A container filling production line control method, system, device and medium based on whole-process closed-loop control

Abstract

The application relates to a container filling production line control method, system, equipment and medium based on a whole-process closed-loop control, and relates to the technical field of production line process control. The container filling production line control method comprises the following steps: constructing a production line coordinate system according to a liquid filling production process, and determining the coordinates of each station in the liquid filling process; planning a container filling path in the liquid filling process according to the container model and the station coordinates; monitoring the liquid filling process according to the container filling path, obtaining the running posture of the container, and calculating the container posture deviation in combination with the target posture of the station; generating a station container compensation instruction according to the process target parameters and the container posture deviation, correcting the liquid filling process, and generating re-filling execution parameters; monitoring the container conveying process according to the re-filling execution parameters, judging the mechanism execution beat, determining the mechanism execution correction strategy; and generating a process control signal according to the mechanism execution correction strategy and the abnormal station, and globally regulating and controlling the container filling production line.

Description

Technical Field

[0001] This application relates to the field of production line process control technology, and in particular to a control method, system, equipment and medium for a container filling production line based on full-process closed-loop control. Background Technology

[0002] Traditional container filling production lines typically employ segmented, independent control, lacking a unified benchmark and coordination mechanism between processes. During container transport, the conveyor system only performs simple start-stop control, failing to ensure containers enter subsequent stations with consistent posture. The cleaning stage often uses timed or fixed-stroke control, unable to dynamically adjust cleaning parameters based on the actual arrival position and condition of the containers. While vision-based positioning technology has emerged in the filling stage, it is often disconnected from preceding and following processes, preventing the effective transmission and compensation of accumulated positional deviations during transport and cleaning. This necessitates secondary adjustments after mechanical stops during filling, increasing the production cycle time per unit.

[0003] Existing patents disclose a method and system for automated quantitative filling of liquid chemicals based on dynamic verification control. The method includes: S1: setting up a dual-mode controlled automated quantitative filling production line for liquid chemicals connected to the liquid chemical supply pipeline; S2: starting one filling station of the production line for filling; S3: the front-end controller controls the flow control valve for filling, and the data obtained by each sensing unit are interactively verified; S4: if the error value E exceeds the set range, reliable data is verified, and filling continues until the target filling capacity or weight is reached. The method and system provided by the above invention, through verification and calculation of multiple channels and sets of sensor data, eliminates variable data, avoids false alarms and missed alarms, and achieves dynamic adjustment, interactive verification, accurate filling, and safe filling. It supports two working modes: fixed volume and fixed weight, and supports automated filling of various liquid chemicals and various filling containers.

[0004] The existing technical solutions mentioned above have the following drawbacks: 1. In the finished product off-line stage, only counting and transfer are usually performed, and the abnormal information of the previous process is not used for off-line path optimization, which can easily cause off-line congestion or container posture confusion. The segmented control mode results in poor overall line coordination, long changeover adjustment time, and delayed response to anomalies, making it difficult to meet the requirements of high-precision and high-efficiency automated filling. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the purpose of this application is to provide a control method, system, equipment, and medium for a container filling production line based on full-process closed-loop control. Through full-process closed-loop control, the production line can adaptively handle container material deviations and equipment cumulative errors, reducing the frequency of manual intervention and production changeover adjustments, and effectively improving the overall efficiency and operational stability of container filling.

[0006] This was achieved using the following technical solutions: Firstly, this application provides a control method for a container filling production line based on full-process closed-loop control, including: Construct a production line coordinate system based on the liquid filling production process to determine the coordinates of each workstation during the liquid filling process; Based on the container model and workstation coordinates, plan the container filling path during the liquid filling process; Based on the container filling path, monitor the liquid filling process, obtain the container running posture, and calculate the container posture deviation by combining it with the target posture of the workstation. Based on the process target parameters and the container position deviation, generate station container compensation instructions, correct the liquid filling process, and generate refill execution parameters. Monitor the container conveying process based on the refilling execution parameters, determine the mechanism's execution cycle time, and identify mechanism execution correction strategies. Based on the mechanism's execution correction strategy and abnormal workstations, process control signals are generated to globally regulate the container filling production line.

[0007] By adopting the above technical solution, a production line coordinate system is constructed based on the liquid filling process. The filling path is planned by the container model, the container posture is monitored in real time and compared with the target posture. After calculating the posture deviation, a compensation command is generated. The refilling parameters are adjusted in combination with the process target. At the same time, the conveying cycle correction mechanism strategy is monitored, and global process control signals are generated for abnormal workstations. This realizes adaptive collaborative control and high-precision dynamic correction of the filling process, which significantly improves production flexibility and finished product qualification rate.

[0008] This application further specifies: constructing a production line coordinate system based on the liquid filling production process, and determining the coordinates of each station during the liquid filling process, including: The liquid filling production process is broken down according to the container conveying direction, the production and processing stations are determined, and corresponding station labels are assigned. Based on the container conveying direction and the container conveying starting point, establish a global reference origin, and construct a production line coordinate system in conjunction with the production and processing station; Based on the workstation layout diagram, the liquid filling production process is deconstructed to determine the straight conveying section and the non-straight conveying section of the workstation. Based on the production line coordinate system and workstation markings, the linear conveyor section of the workstation is located, and the coordinates of each workstation in the linear section are determined. Based on the rotation center radius and the filling phase angle, a virtual mapping is performed on the non-linear conveying section of the workstation to determine the coordinates of each workstation in the non-linear section.

[0009] By adopting the above technical solution, a global production line coordinate system is established based on process decomposition and reference origin positioning algorithm. The coordinates of the workstations on straight lines and non-straight lines are determined by linear mapping and virtual mapping (using the rotation center radius and filling phase angle), respectively. This achieves systematic and accurate positioning of the filling workstations and significantly improves the integrity and adaptability of the coordinate system construction.

