Roll gap control method for a calender system

Abstract

The application discloses a kind of roll gap control methods of calender system, and roll gap control subsystem is adjusted according to roll gap signal, diaphragm thickness signal to the roll gap distance of two film-forming pressure rollers and / or two thinning pressure rollers, and the control center of roll gap control subsystem carries out deviation rate one calculation according to the actual thickness detected by first thickness gauge component and the calibration thickness here, if deviation rate one is in the range of ±5%, data is stored in storage, and control operation is not carried out, if deviation rate one is in the range of ±5% to ±10%, then control center takes first control module and adjusts the roll gap of thinning component, if deviation rate one exceeds the range of ±10%, then control center takes second control module and adjusts the roll gap of film-forming component.The application can accurately control roll gap, improve the control performance of system, improve the stability and accuracy of system, also improve the production efficiency of calender and product quality.

Description

Technical Field

[0001] This invention relates to the field of calender control technology, and in particular to a method for controlling the roll gap in a calendering system. Background Technology

[0002] Calenders are used in industries such as calendering, film stretching, and rolling. In the control system of a calender, several structural units involve feedback control: the roll gap adjustment unit, the electrode correction unit, and the electrode tension control unit. The roll gap adjustment structure adjusts the gap between the main pressure rolls to adjust the rolling force and electrode thickness. The electrode correction structure corrects electrode misalignment, preventing phenomena such as coiling caused by misalignment that affect electrode quality. The electrode tension control system maintains stable electrode tension, ensuring the rolled electrode is straight and preventing wavy phenomena caused by uneven elongation.

[0003] The existing control system of the calender has poor roll gap control performance, which affects the control stability and accuracy of the system, and thus affects the quality of the electrode sheets. Summary of the Invention

[0004] In response to the shortcomings of existing calender control systems, the applicant provides a reasonable roll gap control method for calender systems, which improves the stability and accuracy of roll gap adjustment and ensures electrode quality.

[0005] The technical solution adopted in this invention is as follows: A method for controlling the roll gap in a calendering system includes a roll gap control subsystem that adjusts the roll gap distance between two film-forming rollers and / or two thinning rollers based on roll gap signals and film thickness signals. The control center of the roll gap control subsystem calculates a deviation rate between the actual thickness detected by a first thickness gauge and the calibrated thickness. If the deviation rate is within ±5%, the data is stored in a storage library without any control operation. If the deviation rate is between ±5% and ±10%, the control center uses a first control module to adjust the roll gap of the thinning component. If the deviation rate exceeds ±10%, the control center uses a second control module to adjust the roll gap of the film-forming component.

[0006] As a further improvement to the above technical solution: After the second control module adjusts the roller gap of the film-forming component to a deviation rate within ±10%, it switches to the first control module.

[0007] The control process of the first control module is as follows: taking the current time-series thickness data of the first thickness gauge as a reference, it retrieves the 10 time-series thickness data points before this timestamp from the storage library and combines them with the current time-series thickness data to form a time-series dataset {T1, T2, ..., T...}. 11}, the time series dataset {T1, T2, ..., T} 11The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P T I T D The parameters tsd, Mpd, ess, e(t)} are fed into the PID controller. After receiving the parameter set, the PID controller performs calculations on e(t) to obtain the position adjustment command to the servo hydraulic proportional valve of the second roll gap control actuator of the thinning component. This process takes milliseconds. The servo hydraulic proportional valve controls the second roll gap control actuator to extend or retract to adjust the thinning support roller and thus adjust the roll gap of the thinning component.

[0008] After the first control module adjusts the roll gap of the thinning component, the deviation diaphragm is thinned after being corrected by the thinning component. The control center calculates the deviation rate 2 between the actual thickness detected by the second thickness gauge and the rated thickness at this location. When the deviation rate 2 is within the range of ±6%, no correction is made to the roll gap. When the deviation rate 2 exceeds the range of ±6%, the roll gap of the thinning component is adjusted based on the thickness signal of the second thickness gauge until the deviation rate 2 is reduced to within the range of ±6%.

