thirdly, this application provides a fully automated placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks, comprising: a processor and a storage medium; the storage medium includes instructions, and the processor is used to execute the instructions to implement the system as described in the first aspect and any possible implementation thereof. This fully automated placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks can be an electronic device or a chip within an electronic device.

[0016] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a fully automatic placement and mixed packaging control device for horizontal pressure strips of a photovoltaic glass stack, cause the device to perform the system as described in the first aspect and any possible implementation thereof.

[0017] Fifthly, this application provides a computer program product containing instructions that, when the computer program product is run on a fully automatic placement and mixed packaging control device for horizontal pressure strips of a photovoltaic glass stack, causes the fully automatic placement and mixed packaging control device for horizontal pressure strips of a photovoltaic glass stack to perform the system as described in the first aspect and any possible implementation thereof.

[0018] This application provides a fully automated placement and mixed packaging control system for photovoltaic glass stack horizontal pressure strips. By combining a pose planning unit and a material storage unit, it effectively integrates glass stack pose calculation, intelligent silo matching, and collision-free path planning into a unified algorithm framework. This not only enables adaptive decision-making based on real-time data but also completely "digitizes" the physical silo through the layered architecture (perception layer, abstraction layer, collaboration layer, and visualization layer) of the material storage unit. This achieves real-time transparent management of material status and intelligent decision support. Furthermore, by collaborating with the material transfer unit, the system achieves deep integration of the material transfer unit and the pose planning unit, serving as a key information feedback source. This deeply merges material flow and information flow, forming a continuously self-updating dynamic control closed loop of planning, execution, and feedback. This effectively solves the shortcomings of traditional manual, semi-automatic, and fully automatic packaging methods in terms of efficiency, accuracy, flexibility, continuity, and labor costs.

[0019] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0020] Figure 1 A system architecture diagram of a fully automatic placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks provided in this application embodiment; Figure 2 A flowchart illustrating a fully automated placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application; Figure 3 A flowchart illustrating a fully automated placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application; Figure 4 A flowchart illustrating a fully automated placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application; Figure 5 A flowchart illustrating a fully automated placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application; Figure 6 A flowchart illustrating a fully automated placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application; Figure 7 This is a schematic diagram of a fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks, provided in an embodiment of this application. Detailed Implementation

[0021] In the description of this application, unless otherwise stated, "" means "or," for example, A / B can mean A or B. "And / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. Furthermore, "at least one" means one or more, and "multiple" means two or more. The words "first," "second," etc., do not limit the quantity or order of execution, and "first," "second," etc., do not necessarily imply differences.

[0022] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0023] To address the problems of manual placement in existing technologies, which requires operators to manually place the pressure strips at designated positions on the glass stack according to glass size and packaging specifications, resulting in low efficiency (relying on manual handling of pressure strips, which is labor-intensive and slow), insufficient positioning accuracy (relying on manual judgment of position, which cannot be precisely controlled), high safety risks (operators working next to the equipment for extended periods, leading to a higher probability of accidents), and high labor costs; semi-automatic placement utilizes simple mechanical devices or pneumatic / electric auxiliary equipment, but still requires manual intervention, resulting in limited efficiency improvement and difficulty in adapting to the needs of multiple pressure strips switching in mixed packaging, and also incurring high labor costs; fully automatic placement, while automating the gripping, positioning, and placement of pressure strips through robotic arms or gantry structures without direct manual operation, suffers from low efficiency in single-grip gripping (becoming one of the bottlenecks to improving the overall efficiency of packaging line equipment), fixed gripping in multiple gripping (the material hopper unit lacks data memory, making it impossible to achieve coordinated cooperation between multi-specification material hoppers and gripping units, making it difficult to meet the needs of mixed packaging), and long replenishment time (manual replenishment is required after each material hopper gripping, resulting in production gaps), and also... The lack of real-time transparency and intelligent decision-making capabilities in fully automated technologies, coupled with the inability to coordinate multi-specification silos and gripping units, makes it difficult to meet the technical problems of mixed packaging requirements. This application provides a fully automated placement and mixed packaging control system for photovoltaic glass stack horizontal pressure strips. This system, through the combination of a pose planning unit and a material storage unit, effectively integrates glass stack pose calculation, intelligent silo matching, and collision-free path planning into a unified algorithm framework. It not only achieves adaptive decision-making based on real-time data, but also completely "digitizes" the physical silos through the hierarchical architecture (perception layer, abstraction layer, collaboration layer, and visualization layer) of the material storage unit, realizing real-time transparent management and intelligent decision support for material status. Furthermore, it collaborates with the material transfer unit, deeply integrating the material transfer unit and the pose planning unit, and serving as a key information feedback source. This deeply integrates material flow and information flow, forming a continuously self-updating dynamic control closed loop of planning, execution, and feedback. This effectively solves the shortcomings of traditional manual, semi-automatic, and fully automated packaging methods in terms of efficiency, accuracy, flexibility, continuity, and labor costs.

[0024] like Figure 1 As shown in the embodiment of this application, a fully automatic placement and mixed packaging control system for horizontal pressure strips of photovoltaic glass stacks includes: Step 101: Pose planning unit. The pose planning unit is used to receive the planar structure data of the glass stack, perform real-time pose calculation, obtain the target placement coordinates of the pressure strip, acquire the scanning data of the multi-layer silos inside the material storage unit, identify the independent silo position corresponding to the target placement coordinates through position transformation, and obtain the material trajectory data through the path optimization module. The pose planning unit is used to process the structural information of the glass stack and the data of the storage bins. The planar structural data of the glass stack refers to the geometric parameters describing the glass stack in a two-dimensional plane, including length, width, height, number of support pillars, and the relative position offset of each support pillar. The target placement coordinates of the pressure strip refer to the precise position coordinates of the pressure strip on the glass stack, calculated from these coordinates. The scanning data of the multi-layer storage bins within the material storage unit is the status information of each slot inside the bin, collected in real time by a high-speed circulating scanning unit. Position transformation refers to the process of mapping the spatial position of the physical storage bins to virtual coordinates recognizable by the system. The independent storage bin position refers to the unique identifier position of each storage slot in the storage bin matrix. The path optimization module is the algorithm component in the pose planning unit used to generate efficient, collision-free material transfer trajectories. The material trajectory data describes the complete motion path information of the material from being picked up from the storage bin to being placed on the glass stack.

[0025] In some implementations, the pose planning unit first receives the glass stack planar structure data from the upstream system, extracts the length, width, height, number of pillars N, and relative position offset of each pillar, and establishes a two-dimensional planar coordinate system with the center of the glass stack as the origin. Simultaneously, the pose planning unit acquires the scanning data of the multi-layer silos inside the material storage unit, uniquely numbers each layer and slot of the silo, forming a two-dimensional array [number of layers, number of slots], and maps this array to a set of virtual coordinate points corresponding one-to-one with the physical silos. Using the nearest neighbor matching algorithm, the Euclidean distance between the target placement coordinates and each virtual coordinate point is calculated, and points that meet the minimum distance threshold are identified as the corresponding independent silo locations, and their three-dimensional array [silo location, number of layers, number of slots] is output. Then, the path optimization module, with the goal of minimizing the total movement time and conflict cost, uses a constrained mixed integer linear programming algorithm to generate collision-free material trajectory data, which is then sent to the material transfer unit for execution.

[0026] For example, when receiving planar structural data of a photovoltaic glass stack with a length of 3000mm, a width of 2000mm, and 4 pillars, the pose planning unit first establishes a coordinate system and calculates the actual coordinates of each pillar; at the same time, it scans the material storage unit to identify the location of the independent hopper storing the corresponding specification pressure strip (e.g., located in the 3rd slot of the 2nd layer); then the path optimization module can plan an optimal motion trajectory from the location of the hopper to the target placement point of the glass stack, ensuring that the transfer process is fast and collision-free.