[0010] This application further specifies: planning the container filling path during the liquid filling process based on the container model and workstation coordinates, including: The container to be filled is scanned and inspected to acquire a 3D point cloud of the container and extract its appearance features. The container appearance features are matched according to the preset container parameter library to determine the container model and match the container dynamic constraint spacing. Based on the container model and workstation coordinates, a filling simulation is performed on the container to be filled to generate the container arrival sequence. The static clearance of the container is calculated based on the workstation coordinates and workstation boundaries, and then compared with the preset clearance safety threshold. If the static clearance of the container is less than the clearance safety threshold, it is determined that there is an interference risk at the current workstation and it is marked as a static constraint segment; The distance between transfer containers is calculated based on the conveyor baseline speed and the inherent cycle time of the workstation. If the distance between transfer containers is greater than the dynamic constraint distance between containers, it is determined that there is a speed mismatch risk at the current workstation and it is marked as a global buffer segment; Based on the container arrival sequence, the static constraint segment and the global buffer segment are segmented and interpolated and globally merged to generate the container filling path.

[0011] By adopting the above technical solution, dynamic constraint spacing is obtained by matching container models through 3D scanning, filling is simulated to generate the arrival sequence, static gap and transfer spacing are calculated to identify interference and speed mismatch risks and mark constraint segments, and finally, a safe and efficient container filling path is generated through segmented interpolation and global fusion, which significantly improves the accuracy of path planning and the stability of production line operation.

[0012] This application further specifies: based on the container filling path, monitoring the liquid filling process, obtaining the container's running posture, and combining this with the target posture of the workstation, calculating the container's posture deviation, including: The liquid filling process is monitored according to the container filling path to obtain the container running posture and extract the actual container posture, container radial coordinates and actual arrival time. The target posture of the workstation is compared with the actual posture of the container based on the workstation identification, and the posture deviation vector is calculated. Calculate the absolute value of the radial deviation based on the workstation coordinates and the container radial coordinates, and compare it with the preset tolerance threshold. If the absolute value of the radial deviation is greater than the tolerance threshold, the deviation type of the current container to be filled is determined to be a deviation state. The actual arrival time is compared with the container arrival time, and the arrival deviation time is calculated. If the arrival deviation time is positive, the deviation type of the current container to be filled between the two stations is determined to be too slow. If the arrival deviation time is negative, the deviation type of the container to be filled between the two stations is determined to be too fast. The container pose deviation is generated by temporally associating the deviation type with the container's operating posture.

[0013] By adopting the above technical solution, based on path tracking and state comparison algorithms, the pose deviation vector is calculated by comparing the actual posture with the target posture at the workstation. The radial coordinate and arrival time sequence are used to identify the type of deviation and speed anomaly. This achieves a refined, real-time quantitative assessment of pose deviation during container filling, significantly improving the accuracy and response efficiency of deviation detection.

[0014] This application further specifies: based on the process target parameters and the container posture deviation, a station container compensation instruction is generated to correct the liquid filling process and generate refill execution parameters, including: Feature identification is performed on container orientation deviations to obtain filling anomaly characteristics; these characteristics include pressure attenuation range and pressure amplitude. If the pressure attenuation range is within a preset specific angle range, the deviation type is determined to be circumferential positioning deviation. If the pressure amplitude is less than the preset pressure standard value and the pressure fluctuation is stable, the deviation type is determined to be radial positioning deviation. If both exist, the deviation type is determined to be superimposed positioning deviation, and the container pose error vector is recorded. If the pose error vectors of containers of different models are in the same direction and have the same amplitude, they are determined to be fixed deviations. If the pose error vectors of containers of a certain type N container model are in the same direction and have the same amplitude, it is determined to be an occasional deviation, and the abnormal container identifier is recorded. Calculate the deviation compensation amount based on the process target parameters combined with the abnormal container identifier and / or the container pose error vector; The deviation compensation amount is converted according to the type of actuator to obtain the execution compensation instruction; The execution compensation instructions are allocated according to the liquid filling production process to obtain the station execution instructions that include container correction parameters and signal execution timing. The container correction parameters are weighted according to the source of the deviation to obtain the container compensation parameters; Based on the container model and workstation coordinates, the signal execution timing is corrected to obtain the corrected execution timing. Based on the container compensation parameters and the execution correction sequence, the workstation execution instructions are corrected to generate workstation container compensation instructions. Based on the workstation container compensation instructions and the workstation cumulative deviation, the parameters of the liquid filling process are corrected to obtain the refill execution parameters.

[0015] By adopting the above technical solution, based on pressure feature analysis and deviation classification algorithm, circumferential, radial and superimposed positioning deviations are identified, and fixed and occasional errors are distinguished. Then, the compensation amount is calculated and converted into workstation execution instructions. Combined with time-series correction, container compensation parameters and refilling parameters are generated, realizing adaptive and precise compensation in the liquid filling process, which significantly improves filling quality and production line flexibility.

[0016] Secondly, this application also provides a container filling production line control system based on full-process closed-loop control, which adopts the following technical solution: A container filling production line control system based on full-process closed-loop control, used to implement a container filling production line control method, including: The baseline construction module is used to construct a production line coordinate system based on the liquid filling production process and determine the coordinates of each station in the liquid filling process. The path planning module is used to plan the container filling path during the liquid filling process based on the container model and workstation coordinates. The deviation positioning module is used to monitor the liquid filling process according to the container filling path, obtain the container running posture, and calculate the container posture deviation. The execution correction module is used to correct the liquid filling process based on the process target parameters and the container posture deviation, and generate refill execution parameters. The strategy reconfiguration module is used to monitor the container conveying process based on the refilling execution parameters, determine the mechanism execution cycle, and determine the mechanism execution correction strategy. The global control module is used to generate process control signals based on the mechanism's execution correction strategy and abnormal workstations, and to globally control the container filling production line.