[0009] When the deviation rate exceeds ±6%, the time-series thickness data of the first thickness gauge is blocked. Using the current time-series thickness data of the second thickness gauge as a reference, the 10 time-series thickness data points preceding this timestamp are retrieved from the storage database and combined with the current time-series thickness data to form a time-series dataset {T1', T2', ..., T...}. 11 '}, and then the time series dataset {T1', T2', ..., T} is... 11 The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P T I T D The parameters tsd, Mpd, ess, e(t)} are fed into the PID controller. After receiving the parameter set, the PID controller performs calculations on e(t) to obtain the position adjustment command, which is sent to the servo hydraulic proportional valve of the second roll gap control actuator of the thinning component. This process takes milliseconds. The extension or retraction of the second roll gap control actuator is controlled by the servo hydraulic proportional valve to adjust the roll gap of the thinning component to within the range of ±6% of the deviation rate. After the adjustment is completed, the roll gap is detected by the second roll gap detection sensor and fed back to the PID controller. Then, the shielding of the first thickness gauge is removed, and the adjustment is performed based on the time-series thickness signal of the first thickness gauge.

[0010] When the deviation rate is within ±6%, the time series thickness data is stored in the storage library.

[0011] While the first control module adjusts the roll gap of the thinning component, it performs secondary optimization on the zebra optimized recurrent neural network algorithm (RNN-ZOA) that controls the roll gap of the thinning component.

[0012] The control process of the second control module is as follows: taking the current time-series thickness data of the first thickness gauge as a reference, it retrieves the 10 time-series thickness data points before this timestamp from the storage library and combines them with the current time-series thickness data to form a time-series dataset {T1, T2, ..., T...}. 11}, the time series dataset {T1, T2, ..., T} 11 The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P T I T D The parameters tsd, Mpd, ess, e(t)} are fed into the PID controller. After receiving the parameter set, the PID controller performs calculations on e(t) to obtain a position adjustment command. This command is then sent to the servo hydraulic proportional valve of the first roller gap control actuator of the film forming component. This process takes milliseconds. The position command is sent through the servo hydraulic proportional valve to control the first roller gap control actuator to extend or retract, thereby adjusting the film forming support roller and thus adjusting the roller gap of the film forming component, until the thickness deviation rate of the film produced by the film forming component is within ±10%.

[0013] During the time controlled and adjusted by the second control module, the second thickness gauge monitors and records the timing data in real time, and removes the unqualified film during this period after the winding is completed.

[0014] During the roll gap adjustment process, the first roll gap detection sensor synchronously monitors the actual roll gap and feeds the signal back to the control center. All collected data are time-series data.

[0015] The beneficial effects of this invention are as follows: This invention can precisely control the roll gap. When facing nonlinear, time-varying, or multivariable systems, it can also accurately handle complex dynamic performance and multivariable coupling. It has better adaptability and robustness, improves the control performance of the system, improves the stability and accuracy of the system, and also improves the production efficiency and product quality of the calender.

[0016] This invention relates to fitting or analyzing time-series data. Time-series data refers to data series recorded in chronological order. It focuses on the trend of a period of time. Compared with data at a specific point in time, calculating time-series data can more accurately reflect the periodic or quasi-periodic changes of the data. Based on the periodic or quasi-periodic nature, more accurate and real-time monitoring and prediction can be made. The calculated control quantities are more precise and better meet actual requirements. The analysis time is shorter and the feedback is faster. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the calendering system of the present invention.

[0018] Figure 2This is a schematic diagram of the micro-tension control unit of the present invention.

[0019] Figure 3 This is a schematic diagram of the structure of the swing roller seat.

[0020] Figure 4 This is the control flowchart for the roll gap control subsystem.

[0021] Figure 5 This is the control flowchart for the belt speed control subsystem.

[0022] Figure 6 This is the control flowchart for the micro-tension control subsystem.

[0023] Figure 7 This is the control flowchart for the correction subsystem.

[0024] Figure 8 The optimization process of the Zebra Optimized Recurrent Neural Network (RNN-ZOA) model.

[0025] Figure 9 The flowchart shows the neural network control process for each subsystem control center.

[0026] Figure 10 This is a flowchart showing the process from neural network to PID controller and then to actuator.