[0027] The material storage unit specifically includes: The material status sensing layer is used to collect the slot status data of each independent silo in real time through a high-speed cyclic scanning unit. The slot status data includes the material presence status, material type code and slot physical coordinates. The high-speed cyclic scanning unit is a continuously operating sensing device used to acquire real-time data from the silo. The silo status data is a set of information describing the current status of each silo, including the material presence status (whether there is material), material type code (identifying material specifications), and silo physical coordinates (the actual spatial location of the silo).

[0028] The data abstraction and mapping layer connects to the material state perception layer. It is used to abstract the physical silo space into a two-dimensional array model and convert the physical coordinates of the slot into virtual coordinates in the system coordinate system through a calibration conversion algorithm. The virtual coordinates in the system coordinate system are digital position representations relative to the system origin.

[0029] The collaborative control layer connects the data abstraction and mapping layer and the material transfer unit. It is used to generate dynamic grasping strategies based on the grasping signals of the material transfer unit and the pressing strip demand data of the pose planning unit, and to trigger replenishment warnings based on the material inventory prediction model. Among them, the material inventory forecasting model is an algorithm that predicts future material demand based on historical data.

[0030] The visualization and control layer connects to the collaborative control layer and is used to render the silo status view in real time on the HMI interface. The view includes the material inventory, type identification and warning status of each slot, and provides a manual intervention interface.

[0031] The HMI (Human-Machine Interface) is a graphical user interface for human-computer interaction. The warning status is a prompt or alarm message issued by the system based on inventory levels.

[0032] For example, suppose the system has six independent silos, each with five layers, and each layer with three slots. The material status sensing layer detects material in the second slot of the third layer of silo 1 through scanning, and its type is coded as "TYPE". A The physical coordinates are (X) 100 ,Y 200 Z 300 The data abstraction and mapping layer abstracts this information into a two-dimensional array [3,2] and converts it into system virtual coordinates (10.5,20.3). At this point, the pose planning unit issues a request for "TYPE". A "Regarding the demand for the pressure bar, after receiving the demand, the collaborative control layer finds that the material is available in silo 1, and thus generates a dynamic strategy to retrieve it from position [3,2]. Simultaneously, the inventory forecasting model calculates the 'TYPE' of silo 1 based on the recent consumption rate." A "The pressure bar will fall below the safety stock level in 10 minutes, triggering a replenishment alert. All this information includes the material inventory in the 3rd layer, 2nd slot of silo #1." TYPE A"The labels and yellow warning status will be displayed in real time on the hopper status view of the HMI interface for operators to monitor."

[0033] Step 102: Material transfer unit. The input end of the material transfer unit is connected to the pose planning unit to obtain the target placement coordinates and material transfer trajectory data. It also parses the material trajectory data to drive the three-degree-of-freedom precision linear motion module and multi-jaw parallel gripping mechanism to grasp the material inside the material storage unit and encapsulates the grasping execution result into a grasping signal to feed back to the material storage unit. Among them, the three-degree-of-freedom precision linear motion module is a mechanical mechanism capable of precise linear motion in the X, Y, and Z directions. The multi-jaw parallel gripping mechanism is a grasping device equipped with multiple jaws that can operate simultaneously. The grasping execution result refers to the success or failure status information recorded after each grasping action is completed.

[0034] In some implementations, the target placement coordinates of the pressure bar and the planned material transfer trajectory data are first obtained from the pose planning unit. The material trajectory data can then be analyzed by identifying key points, speed parameters, and posture requirements along the motion path. The generated control commands after analysis drive the three-degree-of-freedom precision linear motion module to move. This module is responsible for precisely moving the multi-jaw parallel gripping mechanism in three-dimensional space, positioning it to the designated hopper slot within the material storage unit. Simultaneously, based on the analyzed parameters, the multi-jaw parallel gripping mechanism is activated, causing multiple jaws to coordinate their movements according to preset clamping force, opening and closing stroke, and other configuration parameters to perform a gripping operation on the material in the target slot. After the gripping action is completed, the success rate and abnormal status of each jaw can be detected in real time using built-in sensors (such as displacement sensors and pressure sensors). The results of this gripping operation (such as success, failure, and possible deviation) are encapsulated to form a structured gripping signal. This encapsulated gripping signal is then fed back to the material storage unit through the communication interface, allowing the material storage unit to update its internal hopper status data accordingly.

[0035] Step 103: Material storage unit. The material storage unit is connected to the high-speed circulating scanning unit to monitor the slot status of each independent silo in real time and receive the grab signal fed back by the material transfer unit. Combined with the replenishment mechanism, a material feedback and update mechanism is constructed.

[0036] In this system, each independent silo refers to multiple physically separate storage spaces. Each silo can independently store materials and can be divided into main silos and backup silos according to different needs, or all silos can be used simultaneously to store different specifications of pressing bars. The grab signal is feedback information generated by the material transfer unit after performing a grab operation, containing the grab result (success / failure, timestamp, target silo number, etc.). The replenishment mechanism is the rule and logic for automatically or prompting material replenishment when the material inventory falls below a threshold. The material feedback and update mechanism is a closed-loop control process that dynamically adjusts the silo status based on real-time data and maintains information accuracy.

[0037] The material feedback mechanism process specifically includes: The high-speed cyclic scanning unit collects the slot status data of each independent silo in real time. The slot status data includes the material presence status, which indicates whether there is material in each slot, the material type code that identifies the material specification type, and the physical coordinates of the slot with three-dimensional coordinate data of the standard slot. The material storage unit receives the grab signal sent by the material transfer unit, parses the information contained in the grab signal, and updates the silo status according to the grab signal; The material inventory is estimated by calculating the replenishment demand based on the material inventory forecasting model. When the material inventory estimate is lower than the minimum inventory requirement, a replenishment warning signal is generated. The future material demand is predicted by using time series analysis method combined with historical data. Based on real-time silo status and replenishment warning signals, the grabbing strategy is adjusted. The grabbing strategy includes automatically switching to the backup silo when the main silo is low on material, and optimizing the grabbing order and path according to the material inventory ratio of each silo.

[0038] In some implementations, a high-speed cyclic scanning unit cyclically scans each independent silo, collecting real-time status data for each slot, and continuously listens for and receives grab signals from the material transfer unit. Upon receiving a grab signal, the material storage unit immediately parses the information and updates the status of the corresponding slot based on the grab result (success or failure). (If the grab is successful, the corresponding slot is marked as idle; if the grab fails, an anomaly is recorded and an alarm may be triggered.) Simultaneously, by analyzing historical grab data and current inventory levels, the unit calculates an estimated material inventory value to form a replenishment mechanism that includes a material inventory prediction model. When the estimated value is lower than the preset minimum inventory requirement, a replenishment warning signal is automatically generated. Finally, by integrating real-time scanning data, grab feedback signals, and replenishment warning information, the material storage unit constructs a complete material feedback and update mechanism, ensuring that the system can dynamically and accurately grasp the material inventory situation and make timely adjustments based on production needs, such as activating spare silos or optimizing the grab sequence.

[0039] For example, suppose that both bins 1 and 2 store type A pressure strips. A high-speed cyclic scan shows that only one of the three slots in the working layer (e.g., layer 3) of bin 1 has material (status: "present"), while all three slots in the working layer of bin 2 are full. At this point, the material transfer unit successfully retrieves one type A pressure strip from the remaining slot in bin 1 and sends a successful retrieval signal. Upon receiving the signal, the material storage unit updates the status of that slot in bin 1 to "empty." The inventory prediction model, based on recent historical data of consuming 2 type A pressure strips per minute and the current state of bin 1 (no material in the working layer), calculates that the estimated inventory will fall below the safety stock level in one minute, immediately generating a replenishment warning. This allows for immediate adjustment of the retrieval strategy, ensuring that subsequent retrieval requests for type A pressure strips are no longer sent to bin 1 but are automatically switched to bin 2. Meanwhile, based on the current sufficient stock in hopper 2, the grabbing path is optimized, prioritizing the grabbing of the outermost slots to reduce the robotic arm's travel distance. During this process, the hopper status view on the HMI interface is updated in real time, displaying the material shortage warning for hopper 1 and the activation status of hopper 2.