[0017] By adopting the above technical solution, constructing a production line coordinate system and planning the filling path, using real-time attitude monitoring and posture deviation calculation algorithms to generate compensation instructions, correcting the filling process and reconstructing execution parameters, and monitoring the conveying cycle time judgment mechanism correction strategy, combined with abnormal workstations to generate global process control signals, the adaptive collaborative control of the filling production line is realized, significantly improving production flexibility and finished product qualification rate.

[0018] Thirdly, this application also provides an electronic device, comprising: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by one or more processors, the one or more processors implement any of the methods in the above scheme.

[0019] Fourthly, this application also provides a storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the container filling production line control method based on full-process closed-loop control as described above.

[0020] In summary, the beneficial technical effects of this application are as follows: By extending the coordinate origin of the filling station to a unified benchmark for the entire line, coordinated control and cross-process compensation for deviations in the transportation, cleaning, filling, and unloading stages are achieved, significantly shortening the waiting time caused by posture adjustment. By implementing closed-loop control throughout the entire process, the production line can adaptively handle deviations in incoming container materials and cumulative equipment errors, reducing the frequency of manual intervention and production changeover adjustments, and effectively improving the overall efficiency and operational stability of container filling. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall process of the container filling production line control method in this application; Figure 2 This is a flowchart illustrating step S3 in this application; Figure 3 This is a schematic diagram of the container filling production line control system in this application. Detailed Implementation

[0022] The present application will be further described in detail below with reference to the accompanying drawings.

[0023] Reference Figure 1 This application discloses a container filling production line control method based on full-process closed-loop control, comprising: S1: Construct a production line coordinate system based on the liquid filling production process to determine the coordinates of each workstation during the liquid filling process; S2: Based on the container model and workstation coordinates, plan the container filling path during the liquid filling process; S3: Based on the container filling path, monitor the liquid filling process, obtain the container running posture, and calculate the container posture deviation in combination with the target posture of the workstation; S4: Based on the process target parameters and container position deviation, generate station container compensation instructions, correct the liquid filling process, and generate refill execution parameters. S5: Monitor the container conveying process based on the refilling execution parameters, determine the mechanism execution cycle time, and determine the mechanism execution correction strategy; S6: Based on the mechanism's execution correction strategy and abnormal workstations, generate process control signals to globally regulate the container filling production line.

[0024] In this embodiment, the process flow of the liquid filling production line mainly includes core steps such as empty bottle feeding, cleaning / disinfection, filling, sealing, labeling, testing, packing and palletizing.

[0025] Empty bottle loading and conveying: Empty bottles are usually stacked on pallets in boxes and fed into the production line via conveyor belt. The unpacking machine removes the empty bottles from the boxes, and the empty boxes are sent to the box washing machine for cleaning and reuse.

[0026] Cleaning and disinfection: Empty bottles enter the bottle washer for internal and external rinsing and high-temperature disinfection (such as using hot water, alkaline solution, or ozone) to ensure a sterile or clean state. Some high-requirement industries (such as pharmaceuticals) also require simultaneous cleaning of pipelines and filling heads using a CIP (Cleaning in Place) system.

[0027] Filling: After cleaning and passing inspection, the bottles enter the filling machine. Different filling methods are used depending on the material characteristics (e.g., liquids, pastes, carbonated beverages, etc.). Atmospheric pressure filling: relies on the liquid's own weight to flow into the bottle, suitable for low-viscosity liquids such as milk and mineral water.

[0028] Isobaric filling: The pressure inside the bottle is equal to that in the storage tank, and the liquid flows in by its own weight. This method is used for carbonated beverages such as beer and soft drinks.

[0029] Vacuum filling: Liquid is drawn in after vacuuming inside the bottle, suitable for high-viscosity materials such as oils and syrups.

[0030] Aseptic cold filling: Low-temperature filling in an aseptic environment to retain nutrients and flavor to the maximum extent. It is often used for high-end fruit juices, coconut milk, etc.

[0031] The filling accuracy is usually controlled within ±0.5%, and a servo control system and a high-precision flow meter are used to ensure consistency.

[0032] Sealing: The bottle is sealed immediately after filling to prevent leakage and contamination. Common methods include: screw caps (such as plastic bottles), compression caps (such as glass bottles), aluminum foil sealing (such as oral liquids), and heat sealing (such as flexible bag packaging).

[0033] Labeling and coding: After sealing, the product enters the labeling machine, where labels are automatically affixed, and the coding machine prints information such as production date and batch number for easy traceability and management.

[0034] Quality Inspection: Through visual inspection systems, weighing inspections, liquid level inspections, and other methods, defective products such as those with incomplete filling, insufficient filling, poor sealing, and misaligned labels are removed to ensure the quality of products leaving the factory.

[0035] Packaging and Palletizing: Qualified products enter the case packer, are packed into cartons according to the set quantity, and are then automatically stacked onto pallets by the palletizer and sent to the warehouse to await shipment, realizing full-process automation.

[0036] In this embodiment, for an automated filling production line of 200L open-top steel drums in a chemical plant, a production line coordinate system is first constructed based on the liquid filling process, including each station such as feeding, positioning, filling, capping, and discharging, to determine the three-dimensional coordinates of each station. Then, based on the container model (steel drum) and station coordinates, the precise path for the container to enter the filling station from the conveyor roller is planned. During the filling process, the running posture of the steel drum (including the tilt angle and center offset of the drum opening flange) is monitored in real time by a top lidar and a side vision sensor, and compared with the target posture of the station (verticality 0°, center deviation ≤1mm) to calculate the container posture deviation (e.g., the drum opening tilts forward 2.3° and deviates to the left 1.5mm).