[0027] In the picture: 10. Micro-tension control unit; 1. Swing roller; 2. Swing arm; 3. Swing roller seat; 310. Top screw hole; 4. Mandrel; 5. Bearing with seat; 6. Swing roller motor; 7. Angle sensor; 8. Tension sensor; 100. Film-forming component; 11. Film-forming support roller; 12. Film-forming pressure roller; 13. First roller gap control actuator; 14. First roller gap detection sensor; 15. First bearing backlash elimination mechanism; 16. First guide roller; 200. Thinning component; 21. Thinning support roller; 22. Thinning pressure roller; 23. First tension isolation device; 24. Floating roller; 25. Second bearing backlash elimination mechanism; 26. Second guide roller; 27. First thickness gauge; 28. Second roll gap control actuator; 29. ​​Second roll gap detection sensor; 300. Rewinding component; 31. Rewinding shaft assembly; 32. Second tension disconnect device; 33. Second thickness gauge; 34. Ultrasonic sensor; 35. Web correction mechanism; 36. Third guide roller; 400. Membrane. Detailed Implementation

[0028] The specific embodiments of the present invention will now be described with reference to the accompanying drawings.

[0029] like Figure 1As shown, the calendering system of the present invention includes a film-forming component 100, a thinning component 200 and a winding component 300 arranged sequentially along the production direction of the film 400.

[0030] The film-forming component 100 includes a pair of film-forming support rollers 11 and film-forming pressure rollers 12. The film-forming pressure roller 12 is located inside the film-forming support roller 11 and is eccentrically positioned with respect to the film-forming support roller 11. The film-forming pressure roller 12 is connected to and driven by a servo motor (not shown in the figure). The film-forming pressure roller 12 can be adjusted vertically relative to the film-forming support roller 11. A first roll gap control actuator 13 is provided on the outer side of one of the film-forming support rollers 11. The film-forming support roller 11, film-forming pressure roller 12, and first roll gap control actuator 13 are arranged horizontally. The first roll gap control actuator 13 adjusts the roll gap width between the film-forming pressure rollers 12 by controlling the displacement of the film-forming support roller 11 through a control valve group. A first roll gap detection sensor 14 is provided on the outer side of the pressing area between the two film-forming pressure rollers 12 for real-time detection of the roll gap between the two film-forming pressure rollers 12. A first bearing clearance elimination mechanism 15 is provided on the outer side of the two film-forming support rollers 11 to eliminate the bearing clearance of the film-forming support rollers 11 and the film-forming pressure rollers 12, so as to avoid the bearing clearance affecting the roller gap adjustment and ensure the adjustment accuracy of the first roller gap control actuator 13. The film-forming component 100 is also provided with several first guide rollers 16 for guiding the film sheet 400.

[0031] The thinning component 200 includes a pair of thinning support rollers 21 and thinning pressure rollers 22. The thinning pressure roller 22 is located inside the thinning support roller 21 and can be adjusted left and right relative to the thinning support roller 21. The thinning pressure roller 22 is connected to and driven by a servo motor (not shown in the figure). A second roll gap control actuator 28 is provided on the outer side of one of the thinning support rollers 21. The thinning support roller 21, the thinning pressure roller 22, and the second roll gap control actuator 28 are arranged vertically. The second roll gap control actuator 28 adjusts the roll gap width between the thinning pressure rollers 22 by controlling the displacement of the thinning support roller 21 through a control valve group. A second roll gap detection sensor 29 is provided on the outer side of the pressing point of the two thinning pressure rollers 22 for real-time detection of the roll gap between the two thinning pressure rollers 22. A first tension isolation device 23 is provided on the feeding side of the thinning pressure roller 22 to isolate the tension of the film 400 and prevent the tension of the film forming component 100 and the thinning component 200 from affecting each other. A floating roller 24 is provided on the discharge side of the thinning pressure roller 22. The floating roller 24 tensions the film 400 by gravity. The floating roller 24 can be counterweighted according to the tension requirements of the film 400. A second bearing clearance elimination mechanism 25 is provided on the outer side of the two thinning support rollers 21 to eliminate the bearing clearance between the thinning support rollers 21 and the thinning pressure roller 22, ensuring the accuracy of the roller gap adjustment. A first thickness gauge 27 is provided on the feeding side of the first tension isolation device 23 to detect the thickness of the film 400 formed. The thinning component 200 is also provided with several second guide rollers 26 for guiding the film 400.