[0040] Based on the above technical solution, the pose planning unit receives glass stack planar structure data, establishes a coordinate system, calculates column base coordinates using a position calibration algorithm, and superimposes real-time corrected pressure strip offsets to achieve high-precision target coordinate calculation. This ensures that the pressure strip placement position is accurately adapted to glass stacks of different specifications, thus solving the problems of insufficient positioning accuracy in traditional manual and semi-automatic methods, reliance on human visual judgment of position, and the rigidity of traditional fully automatic methods in adapting quickly to changes in glass stack size. Simultaneously, by acquiring silo scanning data, the physical silo is mapped to a virtual coordinate array. The nearest neighbor matching algorithm is used to locate the target silo, and combined with the path optimization module to plan a collision-free and time-optimal material trajectory, the optimal path can be generated and matched with the intelligent silo. This solves the problem of low efficiency in single-grab and single-place in traditional fully automatic methods, which has become a bottleneck in the packaging line, and lays the algorithmic foundation for efficient and multi-specification mixed packaging. Simultaneously, by integrating the material status perception layer and high-speed circulating scanning unit, the status of the storage slots is continuously collected; the data abstraction and mapping layer converts this data into unified virtual coordinates for the system; and the visualization control layer renders the silo status view in real time on the HMI interface, thus achieving real-time and accurate visualization control of the material status. This avoids the black-box status problems of traditional fully automated packaging, where silo units lack data memory, resulting in unclear material inventory, chaotic types, and reliance on manual inventory checks and confirmation. Furthermore, the collaborative control layer, based on a material inventory prediction model and time series analysis, can predict demand and trigger replenishment warnings in advance, supporting non-stop replenishment. This addresses the pain points of long replenishment times, production gaps, and inability to meet mixed packaging requirements, achieving production continuity and flexibility. Combined with warning signals and adjustment strategies, this enhances the system's robustness against material consumption fluctuations and prevents production line interruptions caused by localized material shortages.

[0041] In one possible implementation of the embodiments of this application, combined with Figure 1 ,like Figure 2 As shown, the pose planning unit is used to receive the planar structural data of the glass stack and perform real-time pose calculation. The target placement coordinates of the pressure strip can be obtained through the following steps 201 to 205, which are explained in detail below: Step 201: Receive the planar structural data of the glass stack, and extract the length, width, height, number of pillars N, and relative position offset (Δx) of each pillar. i ,Δy i） where i∈[1,N]; Wherein, the number of column feet N refers to the total number of columns supporting and fixing the glass stack. The relative positional offset of each column foot (Δx) i ,Δy i） It refers to the distance difference between the i-th column foot and a preset reference point (such as the center point or corner point) in the plane of the glass stack in the X and Y axes.

[0042] Step 202: Based on the planar structural data, establish a two-dimensional planar coordinate system with the center of the glass stack as the origin, the length direction as the X-axis, and the width direction as the Y-axis; Step 203: Using the position calibration processing algorithm Calculate the actual coordinates of each column base, where d k Δx is the preset gap between the k-th column and the (k+1)-th column. i and Δy i Real-time acquisition and calibration are achieved through a laser scanning unit; The actual coordinates of the column bases refer to the specific position coordinates of each support column base in the established two-dimensional plane coordinate system. The laser scanning unit refers to a sensing device that uses laser technology for non-contact distance measurement and position detection.

[0043] In some implementations, after establishing a two-dimensional plane coordinate system with the center of the glass stack as the origin, the gap d is preset according to the fixed parameters set in the packaging specifications. k This will activate the position calibration processing algorithm, accurately calculating the actual coordinate position of each column foot in the coordinate system.

[0044] Step 204: Pre-set the standard offset reference value δ according to the standard bonding specifications of the pressure strip and the glass stack. xbase ,δ ybase The offset of the pressure bar (δ) is obtained by real-time correction of the reference offset based on the gripper parameters of the material transfer unit. x ,δ y） ; Among them, the standard fitting specification refers to the standard relative positional relationship between the pressure strip and the glass stack according to the packaging process requirements. The standard offset reference value δ xbase ,δ ybase This refers to the theoretical offset distance of the pressure strip relative to the column base reference point, predetermined according to specifications. Gripper parameters refer to the specific technical specifications of the clamping mechanism, including clamping force, opening and closing stroke, dimensions, and other characteristics. Pressure strip offset (δ) x ,δ y） This refers to the offset value of the actual placement position of the pressure strip after correction.

[0045] In some implementation methods, a standard offset reference value δ is pre-set according to the standard fitting specifications of the pressure strip and the glass stack. xbase and δ ybase It acquires key information such as the physical dimensions and gripping point positions of the grippers in the material transfer unit; then, it comprehensively analyzes the gripper parameters and standard offset reference values, calculates the required offset through a specific correction algorithm, and performs real-time correction of the reference value to generate the final pressure bar offset δ applied to actual operation. x and δ y。 This ensures that the placement of the pressure strip accurately compensates for the dimensional influence of the gripper mechanism itself, achieving precise fit between the pressure strip and the glass stack.

[0046] Step 205: Add the pressure strip offset (δ) to the column base coordinates. x ,δ y） Generate the final target placement coordinates.

[0047] Based on the above technical solution, the system automatically receives and parses glass stack data packets from upstream devices (such as scanners or MES systems) through a host computer control system (such as a PLC) and data interface. It then uses a built-in algorithm to automatically extract key geometric parameters and dynamically construct a customized two-dimensional plane coordinate system based on the physical characteristics of the glass stack itself. This transforms the process from relying on manual measurement and input to fully automated, digital parameter acquisition and coordinate system establishment, solving the problems of low efficiency, large subjective errors, and poor consistency caused by relying entirely on manual observation, measurement, and estimation of glass stack dimensions and column base positions in manual and semi-automatic methods. Simultaneously, the system calculates the actual coordinates of the column bases through a position calibration processing algorithm and integrates preset gaps using a laser scanning unit to acquire and calibrate Δx in real time. i and Δy i This allows for the acquisition of extremely high-precision actual coordinates of the column bases, thus forming a primary closed loop for perception and decision-making. This solves the problem that traditional fully automated methods (such as robotic arms) may lack real-time feedback and simply follow preset programs, failing to adapt to minute deviations that occur during the actual transport and positioning of the glass stack. Simultaneously, real-time sensing and calibration ensure the adaptability and robustness of the positioning.

[0048] In one possible implementation of this application embodiment, acquiring scan data of multiple layers of silos inside the material storage unit and identifying the location of the independent silo corresponding to the target placement coordinates through position transformation can be achieved through the following steps 301 to 303, which are described in detail below: Step 301: Obtain structured scan data of multi-layer silos, assign unique numbers to each layer and each slot in the multi-layer silos, and extract the layer number and slot number of the slot to form a unique two-dimensional array [layer number, slot number]. In some implementations, after acquiring structured scanning data of multi-layer silos, the high-speed cyclic scanning unit first collects the slot status data of each independent silo in real time, including the material presence status, material type code, and slot physical coordinates. Then, each layer of the multi-layer silo and each slot on each layer can be uniquely numbered (e.g., a unique identifier consisting of the layer number and slot number is assigned to each slot). Then, the layer number and slot number of the slot are extracted to form a unique two-dimensional array [layer number, slot number]. The two-dimensional array can then be used for subsequent coordinate mapping and path planning to ensure that the material transfer unit can accurately locate and grasp the target.

[0049] Step 302: Map the two-dimensional array set to a set of virtual coordinate points that correspond one-to-one with the physical silo, and convert the physical dimensions into actual displacements in the coordinate system through a calibration algorithm; The virtual coordinate point set refers to the set of digital coordinates that convert the spatial location of the physical silo into a system-recognizable location through mathematical mapping. Physical dimensions refer to the measured length, width, and height of the silo slot in actual space. Actual displacement in the coordinate system refers to the specific distance the material transfer unit needs to move in the X, Y, and Z directions when performing a grasping operation.