[0037] Based on the process target parameters (filling accuracy ±0.5L, filling head insertion depth 30mm) and the positional deviation, a workstation container compensation command is generated (driving the six-degree-of-freedom parallel platform to rotate the steel drum in the opposite direction by 2.3° and move it to the right by 1.5mm) to correct the filling process and calculate the refilling execution parameters (if the deviation exceeds the limit, the refilling process is triggered). At the same time, based on the refilling execution parameters, the conveying speed and arrival signal of the steel drum are monitored to determine whether the execution cycle of the filling head lifting mechanism and the conveyor roller is synchronized (if it is found that the steel drum still shakes for more than 0.2 seconds after the filling head descends to the position), and the mechanism execution correction strategy is determined (increasing damping buffer or extending the positioning delay). Finally, for abnormal workstations (such as the workstation where the filling head positioning timeout occurs), process control signals are generated (adjusting the upstream conveying speed and activating the backup positioning fixture) to achieve global control of the entire filling production line, effectively solving the problem of filling deviation and cycle mismatch caused by steel drum deformation or conveying inertia.

[0038] Preferably, step S1 includes: The liquid filling production process is broken down according to the container conveying direction, the production and processing stations are determined, and corresponding station labels are assigned. Based on the container conveying direction and the container conveying starting point, establish a global reference origin, and construct a production line coordinate system in conjunction with the production and processing station; Based on the workstation layout diagram, the liquid filling production process is deconstructed to determine the straight conveying section and the non-straight conveying section of the workstation. Based on the production line coordinate system and workstation markings, the linear conveyor section of the workstation is located, and the coordinates of each workstation in the linear section are determined. Based on the rotation center radius and the filling phase angle, a virtual mapping is performed on the non-linear conveying section of the workstation to determine the coordinates of each workstation in the non-linear section.

[0039] In this embodiment, the origin is set as follows: the starting point of the bottle infeed screw (bottle separating worm gear) or the intersection of the bottle infeed star wheel and the tangent of the conveyor belt is selected as the origin (0,0,0). This point is the starting point for the bottle to leave the irregular conveying and enter precise controlled motion.

[0040] X-axis definition: It is usually set to the main conveying direction of the bottles (to the right / left along the conveyor belt).

[0041] Y-axis definition: Perpendicular to the conveying direction (in the horizontal plane), used to describe lateral offset or rotation radius.

[0042] Z-axis definition: Vertical height direction (zero point is usually taken as the working surface of the conveyor belt or the support surface of the bottle bottom).

[0043] The core process of liquid filling is concentrated on the rotary filling table, where coordinate calculations are mainly based on the conversion of polar coordinates to rectangular coordinates.

[0044] Input parameters: number of filling heads N, rotation center radius R, and the position of the rotation center in the coordinate system (X_center, Y_center).

[0045] Determine the angular position θ1 of filling valve No. 1 (usually the first valve after the bottle filling is completed).

[0046] The polar coordinate angle of the i-th workstation is: θi = θ1 + (i-1) * (360° / N).

[0047] Convert to Cartesian coordinates: X_i = X_center + R * cos(θ_i), Y_i = Y_center + R * sin(θ_i).

[0048] Z-axis handling: The Z-axis of the filling head station is variable. The initial Z-axis is higher (to avoid the bottle mouth), and during filling, the Z-axis drops to the height of the bottle mouth (about 80% of the bottle height) and then rises again.

[0049] The bottle washing, capping, labeling, and coding stations are located in the linear conveyor section, making coordinate calculation relatively simple.

[0050] X-axis coordinate: X_n = X-coordinate of the reference star wheel outlet + cumulative pitch × n.

[0051] Pitch calculation: Pitch = conveyor belt linear speed × station interval time.

[0052] Filling valve alignment: The X coordinate must be precisely matched with the center of the bottle mouth, and the Y coordinate needs to be offset according to the bottle diameter D: Y_offset=D / 2.

[0053] Screw cap: In addition to XY positioning, the Z-axis needs to introduce torque feedback value as a virtual coordinate component to determine whether it is tightened properly.

[0054] Liquid filling often uses electronic cam coupling, so in addition to the static coordinate system, a virtual principal axis phase angle that follows the bottle also needs to be defined.

[0055] Establish mapping relationship: Linearly map the pulse value Pulse of the conveyor belt encoder to the X-axis travel.

[0056] Calibrate 1mm = X encoder pulses. When a bottle triggers the bottle-in photoelectric switch, record the current absolute position of the main spindle and create a virtual slave axis. The action triggering of each station no longer depends solely on the proximity switch, but rather on: current spindle position - initial recorded position = target coordinate distance.

[0057] Coordinate offset: The filling valve needs to dynamically track and fill the bottle based on its real-time X(t) coordinate.

[0058] If there is an angle α between the production line conveyor belt and the X-axis guide rail, a two-dimensional rotation matrix needs to be used to correct the Y-axis deviation: X'=X*cosα-Y*sinα, Y'=X*sinα+Y*cosα; Liquid level compensation coordinates: For weighing filling, the endpoint of the Z-axis descent needs to be finely adjusted based on the feedback liquid level height (micron-level Z-axis correction).