[0032] The winding component 300 includes a winding shaft assembly 31, whose winding shaft is connected to and driven by a servo motor (not shown in the figure). Along the feeding direction of the film 400, a second tension isolation device 32, a second thickness gauge 33, and an ultrasonic sensor 34 are sequentially arranged on the receiving side of the winding shaft assembly 31. The second tension isolation device 32 is used to isolate the tension of the film 400, preventing the tension of the thinning component 200 and the winding component 300 from interfering with each other. The second thickness gauge 33 is used to detect the thickness of the film 400. The ultrasonic sensor 34 is used to detect the position of the film 400. A web-correcting actuator 35 is provided on the frame of the winding shaft assembly 31 for moving the frame to perform web-correcting, ensuring accurate centering of the film 400 during winding. The winding component 300 also includes several third guide rollers 36 for guiding the film 400.

[0033] Each of the film-forming component 100, the thinning component 200, and the winding component 300 is equipped with a micro-tension control unit 10. The micro-tension control unit 10 of the film-forming component 100 is located on the discharge side of the film-forming roller 12; the micro-tension control unit 10 of the thinning component 200 is located between the thinning roller 22 and the first tension isolation device 23, and between the thinning roller 22 and the floating roller 24; the micro-tension control unit 10 of the winding component 300 is located between the winding shaft assembly 31 and the second thickness gauge 33.

[0034] like Figure 2 As shown, the micro-tension control unit 10 includes a tension roller assembly and at least one tension sensor 8. The tension sensor 8 is located on the discharge side of the tension roller assembly and is used to detect the tension of the diaphragm 400. Figure 1 As shown, in this embodiment, the film-forming component 100 and the winding component 300 are each equipped with a tension sensor 8; the thinning component 200 is equipped with two tension sensors 8, which are respectively arranged on the feed side and the discharge side of the thinning pressure roller 22. Figure 2 As shown, the tension swing roller assembly includes two symmetrical swing rollers 1. Both ends of the swing rollers 1 are connected to swing arms 2 via swing roller seats 3. A spindle 4 passes through the swing arms 2. One side of the spindle 4 is connected to a swing roller motor 6 via a bearing 5. An angle sensor 7 is mounted on the spindle 4 on the same side as the swing roller motor 6. The angle sensor 7 is used to detect the rotation angle of the swing rollers 1. The swing roller motor 6 is a servo motor. Figure 1 As shown, the swing roller seat 3 has set screw holes 310 on opposite sides. The central axis of the set screw holes 310 on both sides passes through the center of the swing roller seat 3 and is located on the same plane. The swing roller seats 3 at both ends of the swing roller 1 are offset by 90° around the axial direction. The central axes of the set screw holes 310 at both ends of the swing roller seat 3 are perpendicular to each other. One end of the swing roller seat 3 can be adjusted from two directions through its two set screw holes 310, and the other end of the swing roller seat 3 can be adjusted from two other directions through its two set screw holes 310. Thus, the two ends of the swing roller 1 can be adjusted in a total of four directions. Without disassembling the machine, each swing roller 1 can be finely adjusted in four directions to ensure the symmetry of the swing roller 1 and its parallelism with each other, and to ensure the accuracy of tension adjustment.

[0035] This invention utilizes a calendering control system for real-time monitoring and parameter adjustment, ensuring the stability and accuracy of the calendering process. The calendering control system includes a roll gap control subsystem, a belt speed control subsystem, a micro-tension control subsystem, and a deviation correction subsystem.

[0036] The roll gap control subsystem is used to control the adjustment of the roll gap of the film-forming roller 12 of the film-forming component 100 and the thinning roller 22 of the thinning component 200. The roll gap control actuators of the film-forming component 100 and the thinning component 200 are electrically connected to the roll gap detection sensor, the first thickness gauge 27 of the thinning component 200, and the second thickness gauge 33 of the winding component 300. The roll gap control actuators adjust the roll gap of the film-forming roller 12 and the thinning roller 22 according to the roll gap information detected by the roll gap detection sensor and the film thickness information detected by the first thickness gauge 27 and the second thickness gauge 33.