[0050] In some implementations, the established two-dimensional array set is spatially transformed through a data abstraction and mapping layer, so that each two-dimensional array point corresponds one-to-one with the actual slot of the physical silo. Thus, the dimensional parameters of the physical silo (such as slot spacing, layer height, etc.) can be accurately matched with the system coordinate system through the transformation formula in the calibration algorithm. Then, the actual displacement corresponding to each virtual coordinate point is calculated, including the horizontal X and Y coordinates and the vertical Z coordinate. Finally, these displacements are integrated into the trajectory instructions that the material transfer unit can execute, ensuring that the three-degree-of-freedom precision linear motion module can accurately locate the target slot.

[0051] Step 303: Obtain the two-dimensional array (x) 层， y 槽 ) and the corresponding target placement coordinates (x 目标， y 目标 ), through the nearest neighbor matching algorithm Calculate the Euclidean distance between the target placement coordinates and the two-dimensional array points. Record the two-dimensional array points that satisfy the minimum distance threshold θ as the target silo positions, and output the three-dimensional array [silo position, number of layers, number of slots] of the independent silos corresponding to the target placement coordinates.

[0052] The nearest neighbor matching algorithm is an optimized algorithm that finds the closest matching point by calculating the Euclidean distance between two points. Euclidean distance refers to the straight-line distance between two points in two-dimensional space. The minimum distance threshold θ is a preset matching tolerance parameter used to determine the critical value for whether the coordinate points are successfully matched. The target hopper location refers to the final grabbing slot determined by the algorithm.

[0053] In some implementations, the converted two-dimensional array and the corresponding target placement coordinates are first obtained from the pose planning unit. Then, the spatial distance between the target placement coordinates and all two-dimensional array points is calculated one by one using the Euclidean distance formula in the nearest neighbor matching algorithm. Next, the calculation results are compared and filtered with the preset minimum distance threshold θ. The two-dimensional array points that meet the distance threshold condition are marked as the target silo locations. Finally, the two-dimensional coordinates of the point are expanded into a three-dimensional array [silo, layer number, slot number] containing the silo number and output to the material transfer unit.

[0054] Based on the above technical solution, by coordinating the data abstraction and mapping layer in the material storage unit and the path optimization module in the pose planning unit, real-time transparent management of the silo status and intelligent and precise matching of the grasping target are achieved. The system can monitor the material quantity and type of each slot at any time, and automatically calculate, within milliseconds, which silo, layer, and slot is most suitable for grasping material based on the placement requirements of the glass stack, generating displacement commands that can directly drive the actuator. This effectively solves the problems of silos lacking data memory and difficulty in coordinating with the grasping unit, making the silo status completely transparent to the system and enabling deep collaboration between the grasping unit and the silo.

[0055] In one possible implementation of this application embodiment, obtaining material trajectory data through the path optimization module can be achieved through the following steps 401 to 403, which are described in detail below: Step 401: Obtain the target placement coordinates P of the j-th material in the pose planning unit. j target和 The coordinates B of the corresponding three-dimensional array j window And the initial position pos of the material i(t） ; Wherein, the initial position pos of the material i(t） Describe the spatial coordinates of the i-th material at time t before the material transfer unit performs the gripping action.

[0056] Step 402: Utilize the spatiotemporal occupancy grid method through the conflict cost function. Predict potential motion path intersections and conflict costs C collision pos i(t） pos represents the position of cell i at time t. k(t） This represents the position of element k at time t, where T represents the total time, and d safe I is the preset safe distance, and I is the indicator function; Among them, the spatiotemporal occupancy grid method discretizes motion space and time into grid cells, which is used to predict the conflict of multiple object motion paths.

[0057] In some implementations, the motion space of the material transfer unit is discretized into a spatiotemporal grid according to a time series; and the position data (pos) of each material transfer unit is collected in real time. k(t） and pos i(t） Therefore, the spatial distance between any two units at each time t is calculated using the conflict cost function, when ||pos i(t) -pos k(t) ||≤d safe The time indicator function I is set to 1; the collision situation within all time steps T is accumulated and statistically analyzed; finally, the quantized collision cost C is output. collision This provides a basis for risk assessment for path optimization.

[0058] For example, when two material transfer units i and k are at t=3s, pos i(t） The coordinates are (200, 300, 500), pos k(t） With coordinates (205, 305, 495), the calculated Euclidean distance is 8.66 mm. If d is set... safe If the distance is 10mm, then the indicator function I takes the value 1; if this situation occurs at 3 times within a total duration T=10s, then the conflict cost C... collision =3.

[0059] Step 403: With minimizing the total movement time and conflict cost as the multiple objectives, a constrained mixed-integer linear programming algorithm is used to solve the problem, generating collision-free material transfer trajectory data and sending it to the corresponding material transfer unit.

[0060] The total motion time refers to the total time required for the material transfer unit to complete the entire grasping and placement operation.

[0061] In some implementations, a linear programming model is established with the goal of minimizing the total motion time and conflict cost. The motion time variable and conflict cost variable are used as optimization objectives. Physical motion constraints such as maximum speed, acceleration limit, and motion space boundary limit in the material transfer unit are introduced. The branch and bound method is used to obtain the optimal solution of the continuous problem by relaxing integer constraints. Then, the optimal integer solution is gradually approximated by integer cutting. The optimal path parameters obtained by solving are then converted into trajectory instructions that can be executed by the material transfer unit and sent to the corresponding material transfer unit for execution through the real-time data interaction channel.

[0062] Based on the above technical solution, a unified digital decision-making environment is constructed through the real-time data interaction channel between the pose planning unit and the material storage unit. Combined with the time dimension introduced by the path optimization module, the motion space is discretized into a spatiotemporal grid. The collision risk of any two units at any time is dynamically evaluated through mathematical functions. Finally, through the algorithm core of the pose planning unit, the multi-objective optimization problem is transformed into a computable mathematical model, which is efficiently solved by a constrained mixed-integer linear programming algorithm. This effectively integrates the three major modules of pose planning unit, material storage unit, and material transfer unit, forming a highly collaborative, data-driven, and self-optimizing intelligent control system.

[0063] In one possible implementation of this application embodiment, the process of parsing material trajectory data to drive a three-degree-of-freedom precision linear motion module and a multi-jaw parallel gripping mechanism to grasp materials inside the material storage unit, and encapsulating the grasping execution result into a grasping signal to feed back to the material storage unit, can be achieved through the following steps 501 to 505, which are described in detail below: Step 501: The material transfer unit receives the trajectory data and the three-dimensional array of the target silo, and generates a grasping plan by combining the required quantity and specification type of the pressure strip in the glass stack structure data through the grasping strategy decision function. The required quantity and specifications of the pressure strips refer to the total number of pressure strips needed for the current glass stack, as well as the dimensions, materials, and other attributes of different pressure strips. The grasping scheme is a set of operation instructions generated by the grasping strategy decision function, containing the target placement coordinates, a three-dimensional array, and grasping parameters.

[0064] In some implementations, trajectory data and a three-dimensional array of the target silo are first received from the pose planning unit through a real-time data interaction channel. At the same time, the number and specification types of pressure strips required in the glass stack structure data are parsed. Then, based on the state machine model, multi-condition fusion calculations are performed on the input data to dynamically generate a gripping scheme that includes target placement coordinates, gripper configuration parameters (such as gripping force and opening / closing stroke), and gripping sequence instructions.

[0065] For example, suppose the system needs to place 4 Type A pressure strips for a glass stack. After receiving the trajectory data, the material transfer unit combines the target silo's three-dimensional array [2,3,1] (i.e., the 3rd layer, 1st slot of silo 2) and the pressure strip requirements (quantity 4, specification A) in the glass stack structure data. The grasping strategy decision function calculates that it needs to be grasped twice (grab 3 strips the first time and 1 strip the second time).