[0059] Preferably, step S2 includes: The container to be filled is scanned and inspected to acquire a 3D point cloud of the container and extract its appearance features. The container appearance features are matched according to the preset container parameter library to determine the container model and match the container dynamic constraint spacing. Based on the container model and workstation coordinates, a filling simulation is performed on the container to be filled to generate the container arrival sequence. The static clearance of the container is calculated based on the workstation coordinates and workstation boundaries, and then compared with the preset clearance safety threshold. If the static clearance of the container is less than the clearance safety threshold, it is determined that there is an interference risk at the current workstation and it is marked as a static constraint segment; The distance between transfer containers is calculated based on the conveyor baseline speed and the inherent cycle time of the workstation. If the distance between transfer containers is greater than the dynamic constraint distance between containers, it is determined that there is a speed mismatch risk at the current workstation and it is marked as a global buffer segment; Based on the container arrival sequence, the static constraint segment and the global buffer segment are segmented and interpolated and globally merged to generate the container filling path.

[0060] In this embodiment, the unique model code of the current container is obtained by extracting the container's external features (bottle height, bottle diameter, bottle mouth type) through a visual sensor or by parsing RFID / barcode signals.

[0061] Parameter matching and verification: The identification results are compared with a preset process parameter library. Matching content includes: Geometric dimensions: maximum diameter of bottle body Dmax, neck height Hneck, total height Htotal.

[0062] Dynamic constraints: maximum allowable acceleration (amax), allowable eccentricity during filling (etolerance).

[0063] If the matching fails (unknown model or missing parameters), immediately stop feeding and issue an alarm to prevent subsequent coordinate mapping from being invalid due to missing parameters.

[0064] Extract the coordinates Pi(Xi,Yi,Zi) of each workstation and the boundaries of adjacent mechanical structures.

[0065] Calculate the envelope profile of the container in each critical orientation and the minimum gap gap between it and the station fixture, guide rail, and lower end face of the filling valve.

[0066] If the gap is less than the safety threshold (e.g., 2mm), it is determined that there is an interference risk, and the section is marked as a constraint section.

[0067] Read the conveyor belt baseline speed Vbase and the inherent cycle time Tcycle of each station (especially the rotary filling station): if Vbase × Tcycle ≤ the maximum allowable clamping distance of the bottle feeding star wheel.

[0068] If the speed mismatch exceeds the gripping range of the mechanical finger, it is determined to be a speed mismatch, and the global linear speed needs to be adjusted or the buffer mode needs to be enabled.

[0069] For workstations that require reversal or flipping (such as inverted cleaning or flipped draining), check if there is sufficient safety height margin (Hsafety > Htotal + ΔH) in the Z-axis direction. If the space is insufficient, determine that the planned path cannot be executed directly.

[0070] Linear conveyor section: The posture remains vertical and upright, and the position command is linear interpolation P(t)=Pentry+Vbase⋅t⋅X^.

[0071] Rotary filling section: The posture rotates synchronously with the star wheel, the position command conforms to the circular trajectory equation, and the follow-up height compensation Z(t) of the bottle mouth and filling valve is superimposed.

[0072] Container attitude interpolation: In the transition arc between entering and exiting the star wheel, a fifth-order polynomial is used to program the bottle tilt angle change to ensure continuous angular acceleration and prevent liquid sloshing or bottle tipping.

[0073] Reference Figure 2 Preferably, step S3 includes: The liquid filling process is monitored according to the container filling path to obtain the container running posture and extract the actual container posture, container radial coordinates and actual arrival time. The target posture of the workstation is compared with the actual posture of the container based on the workstation identification, and the posture deviation vector is calculated. Calculate the absolute value of the radial deviation based on the workstation coordinates and the container radial coordinates, and compare it with the preset tolerance threshold. If the absolute value of the radial deviation is greater than the tolerance threshold, the deviation type of the current container to be filled is determined to be a deviation state. The actual arrival time is compared with the container arrival time, and the arrival deviation time is calculated. If the arrival deviation time is positive, the deviation type of the current container to be filled between the two stations is determined to be too slow. If the arrival deviation time is negative, the deviation type of the container to be filled between the two stations is determined to be too fast. The container pose deviation is generated by temporally associating the deviation type with the container's operating posture.

[0074] In this embodiment, timestamp synchronization is achieved by recording the precise timestamp treal for each sampling point based on a global master clock. Longitudinal position capture is performed by calculating the actual X-coordinate Xreal(t) of the container in a unified coordinate system using the difference between the conveyor encoder feedback value and the bottle-entry trigger edge. Lateral and attitude capture is conducted by extracting the container's Y-axis offset Yreal and deflection angle θreal from key detection frames using a smart camera or laser profilometer.

[0075] Arrival Time Record: When the container triggers the photoelectric switch at the workstation entrance, the actual arrival time, Tarrive_real, is locked.

[0076] Longitudinal position deviation calculation: ΔX(t) = Xreal(t) − Xplan(t); calculate the differential of ΔX d(ΔX) / dt for multiple consecutive sampling periods. If this value continues to increase positively, it is determined that the speed is too fast; if it continues to decrease negatively, it is determined that the speed is too slow or slipping.

[0077] Lateral offset deviation calculation: ΔY(t)=Yreal(t)−Yplan(t); if |ΔY|>the centering allowable error (e.g., ±0.5mm), it is marked as a deviation state.

[0078] Arrival time deviation calculation (specific to cleaning / filling station): ΔT = Tarrive_real − Tarrive_plan; Positive deviation (late arrival) may cause the station to wait, while negative deviation (early arrival) may cause the process to be incomplete.

[0079] Positional deviation calculation within the workstation: During the cleaning workstation stop or follow-up process, calculate the eccentricity vector E=(Xerror,Yerror) between the current center of the container and the center of the workstation fixture. Calculate the angle Δθ between the current tilt angle of the container and the vertical line.