[0037] The detailed steps are as follows: During equipment operation, the first thickness gauge 27 detects the thickness of the diaphragm 400 in real time and sends it to the control center. The control center calculates the deviation rate between the actual thickness and the calibrated thickness. The formula for calculating the deviation rate is as follows:

[0038] Where a is the calibrated thickness and x is the actual thickness. If the deviation rate is within the normal fluctuation range (±5%), the data is stored in the database and no control operation is performed.

[0039] If the deviation rate is within the rated range (±5% to ±10%), the control center will activate the first control module. First, using the current time-series thickness data of the first thickness gauge 27 as a reference, it will retrieve the 10 time-series thickness data points preceding this timestamp from the storage database and combine them with the current time-series thickness data to form a time-series dataset {T1, T2, ..., T...}. 11}, then the time series dataset {T1,T2,…,T} is... 11 The Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm is used to compute the output parameter set {K}. P ,T I ,T D The parameter set ,tsd,Mpd,ess,e(t)} is fed into the PID controller. After receiving the parameter set, the PID controller performs calculations on e(t) to obtain the position adjustment command to the servo hydraulic proportional valve of the second roll gap control actuator 28 of the thinning component 200. This process takes milliseconds. The servo hydraulic proportional valve controls the second roll gap control actuator 28 to extend or retract to adjust the thinning support roller 21 and thus adjust the roll gap of the thinning component 200.

[0040] After the deviation diaphragm is corrected by the thinning component 200, it is thinned. The second thickness gauge 33 transmits the thickness of the finished diaphragm to the control center, which calculates the deviation rate between the actual thickness of the finished diaphragm and the rated thickness.

[0041] When the deviation rate is within ±6%, the roll gap is not corrected, and the time-series thickness data is stored in the storage library. When the deviation rate exceeds ±6%, the time-series thickness data of the first thickness gauge 27 will be blocked. Using the current time-series thickness data of the second thickness gauge 33 as a reference, the 10 time-series thickness data points preceding this timestamp will be retrieved from the storage database and combined with the current time-series thickness data to form a time-series dataset {T1', T2', ..., T...}. 11 '}, and then the time series dataset {T1',T2',…,T 11 The Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm calculates the output parameter set {K}. P ,T I ,T D The parameter set ,tsd,Mpd,ess,e(t)} is fed into the PID controller. After receiving the parameter set, the PID controller calculates e(t) to obtain a position adjustment command, which is sent to the servo hydraulic proportional valve of the second roll gap control actuator 28 of the thinning component 200. This process takes milliseconds. The servo hydraulic proportional valve controls the extension or retraction of the second roll gap control actuator 28 to adjust the roll gap of the thinning component 200 until the deviation rate is reduced to within 6%. After the adjustment is completed, the roll gap is detected by the second roll gap detection sensor 29 and fed back to the PID controller. After that, the roll gap is released. The shielding of the first thickness gauge 27 is adjusted based on the timing thickness signal of the first thickness gauge 27. At the same time, the zebra optimized recurrent neural network algorithm (RNN-ZOA) controlling the roll gap of the thinning component 200 is optimized a second time to ensure that the thickness deviation rate of the finished film is within the set range. After the roll gap adjustment of the thinning component 200 is completed, the second thickness gauge 33 continuously monitors the thickness of the finished film 400 and sends it to the control center. If the thickness returns to the fluctuation range, the control is completed and the second roll gap control actuator 28 no longer operates.