[0066] Step 502: The capture scheme is passed to the multi-task switching control logic based on the state machine model; The process of transmitting the grasping scheme to the multi-task switching control logic of the three-degree-of-freedom precision linear motion module and the multi-jaw parallel gripping mechanism specifically includes: The grasping plan is received through the real-time data interaction channel between the pose planning unit and the material storage unit. The grasping plan includes the target placement coordinates, the three-dimensional array, and the grasping parameters. Based on the specifications and types of the clamping bar in the gripping scheme, the clamping parameters of the three sets of mechanical side-pulling grippers are adjusted by querying the gripper configuration mapping table, and the action synchronization error of the three sets of grippers is ensured to be within the allowable range by using a parallel gripping synchronization algorithm. The position and clamping force of the gripper are collected in real time by displacement and pressure sensors. The gripper configuration mapping table is compared with the corresponding clamping parameters to generate a deviation signal. The deviation signal is compared with the deviation threshold to generate a gripping signal. If the deviation signal exceeds the deviation threshold, a gripping failure signal is sent to the material storage unit through the communication interface layer. Otherwise, a gripping success signal is sent. The gripping signal includes a timestamp, target silo number and deviation amount. Based on the target placement coordinates and three-dimensional array in the grasping scheme, the joint angular displacement of the three-degree-of-freedom precision linear motion module is calculated by the inverse kinematics algorithm. The joint angular displacement, gripper parameters and grasping signals are summarized to generate multi-task switching control logic.

[0067] The multi-task switching control logic based on a state machine model is an algorithmic framework for managing motion processes using discrete states, dividing tasks into independent states such as idle, initialization, grasping, and placement. The gripper configuration mapping table stores the correspondence between parameters such as clamping force and opening / closing stroke and the specifications of the clamping bar. The parallel clamping synchronization algorithm is a coordination algorithm that ensures consistent timing of multiple gripper actions. Displacement and pressure sensors are used to detect the physical quantities of gripper position and clamping force, respectively. The inverse kinematics algorithm is a mathematical model that inversely derives the joint angular displacement from the target coordinates.

[0068] For example, suppose the gripping scheme requires gripping an A-type pressure bar from the hopper corresponding to the three-dimensional array [2,3,1]. The control logic queries the mapping table and finds that the required clamping force for the A-type pressure bar is 50N and the opening and closing stroke is 30mm. After adjusting the gripper parameters, the sensor detects that the actual clamping force is 48N (deviation 2N), which is lower than the threshold of 5N, so a gripping success signal is generated. At the same time, the inverse kinematics calculation is used to determine the coordinates that the module needs to move to (X=200mm, Y=150mm, Z=300mm). The state machine switches to the gripping state accordingly and drives the module and gripper to complete the action.

[0069] Step 503: The multi-task switching control logic analyzes the grasping scheme to obtain the gripper configuration parameters and switches to the corresponding running state. The running states include idle state, initialization state, probe state, grasping state, placement state and exception handling state. Among them, the operating status refers to the different working modes of the material transfer unit during the operation process, specifically including idle status (standby without tasks), initialization status (system parameters are loaded and reset), probe status (positioning the material position in the hopper), gripping status (performing gripping actions), placement status (positioning the pressure strip to the glass stack), and abnormal handling status (fault diagnosis and recovery).

[0070] In some implementations, the multi-task switching control logic first parses the pressure strip specification type in the gripping scheme and obtains the corresponding clamping parameters by querying the gripper configuration mapping table. Then, it switches to the target running state according to the current operation requirements: if the system enters the initialization state after startup or reset, it loads the module reference coordinates and the default parameters of the grippers; when a gripping command is received, the control logic switches to the probe state and accurately locates the hopper slot through sensors; after confirming the target position, it immediately switches to the gripping state and drives the three sets of grippers to synchronously grip the pressure strip according to the configuration parameters; after gripping is completed, it switches to the placement state and controls the three-degree-of-freedom precision linear motion module to move the pressure strip to the target coordinates of the glass stack; if any link detects parameter deviation or mechanical conflict, it immediately jumps to the abnormal handling state, interrupts the current task and triggers an alarm signal; the switching between each state strictly follows the transition conditions of the state machine model to ensure the atomicity and safety of task execution.

[0071] For example, assuming the system needs to grab a type B clamping strip, the multi-task control logic analyzes the grabbing scheme and reads the type B parameters (clamping force 60N, opening and closing stroke 25mm) from the mapping table. If the system is currently in an idle state, upon receiving the instruction, it first switches to the initialization state to load the module origin coordinates. Then it enters the probe state and uses the displacement sensor to locate the slot corresponding to the three-dimensional array [3,2,1] of the hopper. After confirming that the position is correct, it switches to the grabbing state and controls the three sets of grippers to simultaneously grab the clamping strip with a clamping force of 60N. After successful grabbing, it immediately switches to the placement state and drives the module to move the clamping strip to the glass stack coordinates (X=300mm, Y=200mm). If the pressure sensor detects abnormal fluctuations in the clamping force during the grabbing process (such as continuously being lower than 55N), it immediately jumps to the abnormal handling state, stops the module movement, and prompts an "insufficient gripper force" alarm.

[0072] Step 504: Adjust the actual parameters of the gripper by configuring the gripper parameters to drive the coordinated action of the three sets of mechanical side-pulling grippers. The gripper configuration parameters include clamping force, opening and closing stroke and response time. Among them, the three sets of mechanical side-pulling grippers are three sets of parallel execution mechanisms that use lateral traction to grasp materials, and achieve synchronous grasping and releasing of multiple materials through coordinated action.

[0073] In some implementations, preset parameters (including clamping force, opening and closing stroke, and response time) for the corresponding clamping strip specifications are first read from the gripper configuration mapping table, and then these configuration parameters are sent to the gripper drive system. The servo control module can then adjust the hydraulic or pneumatic output of the three sets of mechanical side-pulling grippers in real time to ensure the clamping force precisely matches the set value, while controlling the motor stroke to ensure the opening and closing positions meet requirements. During the execution of the action, displacement sensors continuously monitor the actual opening and closing stroke of each gripper, and pressure sensors synchronously collect clamping force data, comparing the actual parameters with the configuration parameters. A parallel clamping synchronization algorithm is used to fine-tune the timing of the three sets of grippers' actions, ensuring that the synchronization error of their closing or opening actions is controlled within ±0.5mm. Finally, based on the response time requirements, the timing of the drive signal transmission is optimized, enabling the three sets of grippers to complete coordinated actions within a specified time window.

[0074] Step 505: Real-time detection of the gripping success rate and abnormal status of each gripper using a gripper status monitoring algorithm, and encapsulation of the gripping execution result into a gripping signal to be fed back to the material storage unit.

[0075] The gripper status monitoring algorithm is a program that uses physical quantities collected by sensors to determine the effectiveness of the action and analyzes the gripper's operating data in real time. The gripping success rate refers to the percentage of successfully completed gripping actions out of the total number of attempts. Abnormal states include gripper timeouts, insufficient gripping force, and excessive travel limits.

[0076] In some implementations, the gripper status monitoring algorithm collects the opening and closing stroke data of each gripper in real time through displacement sensors, and continuously monitors the clamping force value through pressure sensors. The collected actual parameters are compared with the standard parameters in the gripper configuration mapping table. When the gripper is detected to be fully closed within a set time and the clamping force is stable within a preset range, it is determined to be a successful gripping; otherwise, it is marked as an abnormal state. For abnormal states, further diagnosis of the specific type is activated (such as insufficient clamping force, response timeout, or mechanical jamming), and the deviation is recorded. Subsequently, the gripping success rate statistics and abnormal state codes are encapsulated into a gripping signal, which includes a timestamp, target silo number, actual clamping force deviation, and status code. Finally, the gripping signal is fed back to the material storage unit in real time through the communication interface layer, triggering the silo slot status update.