[0080] Preferably, step S4 includes: Feature identification is performed on container orientation deviations to obtain filling anomaly characteristics; these characteristics include pressure attenuation range and pressure amplitude. If the pressure attenuation range is within a preset specific angle range, the deviation type is determined to be circumferential positioning deviation. If the pressure amplitude is less than the preset pressure standard value and the pressure fluctuation is stable, the deviation type is determined to be radial positioning deviation. If both exist, the deviation type is determined to be superimposed positioning deviation, and the container pose error vector is recorded. If the pose error vectors of containers of different models are in the same direction and have the same amplitude, they are determined to be fixed deviations. If the pose error vectors of containers of a certain type N container model are in the same direction and have the same amplitude, it is determined to be an occasional deviation, and the abnormal container identifier is recorded. Calculate the deviation compensation amount based on the process target parameters combined with the abnormal container identifier and / or the container pose error vector; The deviation compensation amount is converted according to the type of actuator to obtain the execution compensation instruction; The execution compensation instructions are allocated according to the liquid filling production process to obtain the station execution instructions that include container correction parameters and signal execution timing. The container correction parameters are weighted according to the source of the deviation to obtain the container compensation parameters; Based on the container model and workstation coordinates, the signal execution timing is corrected to obtain the corrected execution timing. Based on the container compensation parameters and the execution correction sequence, the workstation execution instructions are corrected to generate workstation container compensation instructions. Based on the workstation container compensation instructions and the workstation cumulative deviation, the parameters of the liquid filling process are corrected to obtain the refill execution parameters.

[0081] In this embodiment, the slope and fluctuation pattern of the pressure decay curve are extracted. Simultaneously, the real-time spindle phase angle (corresponding to the circumferential position of the filling rotary table) of the container is acquired.

[0082] Deviation type diagnosis: Circumferential deflection angle error: If the pressure abnormality only occurs in a specific angle range (e.g., leakage only in the 90°~120° arc segment), it is determined that there is an angular or circumferential positioning deviation between the container opening plane and the sealing gasket.

[0083] Lateral offset error: If the pressure is lower than the standard value and fluctuates stably throughout the entire filling cycle, it is determined that there is an overall eccentricity in the XY plane between the center of the filling valve and the center of the bottle mouth.

[0084] Superposition error: If both of the above characteristics exist simultaneously, the system needs to record the error vector Eseal.

[0085] Check whether the deviation vectors of N consecutive containers show the same direction and amplitude. If they are determined to be systematic deviations, execute the subsequent compensation instruction generation process. If they are sporadic deviations, only record the container ID for exclusion analysis, and do not modify the global parameters.

[0086] If the root cause is lateral deviation (i.e., the conveyor belt deviates, causing the bottle to deviate in the Y direction): allocate 70% weight to the centering guide plate adjustment during the transportation stage, and leave 30% weight for the nozzle trajectory fine-tuning at the cleaning station (as a temporary adaptation).

[0087] If the root cause is circumferential deflection (i.e., attitude twisting during bottle handover): allocate 90% weight to the bottle screw or star wheel clamping phase during the transportation stage, and leave only 10% weight for attitude re-inspection during the cleaning stage.

[0088] If the root cause is longitudinal position fluctuation (i.e., speed fluctuation causing misalignment in the time to reach the sealing position): do not assign weight to the cleaning station, and directly adjust the pressure trigger delay of the filling valve or the slope of the follow-up curve.

[0089] The pre-adjustment amount Ctrans during the transportation phase is equal to the total compensation amount × Wtrans; the correction amount Cwash during the cleaning phase is equal to the total compensation amount × Wwash. The reference coordinate offset register of the corresponding workstation is automatically updated, completing the reverse injection of the compensation command.

[0090] Preferably, steps S5 and S6 include: The conveying process of 200L steel drums is monitored in real time based on refilling execution parameters (such as a 2mm compensation for filling head insertion depth and a filling flow rate correction to 1.25L / s). The execution cycle of the filling head lifting mechanism, rotating gripper, and conveyor rollers is determined by the conveyor chain encoder and the positioning photoelectric signal. When the time difference between the filling head descent positioning signal and the steel drum stop signal exceeds the preset cycle threshold of 0.15 seconds, it is determined to be a cycle mismatch, and a mechanism execution correction strategy is determined (such as advancing the filling head descent by 0.1 seconds and increasing the electromagnetic brake delay of the conveyor rollers by 0.05 seconds). This is to address any issues arising from... For abnormal workstations where barrel deformation causes multiple positioning failures of the filling head, a process control signal is generated—a deceleration command is sent to the upstream feeding workstation (the conveying speed is reduced from 0.3m / s to 0.15m / s). At the same time, a backup visual positioning algorithm is activated to increase the number of barrel mouth feature point matching attempts, and position loop gain adjustment parameters are sent to the servo driver of the filling workstation (the proportional gain is increased from 1.2 to 1.5). This enables global control without interrupting production, ensuring consistent cycle time across all workstations on the entire filling production line and effectively preventing chain shutdowns or filling overflow accidents caused by abnormalities at a single workstation.

[0091] In this embodiment, the maximum lateral deviation Ymax_offset is recorded during the transportation stage; the final pose correction vector Ewash is recorded during the cleaning stage; and the residual deviation of the seal alignment Eseal is recorded during the filling stage.

[0092] The theoretical lower limit pose (Poseideal) is superimposed with the cumulative deviation vector to generate the actual predicted pose (Posepredicted).

[0093] Posepredicted=Poseideal+∑Δtransport+∑Δprocess.