[0042] If the deviation rate exceeds the rated range (±10%), it means that the second roll gap control actuator 28 needs a large adjustment or is no longer adjustable. In this case, the control center will use the second control module. First, based on the current sequential thickness data of the first thickness gauge 27, it will retrieve the 10 time-series thickness data before this timestamp from the storage library and combine them with the current time-series thickness data to form a time-series dataset {T1, T2, ..., T...}. 11}, then the time series dataset {T1,T2,…,T} is... 11 The Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm is used to compute the output parameter set {K}. P ,T I ,T DThe parameters are fed into the PID controller. After receiving the parameter set, the PID controller performs calculations on e(t) to obtain a position adjustment command. This command is sent to the servo hydraulic proportional valve of the first roll gap control actuator 13 of the film forming component 100. This process takes milliseconds. The position command is sent through the servo hydraulic proportional valve to control the first roll gap control actuator 13 to extend or retract to adjust the film forming support roller 11 and thus adjust the roll gap of the film forming component 100 until the thickness deviation rate of the film 400 produced by the film forming component 100 is within ±10%. Then, the controller switches to the first control module. During this period, the second thickness gauge 33 monitors and records the timing data in real time. After the winding is completed, the unqualified film 400 produced during this period is removed.

[0043] During the roll gap adjustment process, the first roll gap detection sensor 14 synchronously monitors the actual roll gap and feeds the signal back to the control center to ensure the completion of the roll gap adjustment command. All collected data are time-series data.

[0044] The typical PID control process is as follows: The first thickness gauge 27 monitors the thickness of the diaphragm 400 and sends this signal to the PID controller; the PID controller averages the data over a period of time, and if this average exceeds the set value, the output signal is adjusted to the gap between the membrane pressure rollers 12. Because the traditional PID control method involves adjusting after measurement, it inevitably results in a section of the diaphragm 400 with unacceptable thickness, leading to poor overall thickness uniformity. However, the control method based on the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm can effectively reduce the length of the unacceptable thickness diaphragm 400, resulting in higher overall thickness uniformity of the diaphragm 400.

[0045] like Figure 4 As shown, the control method of the roll gap control subsystem is as follows: 1) The first roll gap detection sensor 14 and the second roll gap detection sensor 29 transmit the real-time detected roll gap signal, and the first thickness gauge 27 and the second thickness gauge 33 transmit the real-time monitored film 400 thickness signal to the roll gap control center; 2) The roll gap control center performs fitting calculations on the roll gap signal and thickness signal based on the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA); 3) The roll gap control center sends the time-series data obtained after fitting calculations to the first roll gap control actuator 13 and the second roll gap control actuator 28; 4) The first roll gap control actuator 13 controls the hydraulic system through the control valve group to adjust the displacement of the film forming support roller 11 and the thinning support roller 21 according to the instructions, thereby adjusting the displacement of the film forming pressure roller 12 and the thinning pressure roller 22, and then adjusting the roll gap distance between the two film forming pressure rollers 12 and the two thinning pressure rollers 22 to accurately adjust the thickness of the film 400.

[0046] The belt speed control subsystem is used to control the conveying speed of the film 400. The servo motors connected to the film forming roller 12, the thinning roller 22, the take-up shaft assembly 31, the first tension interruption device 23, and the second tension interruption device 32 are electrically connected to the belt speed control subsystem. The belt speed control subsystem uses the speed of the servo motor of one of the rollers (here, the thinning roller 22) as a reference to adjust the speeds of the servo motors of the other rollers and the take-up shaft, as well as the oscillating roller motor 6.

[0047] The detailed steps are as follows: The servo motors connected to the film-forming roller 12, thinning roller 22, winding shaft assembly 31, first tension isolation device 23, and second tension isolation device 32 transmit timing data-speed signals to the control center. The control center uses the speed of the servo motor of the thinning roller 22 as a reference to calculate the speed of the other servo motors relative to the reference speed, and calculates the speed deviation rate of each servo motor. When the deviation rate is within the normal range, the control center does not issue commands to other servo motors based on the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA). When the deviation rate exceeds the normal range, commands are issued to the servo motors with deviation rates exceeding the normal range to adjust their speeds.

[0048] When the material or process changes and the belt speed needs to be changed, the control center will simultaneously send commands (including acceleration, final speed, and other parameters) to the servo motors connected to the film-forming roller 12, thinning roller 22, take-up shaft assembly 31, first tension isolation device 23, and second tension isolation device 32 based on the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) to increase or decrease their speeds to meet the speed change requirements. After adjusting their speeds, the speed of the servo motor of the thinning roller 22 is used as the reference to monitor the speeds of the other servo motors to ensure that the deviation rate between the speeds of each servo motor and the speed of the servo motor of the thinning roller 22 is within the normal range.