[0077] Based on the above technical solution, the trajectory data, the three-dimensional array of the target silo, and the glass stack structure data (required quantity and specifications of pressure strips) are fused through a real-time data interaction channel between the pose planning unit, the material storage unit, and the material transfer unit. This data is then calculated by the grasping strategy decision function. Based on the specific glass stack specifications and the real-time silo inventory status, the grasping sequence, grasping quantity (e.g., multi-stage grasping), and grasping path are dynamically determined, rather than executing fixed, preset grasping actions. This improves the real-time performance, adaptability, and optimization of the overall grasping solution. Simultaneously, based on a state machine model, the multi-task switching control logic defines discrete operating states such as idle, initialization, probe, grasping, placement, and exception handling. The target coordinates are converted into joint angular displacements using an inverse kinematics algorithm, and these displacements are combined with gripper parameters and grasping signals to generate control commands. This effectively ensures the conditionality and safety of switching between task states, avoids task conflicts, and enables precise coordination between the three-degree-of-freedom precision linear motion module and the multi-jaw parallel gripping mechanism, improving system stability and response speed. Furthermore, a gripper configuration mapping table presets parameters such as clamping force, opening and closing stroke, and response time for different specifications of clamping strips. Displacement and pressure sensors collect actual parameters in real time, and a parallel clamping synchronization algorithm ensures the coordination of the three sets of grippers. Simultaneously, a gripper status monitoring algorithm diagnoses the success rate and abnormal states in real time, generating a gripping signal containing a timestamp, hopper number, and deviation, which is fed back to the material storage unit. This effectively and automatically adjusts the clamping parameters according to the clamping strip specifications, ensuring reliable gripping without damaging the material. Real-time monitoring ensures the quality of the operation and enables rapid fault location. Closed-loop feedback allows the status of the material storage unit to be updated in real time, forming a closed loop of information and material flow. This not only solves the problems of poor positioning accuracy, reliance on human experience for gripping reliability, and lack of real-time feedback and traceability mechanisms in manual and semi-automatic methods, but also addresses the issues of traditional fully automatic methods where fixed parameters cannot adapt to material differences or where batch gripping failures are difficult to detect due to a lack of effective monitoring.

[0078] In one possible implementation of this application embodiment, a digital definition unit is further included, which can be implemented through steps 601 to 604, as detailed below: Step 601: The high-speed cyclic scanning unit identifies and assigns a unique material type code to each specification of pressure strip in real time, and establishes a dynamic mapping relationship between the material type code and the physical silo slot. The dynamic mapping relationship refers to the association between material type code and physical slot that can be updated in real time with material consumption, replenishment or type change operations, rather than being fixed.

[0079] In some implementations, a high-speed cyclic scanning unit cyclically scans each slot in the silo, capturing the physical characteristics of the material within the slot (such as size, shape, or RFID tags) through sensors and extracting key parameters. Based on preset coding rules (such as a specification classification library), a unique material type code is generated for each specification of pressure bar, and this code is bound to the real-time scanning data of the corresponding slot. Then, through a data abstraction and mapping layer, the three-dimensional coordinates of the physical slot (such as [slot number, layer number, number of slots]) are converted into virtual coordinates recognizable by the system, and a mapping relationship is established with the material type code. This mapping relationship is stored in a dynamic database. When the material position in the silo changes due to grabbing or replenishment, the scanning unit updates the slot status in real time, dynamically adjusting the association between the code and the slot to ensure the accuracy and timeliness of the mapping.

[0080] Step 602: Parse the target pressure strip type code contained in the specification switching instruction of the glass stack, query the set of available slots that match the target type code in each silo, and generate a slot activation strategy; Among them, the slot activation strategy is a priority scheduling scheme dynamically generated by the system based on factors such as the distribution of available slots and the efficiency of the grabbing path, which is used to guide the material transfer unit to grab in sequence.

[0081] In some implementations, the glass stack specification switching command is received through a communication interface, and the target pressure strip type code is parsed. Then, the real-time data abstraction and mapping layer of the material storage unit can be directly queried to scan the material type codes and status of all silo slots. Slots that match the target code and are in the "material available" status are selected to form a set of available slots. Then, based on the path optimization algorithm (such as nearest neighbor matching or conflict cost evaluation), the grabbing efficiency weight of each available slot is calculated, and the slots that are closest to the current position of the material transfer unit or have the least conflict with other grabbing tasks are selected first. Finally, the slot activation strategy is generated according to the weight sorting, the grabbing order is specified (such as grabbing slots [1,3,2] of silo 1 first, and then grabbing slots [3,1,4] of silo 3), and the strategy is sent to the material transfer unit for execution.

[0082] Step 603: Based on the slot activation strategy and the glass stack planar structure data output by the pose planning unit, reconstruct the gripping path and gripper parameters of the material transfer unit to support the sequential gripping of pressure strips of different specifications from different silos. Reconstruction refers to recalculating and adjusting the original control parameters or paths based on new input conditions.

[0083] In some implementations, the explicitly defined grabbing sequence in the slot activation strategy is obtained (e.g., grabbing the A-type pressure strip of slot [1,2,1] in hopper 1 first, and then grabbing the B-type pressure strip of slot [3,1,3] in hopper 3), while simultaneously receiving the glass stack planar structure data output by the pose planning unit (e.g., glass stack length 3000mm, width 2000mm, column base coordinate list); then, based on the physical coordinates of each slot in the grabbing sequence (converted to system virtual coordinates through data abstraction and mapping layer) and the target placement coordinates of the glass stack, the motion trajectory of the material transfer unit can be recalculated using the spatiotemporal occupancy grid method. Simultaneously, based on the specifications of the clamping strips in the current gripping sequence (e.g., type A clamping strip with a thickness of 20mm, type B clamping strip with a thickness of 25mm), the gripper configuration mapping table is queried, and the clamping parameters of the three sets of mechanical side-pulling grippers are dynamically adjusted (e.g., clamping force of 50N and opening / closing stroke of 30mm are set for type A clamping strips, and clamping force of 60N and opening / closing stroke of 25mm are set for type B clamping strips). Thus, the reconstructed gripping path and gripper parameters are integrated into the multi-task switching control logic, and the material transfer unit is driven by the state machine model to execute the gripping tasks of different hoppers in sequence, ensuring path efficiency and gripping accuracy during mixed packaging.

[0084] Step 604: After the specification switch is completed, the consistency between the type code of the actual gripping strip and the target type code is verified by the high-speed cyclic scanning unit. If they are inconsistent, an abnormal handling state is triggered, and the type change failure information is displayed through the visual control layer.

[0085] Specification switching refers to the process where the required pressure strip type changes when the production line switches from processing a glass stack specification to another. The actual pressure strip type code is the physical identification code of the pressure strip read by the scanning unit after the material transfer unit performs the gripping action. The target type code is the unique identifier of the pressure strip specification explicitly required in the specification switching instruction. Anomaly handling status is a fault management mode automatically triggered when the system detects an error, including pausing operations and resetting the equipment.

[0086] In some implementations, after the material transfer unit completes the gripping and placement of the current specification strip, the specification switching process is immediately initiated. A high-speed cyclic scanning unit rapidly scans the strip just gripped onto the gripper, extracting the type code from the physical surface using image recognition or RFID technology. Simultaneously, the target type code of this switching instruction is retrieved from the cache, and a consistency check algorithm is activated to perform a precise string or numerical match between the two. If the match is successful (i.e., the actual code matches the target code), subsequent gripping tasks continue. If the match fails, an exception handling state is immediately triggered, causing the material transfer unit to pause all motion modules. The gripper will maintain its current state to prevent material drop, and an error code is sent to the collaborative control layer. Upon receiving the error signal, the collaborative control layer drives the visualization management layer to display a red warning window on the HMI interface, showing the specific information of the type change failure (e.g., "Target Type TYPE"). C Actual TYPE capture B (Please check the silo slot configuration), and lock the automatic operation mode, waiting for operator intervention.