[0094] The judgment criteria are whether the gripping point Pgrip_std falls within the accessible surface of the container and whether the mechanical finger does not interfere with the bottle body.

[0095] If unreachable, trigger strategy adjustments: Change the gripping point position: Offset by a distance δz along the bottle's axis to avoid deformed or labeled areas. Adjust the gripping angle: Rotate the gripper by Δθgrip to match the container's actual deflection angle. Change the drop path: If the container is severely tilted, plan an outward-expanding arc trajectory to avoid the guardrail.

[0096] The simulation of the adjusted grabbing action takes time Tgrip_new. It determines whether this causes subsequent containers to queue and congest in the handover area. If a conflict exists, the system reduces the grabbing speed smoothness requirement (increasing acceleration), or requests a slight 2% reduction in global linear velocity from the main controller to maintain a safe grabbing window.

[0097] Reference Figure 3A container filling production line control system based on full-process closed-loop control, applied to a container filling production line control method, includes: The baseline construction module is used to construct a production line coordinate system based on the liquid filling production process and determine the coordinates of each station in the liquid filling process. The path planning module is used to plan the container filling path during the liquid filling process based on the container model and workstation coordinates. The deviation positioning module is used to monitor the liquid filling process according to the container filling path, obtain the container running posture, and calculate the container posture deviation. The execution correction module is used to correct the liquid filling process based on the process target parameters and the container posture deviation, and generate refill execution parameters. The strategy reconfiguration module is used to monitor the container conveying process based on the refilling execution parameters, determine the mechanism execution cycle, and determine the mechanism execution correction strategy. The global control module is used to generate process control signals based on the mechanism's execution correction strategy and abnormal workstations, and to globally control the container filling production line.

[0098] In this embodiment, the reference construction module first constructs a production line coordinate system containing the three-dimensional coordinates of each station based on the liquid filling process (feeding, positioning, filling, capping, and stacking), with the center point directly below the filling head as the origin; the path planning module plans an arc path from the inlet of the conveyor roller conveyor through the positioning stop to the filling position based on the steel drum model and the station coordinates, so as to avoid interference between the drum opening and the filling head.

[0099] During the filling process, the deviation positioning module monitors the running posture of the steel drum in real time (drum opening flange plane tilt angle, center offset) using a top lidar and a side vision sensor. It compares this posture with the target posture at the workstation (verticality 0°, center offset ≤0.5mm) to calculate the posture deviation (e.g., drum opening tilt 1.8° forward, right offset 2.1mm). The execution correction module, based on the process target parameters (filling accuracy ±0.3L, filling head insertion depth 35mm) and this posture deviation, drives the six-degree-of-freedom parallel platform to rotate the steel drum in the opposite direction and translate it to the ideal position. Simultaneously, it generates refilling execution parameters. If the deviation exceeds 3° or 3mm, a refilling process is triggered, the flow rate is reduced from 1.2L / s to 0.8L / s and the filling time is extended by 0.6 seconds. The strategy reconfiguration module monitors the steel drum conveying process according to the refilling execution parameters. It determines the execution cycle of the filling head lifting mechanism, rotating gripper and roller conveyor start and stop through the conveyor chain encoder and the position photoelectric signal. When the time difference between the filling head descent signal and the steel drum stop signal exceeds the cycle threshold of 0.12 seconds, the mechanism is determined to execute the correction strategy (the filling head starts to descend 0.08 seconds earlier, and the roller conveyor electromagnetic brake delay is increased by 0.04 seconds).

[0100] For abnormal workstations where filling head positioning fails multiple times due to barrel deformation, the global control module generates process control signals: it sends a deceleration command to the upstream feeding station (the conveying speed is reduced from 0.25m / s to 0.12m / s), activates a backup vision algorithm to increase the number of feature point matching attempts, and sends position loop gain adjustment parameters (the proportional gain is increased from 1.1 to 1.4) to the servo driver of the filling station. At the same time, it sends a waiting delay signal of 0.3 seconds to the downstream capping station. This enables dynamic coordination of the cycle time of each workstation in the entire filling production line without interrupting production, effectively avoiding chain shutdowns or filling overflow accidents caused by abnormalities in a single workstation.

[0101] An electronic device, comprising: One or more processors; Memory, used to store one or more programs; When one or more programs are executed by one or more processors, the one or more processors implement any of the methods in the above scheme.

[0102] A storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the container filling production line control method as described above.

[0103] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A control method for a container filling production line based on closed-loop control throughout the entire process, characterized in that, include: Construct a production line coordinate system based on the liquid filling production process to determine the coordinates of each workstation during the liquid filling process; Based on the container model and the workstation coordinates, plan the container filling path during the liquid filling process; Based on the container filling path, monitor the liquid filling process, obtain the container running posture, and calculate the container posture deviation in combination with the target posture of the workstation. Based on the process target parameters and the container position deviation, a station container compensation command is generated to correct the liquid filling process and generate refill execution parameters.

2. The container filling production line control method based on full-process closed-loop control according to claim 1, characterized in that, The step of constructing a production line coordinate system based on the liquid filling production process and determining the coordinates of each station during the liquid filling process includes: The liquid filling production process is broken down according to the container conveying direction, the production and processing stations are determined, and corresponding station labels are assigned. Based on the container conveying direction and the container conveying starting point, a global reference origin is established, and a production line coordinate system is constructed in conjunction with the aforementioned production and processing station; Based on the workstation layout diagram, the liquid filling production process is deconstructed to determine the linear conveying section and the non-linear conveying section of the workstation. Based on the production line coordinate system and the workstation identifier, the linear conveyor section of the workstation is located, and the coordinates of each workstation in the linear section are determined. Based on the rotation center radius and the filling phase angle, the non-linear conveying section of the workstation is virtually mapped to determine the coordinates of each workstation in the non-linear section.