[0049] The Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) embedded PID is more accurate in terms of output parameters compared to traditional PID control.

[0050] like Figure 5As shown, the control method of the belt speed control subsystem is as follows: 1) Each servo motor feeds back its speed signal to the servo driver; 2) The servo driver transmits the received signal to the belt speed control center; 3) The belt speed control center uses the speed of the servo motor of one of the pressure rollers (reference rollers) in the film-forming component 100 or the thinning component 200 as a reference to analyze and calculate the speed of other servo motors based on the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm; 4) The belt speed control center sends the instructions obtained after analysis and calculation to the servo driver; 5) The servo driver adjusts the speed of the servo motors according to the instructions to stabilize the servo motors at the set speed, avoiding problems such as belt breakage or insufficient compaction caused by excessive speed difference between the servo motors.

[0051] The micro-tension control subsystem is used to control the tension of the diaphragm 400 to prevent the diaphragm 400 from breaking or wrinkling. The oscillating roller motor 6, angle sensor 7, and tension sensor 8 of the micro-tension control unit 10 are electrically connected to the micro-tension control subsystem. The oscillating roller motor 6 adjusts the tension of the diaphragm 400 based on the rotation angle information detected by the angle sensor 7 and the tension information detected by the tension sensor 8.

[0052] The detailed steps are as follows: Here, we take the micro-tension control subsystem at the film-forming component 100 as an example; the logic and steps of the other micro-tension control subsystems are the same. Before the equipment is run, the micro-tension control unit 10 is placed horizontally, so that the swing arm 2 is horizontal and the swing roller 1 is at the same height. Then, the swing roller motor 6 is initialized and adjusted to torque mode to output the rated torque. During equipment operation, tension sensor 8 monitors the tension of diaphragm 400 in real time and transmits this time-series data to the micro-tension control subsystem. The control center collects and inputs the time-series tension data. If the tension time-series data exceeds the fluctuation range at a certain time, the control center retrieves the time-series tension data from a previous period and uses the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm to calculate and fit this time-series dataset to predict its tension trend. If the trend is within the fluctuation range, the control center does not correct it. If the trend is outside the fluctuation range or is predicted to exceed the fluctuation range in the future, the control center calculates a correction value based on this AI algorithm and sends a command to the swing roller motor 6. The output torque of the swing roller motor 6 is changed to adjust the tension in this segment. When the output torque of the swing roller motor 6 is adjusted, tension sensor 8 continues to monitor the diaphragm tension. If the tension is within the normal fluctuation range at this time, the control is completed.

[0053] During the tension fluctuation correction process, the rotation angle of the swing roller motor 6 will change, causing the swing arm 2 to break the overall horizontal dynamic balance. This overall horizontal dynamic balance has a significant impact on the control of micro-tension and the smooth operation of the diaphragm 400 in the tension range. Therefore, the present invention also modifies the position of the swing arm 2 to ensure the overall horizontal dynamic balance of the swing arm 2. During the tension fluctuation correction process, the angle signal of the swing roller motor 6 is synchronously sent to the control center. Based on the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm, the control center pre-constructs a function by combining the angle information of the swing roller motor 6 with the speed signal of the servo motor of the first tension isolation device 23. After calculation by the correlation function, the function is output to the servo motor of the first tension isolation device 23, so that it adjusts its speed to adjust the position of the swing arm 2 and the angle of the swing roller motor 6. As the quasi-static speed of the servo motor of the first tension isolation device 23 increases, the roller speed of the first tension isolation device 23 will be faster than the roller speed of the film forming roller 12, and a speed difference will gradually be generated between the two. The film 400 will be gradually tightened, which will cause the swing arm 2 to gradually tend to be horizontal, thereby adjusting the angle deflection of the swing roller motor 6. When this adjustment process is completed, the angle information of the swing roller motor 6 continues to be sent to the control center. The control center judges this data. If it is within the normal fluctuation range, the control is completed.

[0054] Traditional PID controllers cannot construct an effective function when adjusting the overall horizontal dynamic balance of swing arm 2. They are worse than the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm in terms of variability and adaptability to data changes. The Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm can output parameters more accurately and provide predictions based on these parameters.