[0087] Based on the above technical solution, by constructing a material identity database using a high-speed cyclic scanning unit combined with a material type coding system and a data abstraction mapping layer, automatic identification of pressing strip specifications and real-time transparent management of slot status can be achieved. This replaces the inefficient mode of manual recording and inventory, solving the problems of traditional systems where the silo lacks data memory function and material inventory and type rely entirely on manual visual inspection or periodic checks, easily leading to mis-grabbing, missed grabbing, and delays in type change. Simultaneously, by parsing specification switching instructions and linking the slot query algorithm and path optimization module, a dynamic slot activation strategy is generated, supporting on-demand scheduling and optimized grabbing sequence for multiple specifications of pressing strips. This improves the flexible production capacity of mixed packaging and avoids situations where fixed grabbing strategies cannot adapt to the needs of multiple specification switching, requiring machine stoppage for replenishment in a single silo mode, leading to production interruptions. Furthermore, based on the output data of the pose planning unit, the grabbing path and gripper parameters of the material transfer unit are reconstructed in real time, ensuring that the grabbing accuracy and movement trajectory of different specifications of pressing strips are conflict-free, reducing idle travel and adjustment time. This avoids the problem of fixed paths in traditional fully automated equipment such as robotic arms or gantry cranes, which cannot adaptively adjust according to changes in glass stack size. Finally, by scanning the physical code to verify it, an abnormal handling status is triggered and an alarm is triggered in the visual control layer. This effectively forms a closed-loop control of execution, verification and feedback, preventing replacement errors from flowing to subsequent stages.

[0088] The above primarily describes the solutions of the embodiments of this application from the perspective of device implementation. It is understood that each device, for example, a fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks, includes at least one of the hardware structures and software modules corresponding to the execution of each function in order to achieve the above-mentioned functions. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software-driven hardware manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different systems to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0089] This application embodiment can divide the fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks into functional units based on the above system example. For example, each function can be divided into separate functional units, or two or more functions can be integrated into the same processing unit. The integrated unit can be implemented in hardware or software functional units. It should be noted that the unit division in this application embodiment is illustrative and only represents a logical functional division; other division methods may be used in actual implementation.

[0090] When using integrated units, Figure 7 The above embodiment shows a possible structural schematic diagram of a fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks (referred to as a fully automatic placement and mixed packaging control device 70 for horizontal pressure strips of photovoltaic glass stacks). The fully automatic placement and mixed packaging control device 70 for horizontal pressure strips of photovoltaic glass stacks includes a processing unit 701 and a communication unit 702, and may also include a storage unit 703. Figure 7 The structural diagram shown can be used to illustrate the structure of a fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks involved in the above embodiments.

[0091] when Figure 7 The schematic diagram shown illustrates the structure of the fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks involved in the above embodiments. The processing unit 701 is used to control and manage the operation of the fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks. The communication unit 702 is used for the fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks to communicate with other devices. The storage unit 703 is used to store the program code and data of the fully automatic placement and mixed packaging control device for horizontal pressure strips of photovoltaic glass stacks.

[0092] For example, communication unit 702 is used to receive planar structure data of the glass stack; The processing unit 701 is used to perform real-time pose calculation on the planar structural data of the glass stack to obtain the target placement coordinates of the pressure strip, acquire the scanning data of the multi-layer silos inside the material storage unit, identify the independent silo positions corresponding to the target placement coordinates through position transformation, and obtain material trajectory data through the path optimization module. The input end of the material transfer unit is connected to the pose planning unit to acquire the target placement coordinates and the material transfer trajectory data, and parse the material trajectory data to drive the three-degree-of-freedom precision linear motion module and the multi-jaw parallel gripping mechanism to grasp the material inside the material storage unit, and encapsulate the grasping execution result into a grasping signal to feed back to the material storage unit. The material storage unit is connected to a high-speed cyclic scanning unit to monitor the slot status of each independent silo in real time and receive the grasping signal fed back by the material transfer unit. Combined with the replenishment mechanism, a material feedback and update mechanism is constructed.

[0093] The processing unit 701 can be a processor or a controller, and the communication unit 702 can be a communication interface, transceiver, transceiver circuit, transceiver device, etc. The term "communication interface" is a general term and may include one or more interfaces. The storage unit 703 can be a memory. When the photovoltaic glass stack horizontal pressure bar fully automatic placement and mixed packaging control device 70 is a chip, the processing unit 701 can be a processor or a controller, and the communication unit 702 can be an input interface and / or an output interface, pins, or circuits, etc. The storage unit 703 can be a storage unit within the chip (e.g., a register, cache, etc.) or a storage unit located outside the chip (e.g., read-only memory (ROM), random access memory (RAM, etc.)).

[0094] The communication unit can also be called a transceiver unit. The antenna and control circuit with transceiver functions in the fully automatic placement and mixed packaging control device 70 for photovoltaic glass stack horizontal pressure strips can be considered as the communication unit 702 of the fully automatic placement and mixed packaging control device 70 for photovoltaic glass stack horizontal pressure strips, and the processor with processing functions can be considered as the processing unit 701 of the fully automatic placement and mixed packaging control device 70 for photovoltaic glass stack horizontal pressure strips. Optionally, the device in the communication unit 702 used to implement the receiving function can be considered as the communication unit, which is used to execute the receiving steps in the embodiments of this application. The communication unit can be a receiver, a receiver circuit, etc. The device in the communication unit 702 used to implement the transmitting function can be considered as the transmitting unit, which is used to execute the transmitting steps in the embodiments of this application. The transmitting unit can be a transmitter, a transmitter, a transmitting circuit, etc.

[0095] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, the disclosure, and the appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple components. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0096] Although this application has been described in conjunction with specific features and embodiments, it is apparent that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely illustrative examples of the application as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from the spirit and scope of this application. Thus, if such modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is also intended to include such modifications and variations.

Claims

1. A full-automatic placing and mixing packing control system for photovoltaic glass stack horizontal batten, characterized in that, Comprise: The pose planning unit is used for receiving the planar structure data of the glass stack for real-time pose solution, obtaining the target placement coordinates of the pressing strip, obtaining the scanning data of the multi-layer material storage unit inside the material storage unit, identifying the independent material storage position corresponding to the target placement coordinates through position conversion, and obtaining the material track data through the path optimization module. The material transfer unit is connected with the pose planning unit, used for obtaining the target placement coordinates and the material transfer track data, analyzing the material track data to drive the three-degree-of-freedom precise linear motion module and the multi-claw parallel clamping mechanism to grab the material inside the material storage unit, and encapsulating the grabbing execution result as a grabbing signal feedback to the material storage unit. The material storage unit is connected with the high-speed circulating scanning unit, used for monitoring the slot state of each independent material storage in real time, receiving the grabbing signal feedback by the material transfer unit, combining the material supplement mechanism, and constructing the material feedback and updating mechanism.

2. The full-automatic placing and mixing packing control system for photovoltaic glass stack horizontal batten according to claim 1, characterized in that, The process that the pose planning unit is used for receiving the planar structure data of the glass stack for real-time pose solution, obtaining the target placement coordinates of the pressing strip, specifically comprises: Receiving the planar structure data of the glass stack, extracting the length, width, height, number of column feet N and relative position offset (Δx i ,Δy i） of each column foot of the glass stack, wherein i∈[1,N]; Based on the planar structure data, a two-dimensional plane coordinate system is established with the center of the glass stack as the origin, the length direction as the X axis, and the width direction as the Y axis. By position calibration processing algorithm The actual coordinates of each column foot are calculated, where d k is the preset gap between the kth column foot and the k+1th column foot, Δx i and Δy i are acquired and calibrated in real time by a laser scanning unit; According to the standard offset reference value δ preset by the standard bonding specification of the rubber strip and the glass stack xbase ,δ ybase ), the reference offset is corrected in real time by combining the jaw parameters of the material transfer unit to obtain the rubber strip offset (δ x ,δ y） ; On the basis of the column foot coordinates, the pressure strip offset (δ x , δ y） ) is superimposed to generate the final target placement coordinates.