3. The container filling production line control method based on full-process closed-loop control according to claim 1, characterized in that, The step of planning the container filling path during the liquid filling process based on the container model and the workstation coordinates includes: The container to be filled is scanned and inspected to acquire a 3D point cloud of the container and extract its appearance features. The container appearance features are matched according to a preset container parameter library to determine the container model and match the container dynamic constraint spacing. Based on the container model and workstation coordinates, a filling simulation is performed on the container to be filled to generate the container arrival sequence. The static clearance of the container is calculated based on the workstation coordinates and workstation boundaries, and compared with a preset clearance safety threshold. If the static gap of the container is less than the gap safety threshold, it is determined that there is an interference risk at the current workstation and it is marked as a static constraint segment; The distance between transfer containers is calculated based on the conveyor baseline speed and the inherent cycle time of the workstation. If the distance between the transfer containers is greater than the dynamic constraint distance between the containers, it is determined that there is a speed mismatch risk at the current workstation and it is marked as a global buffer segment. Based on the container arrival sequence, the static constraint segment and the global buffer segment are segmented, interpolated, and globally merged to generate the container filling path.

4. The container filling production line control method based on full-process closed-loop control according to claim 1, characterized in that, The step of monitoring the liquid filling process according to the container filling path, obtaining the container's running posture, and calculating the container's posture deviation in conjunction with the target posture of the workstation includes: The liquid filling process is monitored according to the container filling path to obtain the container running posture and extract the actual container posture, container radial coordinates and actual arrival time. The target posture of the workstation is compared with the actual posture of the container based on the workstation identifier, and the posture deviation vector is calculated. Calculate the absolute value of the radial deviation based on the workstation coordinates and the container radial coordinates, and compare it with the preset tolerance threshold. If the absolute value of the radial deviation is greater than the tolerance threshold, the deviation type of the current container to be filled is determined to be a deviation state. The actual arrival time sequence is compared with the container arrival time sequence to calculate the arrival deviation time; If the arrival deviation time is positive, then the deviation type of the current container to be filled between the two stations is determined to be too slow. If the arrival deviation time is negative, the deviation type of the current container to be filled between the two stations is determined to be too fast. The container pose deviation is generated by temporally associating the deviation type with the container's operating posture.

5. The container filling production line control method based on full-process closed-loop control according to claim 1, characterized in that, The process involves generating station container compensation commands based on process target parameters and container orientation deviations, correcting the liquid filling process, and generating refill execution parameters, including: The container's orientation deviation is identified to obtain filling anomaly features; the filling anomaly features include pressure attenuation range and pressure amplitude. If the pressure attenuation range is located within a preset specific angle range, the deviation type is determined to be circumferential positioning deviation. If the pressure amplitude is less than the preset pressure standard value and the pressure fluctuation is stable, the deviation type is determined to be radial positioning deviation. If both exist, the deviation type is determined to be superimposed positioning deviation, and the container pose error vector is recorded. If the container pose error vectors of different container models are all in the same direction and have the same amplitude, they are determined to be fixed deviations. If the container pose error vectors of a certain type N container model are in the same direction and have the same amplitude, it is determined to be an occasional deviation, and the abnormal container identifier is recorded. Calculate the deviation compensation amount based on the process target parameters combined with the abnormal container identifier and / or the container pose error vector; The deviation compensation amount is converted according to the type of actuator to obtain the execution compensation instruction.

6. The container filling production line control method based on full-process closed-loop control according to claim 1 or 5, characterized in that, The process of generating station container compensation instructions based on process target parameters and container posture deviation, correcting the liquid filling process, and generating refill execution parameters also includes: The execution compensation instructions are allocated according to the liquid filling production process to obtain the station execution instructions that include container correction parameters and signal execution timing. The container correction parameters are weighted according to the source of the deviation to obtain the container compensation parameters; Based on the container model and workstation coordinates, the execution timing of the signal is corrected to obtain the corrected execution timing. Based on the container compensation parameters and the execution correction timing, the workstation execution command is corrected to generate a workstation container compensation command. Based on the workstation container compensation command and the workstation cumulative deviation, the parameters of the liquid filling process are corrected to obtain the refill execution parameters.

7. The container filling production line control method based on full-process closed-loop control according to claim 1, characterized in that, The container filling production line control method further includes: Monitor the container conveying process based on the refilling execution parameters, determine the mechanism's execution cycle time, and identify mechanism execution correction strategies. Based on the correction strategy executed by the aforementioned mechanism and the abnormal workstations, process control signals are generated to globally regulate the container filling production line.

8. A container filling production line control system based on full-process closed-loop control, used to implement the method as described in any one of claims 1-7, characterized in that, include: The baseline construction module is used to construct a production line coordinate system based on the liquid filling production process and determine the coordinates of each station in the liquid filling process. The path planning module is used to plan the container filling path during the liquid filling process based on the container model and the workstation coordinates. The deviation positioning module is used to monitor the liquid filling process according to the container filling path, obtain the container running posture, and calculate the container posture deviation. The execution correction module is used to correct the liquid filling process based on the process target parameters and the container pose deviation, and generate refill execution parameters. The strategy reconfiguration module is used to monitor the container conveying process based on the refilling execution parameters, determine the mechanism execution cycle, and determine the mechanism execution correction strategy. The global control module is used to generate process control signals based on the correction strategy executed by the mechanism and the abnormal workstations, and to globally control the container filling production line.

9. An electronic device, comprising: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the container filling production line control method as described in any one of claims 1 to 7.

10. A storage medium storing at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the container filling production line control method as claimed in any one of claims 1 to 7.

Images