3. The full-automatic placing and mixing packing control system for photovoltaic glass stack horizontal batten according to claim 2, characterized in that, The process that the scanning data of the multi-layer material storage inside the material storage unit is obtained, and the independent material storage position corresponding to the target placement coordinates is identified through position conversion, specifically comprises: Obtain the structured scanning data of the multi-layer material storage, uniquely number each layer and each slot on the layer of the multi-layer material storage, extract the layer number and slot number of the slot to form a unique two-dimensional array [layer number, slot number]; The two-dimensional array set is mapped to a virtual coordinate point set corresponding to the physical material storage one by one, and the physical size is converted to the actual displacement amount in the coordinate system through a calibration algorithm; Obtain a two-dimensional array (x 层， y 槽 ) and the corresponding target placement coordinates (x 目标， y 目标 ), calculate the Euclidean distance between the target placement coordinates and the two-dimensional array points by the nearest neighbor matching algorithm , record the two-dimensional array points that meet the minimum distance threshold θ as the target silo position, and output the three-dimensional array [silo position, layer number, slot number] of the independent silo corresponding to the target placement coordinates.

4. The full-automatic placing and mixing packing control system for photovoltaic glass stack horizontal batten according to claim 3, characterized in that, The process that the material track data is obtained through the path optimization module, specifically comprises: Obtain the target placement coordinate P of the jth material in the pose planning unit j target和 Coordinates B corresponding to the three-dimensional array j window And the initial position pos of the material i(t） ; Using a space-time occupancy grid through a conflict cost function Predicting potential motion path intersections and conflict cost C collision , the pos i(t） represents the position of unit i at time t, pos k(t） represents the position of unit k at time t, T represents the total time, d safe is a preset safety distance, and I is an indication function; With the minimization of total motion time and conflict cost as the multi-objective, a constrained mixed integer linear programming algorithm is used for solving, and collision-free material transfer track data is generated and sent to the corresponding material transfer unit.

5. The full-automatic placing and mixing packing control system for photovoltaic glass quoin horizontal pressure strip according to claim 4, characterized in that, The process that the material track data is analyzed to drive the three-degree-of-freedom precise linear motion module and the multi-claw parallel clamping mechanism to grab the material inside the material storage unit, and the grabbing execution result is encapsulated as a grabbing signal feedback to the material storage unit, specifically comprises: The material transfer unit receives the track data and the three-dimensional array of the target material storage, combines the pressing strip demand quantity and specification type in the glass stack structure data, generates a grabbing scheme through a grabbing strategy decision function; The grabbing scheme is transmitted to the multi-task switching control logic based on the state machine model; The multi-task switching control logic analyzes the grabbing scheme to obtain the claw configuration parameters, and switches to the corresponding running state, the running state includes the idle state, the initialization state, the probe plate state, the grabbing state, the placement state and the abnormal processing state; The actual parameters of the gripper are adjusted by the gripper configuration parameters, which are used to drive the coordinated action of the three groups of mechanical side-pulling grippers, wherein the gripper configuration parameters include clamping force, opening and closing stroke and response time; The success rate of grabbing and abnormal state of each gripper are detected in real time by the gripper state monitoring algorithm, and the grabbing execution result is encapsulated as a grabbing signal and fed back to the material storage unit.

6. The full-automatic placing and mixing packing control system for photovoltaic glass quoin horizontal pressure strips according to claim 5, characterized in that, The process of transferring the grabbing scheme to the multi-task switching control logic of the three-degree-of-freedom precise linear motion module and the multi-gripper parallel clamping mechanism includes: The grabbing scheme is received through the real-time data interaction channel between the pose planning unit and the material storage unit, and the grabbing scheme includes target placement coordinates, three-dimensional array and grabbing parameters; According to the pressing strip specification type in the grabbing scheme, the clamping parameters of the three groups of mechanical side-pulling grippers are adjusted by querying the gripper configuration mapping table, and the action synchronization error of the three groups of grippers is ensured to be within the allowable range by the parallel clamping synchronization algorithm; The gripper position and clamping force are collected in real time by the displacement sensor and the pressure sensor, and the deviation signal is generated by comparing the corresponding clamping parameters in the gripper configuration mapping table; Based on the target placement coordinates and three-dimensional array in the grabbing scheme, the joint angular displacement of the three-degree-of-freedom precise linear motion module is calculated by the inverse kinematics algorithm, and the multi-task switching control logic is generated by summarizing the joint angular displacement, clamping parameters of the gripper and grabbing signal.

7. The full-automatic placing and mixing packing control system for photovoltaic glass quoin horizontal pressure strips according to claim 6, characterized in that, The material storage unit specifically includes: The material state perception layer is used to collect the slot state data of each independent bin in real time by the high-speed cyclic scanning unit, and the slot state data includes material existence state, material type code and slot physical coordinates; The data abstraction and mapping layer is connected to the material state perception layer, which is used to abstract the physical bin space into a two-dimensional array model, and convert the slot physical coordinates into virtual coordinates in the system coordinate system by the calibration conversion algorithm; The collaborative control layer is connected to the data abstraction and mapping layer and the material transfer unit, which is used to generate a dynamic grabbing strategy according to the grabbing signal of the material transfer unit and the pressing strip demand data of the pose planning unit, and trigger the material replenishment warning based on the material inventory prediction model; The visualization control layer is connected to the collaborative control layer, which is used to render the bin state view in real time on the HMI interface, and the view includes the material inventory, type identification and warning state of each slot, and provides a manual intervention interface.

8. The full-automatic placing and mixing packing control system for photovoltaic glass quoin horizontal pressure strips according to claim 7, characterized in that, The process of the material feedback mechanism includes: The slot state data of each independent bin is collected in real time by the high-speed cyclic scanning unit, and the slot state data includes material existence state for indicating whether there is material in each slot, material type code for identifying material specification type, and three-dimensional coordinate data of standard slot physical coordinates; The material storage unit receives the grabbing signal sent by the material transfer unit, analyzes the information contained in the grabbing signal, and updates the bin state according to the grabbing signal; Based on the material inventory prediction model, the material inventory estimation value is calculated, and when the material inventory estimation value is lower than the minimum inventory requirement, a material replenishment warning signal is generated, and a time series analysis method is used to predict future material demand based on historical grabbing data; According to the real-time bin state and the material replenishment warning signal, the grabbing strategy is adjusted, which includes automatically switching to the standby bin when the main bin material is insufficient, and optimizing the grabbing sequence and path according to the proportion of the material inventory of each bin.

9. The full-automatic placing and mixing packing control system for photovoltaic glass quoin horizontal pressure strips according to claim 8, characterized in that, The digital definition unit is also included: Through the high-speed cyclic scanning unit, the unique material type code of each specification is identified and assigned in real time, and a dynamic mapping relationship between the material type code and the physical bin slot is established; The target type code contained in the specification switching instruction of the glass stack is analyzed, the available slot set matching the target type code in the current bin is queried, and the slot activation strategy is generated; According to the slot activation strategy and the glass stack plane structure data output by the pose planning unit, the grabbing path and gripper parameters of the material transfer unit are reconstructed to support sequential grabbing of different specifications from different bins; After the specification switching is completed, the high-speed cyclic scanning unit checks the consistency of the actual grabbing type code and the target type code, and if they are inconsistent, an abnormal processing state is triggered, and the visual management and control layer prompts the change type failure information.

10. A full-automatic placing and mixing packing control device for photovoltaic glass stack horizontal batten, characterized in that, The device includes a communication unit and a processing unit; The communication unit is used to receive the plane structure data of the glass stack; The processing unit is used to perform real-time pose calculation on the plane structure data of the glass stack to obtain the target placement coordinates of the strip, obtain the scanning data of the multi-layer bin inside the material storage unit, identify the independent bin position corresponding to the target placement coordinates through position conversion, and obtain the material trajectory data through the path optimization module; The material transfer unit is connected to the pose planning unit, used to obtain the target placement coordinates and the material transfer trajectory data, and analyze the material trajectory data to drive the three-degree-of-freedom precision linear motion module and the multi-gripper parallel clamping mechanism to grab the material inside the material storage unit, and encapsulate the grabbing execution result as a grabbing signal feedback to the material storage unit; The material storage unit is connected to the high-speed cyclic scanning unit, used to monitor the slot state of each independent bin in real time, and receive the grabbing signal feedback from the material transfer unit, and combine the material replenishment mechanism to construct the material feedback and update mechanism.

Images