[0055] like Figure 6 As shown, the control method of the micro-tension control subsystem is as follows: 1) Angle sensor 7 transmits the real-time detected rotation angle signal of the swing roller 1 and tension sensor 8 transmits the real-time detected tension signal of the diaphragm 400 to the micro-tension control center; 2) The micro-tension control center analyzes and calculates the rotation angle signal and tension signal based on the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm; 3) The micro-tension control center sends the analyzed and calculated instructions to the swing roller motor 6; 4) The swing roller motor 6 adjusts the rotation angle and speed of the swing roller 1 according to the specified parameters, thereby adjusting the tension of the diaphragm 400, resulting in more accurate tension adjustment. In addition to real-time monitoring of the tension of the diaphragm 400 through tension sensor 8, the micro-tension control subsystem also uses angle sensor 7 to detect the angle of the swing roller 1 in real time. Angle sensor 7 can provide real-time feedback on the actual rotation angle of the swing roller 1 during tension adjustment, providing more precise parameters for control and making tension adjustment more stable and accurate.

[0056] The web correction subsystem is used to adjust the position of the take-up shaft assembly 31 to ensure that the diaphragm 400 is accurately centered during winding. The ultrasonic sensor 34 of the take-up component 300 is electrically connected to the web correction actuator 35. The web correction actuator 35 adjusts the position of the take-up shaft assembly 31 to correct the web alignment based on the position information detected by the ultrasonic sensor 34.

[0057] The detailed steps are as follows: During equipment operation, the ultrasonic sensor 34 continuously monitors the edges of the diaphragm 400. Due to the influence of factors such as tension and friction during the operation of the diaphragm 400, there will be a relative positional deviation between the diaphragm 400 and the winding shaft assembly 31 at different times, resulting in uneven edges. Therefore, the ultrasonic sensor 34 monitors the edges of the diaphragm and converts them into positional offsets, and transmits this offset signal to the control center. The control center judges the offset in real time. If it does not exceed the fluctuation range, the control center does not issue a command. If a certain time offset exceeds the fluctuation range, the control center immediately uses this time offset as a reference, retrieves some time data from before this time period, and uses the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) to fit and calculate to predict its future trend. Based on this, the rotation angle of the motor of the correction actuator 35 is calculated, and this angle signal is transmitted to the servo motor of the correction actuator 35. The servo motor rotates the corresponding angle, causing the motion rod of the correction actuator 35 to extend and retract, thereby pushing the winding shaft assembly 31 to move, ensuring the neatness of the edges of the diaphragm 400 after winding. While this step is being completed, the ultrasonic sensor 34 continues to monitor and send signals. If the deviation is within the normal fluctuation range at the end, the angle of the servo motor of the correction actuator 35 will no longer change, and the correction control is completed.

[0058] like Figure 7 As shown, the control method of the correction subsystem is as follows: 1) The ultrasonic sensor 34 transmits the position signal of the diaphragm 400 detected in real time to the correction control center; 2) The correction control center analyzes and calculates the position signal based on the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA); 3) The correction control center sends the command obtained from the analysis and calculation to the correction execution mechanism 35; 4) The correction execution mechanism 35 controls the overall movement of the frame of the take-up shaft assembly 31 according to the command, adjusts the position of the take-up shaft assembly 31 to achieve correction, and ensures the take-up is aligned.

[0059] The control centers of the roll gap control subsystem, belt speed control subsystem, micro-tension control subsystem, and web correction subsystem all employ the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm to fit or analyze time-series data. Based on the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm, reasonable control quantities are predicted. These control quantities correspond to the displacements of the film-forming support roller 11 and the thinning support roller 21 in the roll gap control subsystem; the rotational speed of the servo motor in the belt speed control subsystem; the rotation angle and speed of the oscillating roller motor 6 in the micro-tension control subsystem; and the movement of the take-up shaft assembly 31 frame in the web correction subsystem. The neural network structure includes an input layer, a hidden layer, and an output layer, and the neural network also has a suitable activation function. The Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) is used for fitting and analyzing time-series data. Time-series data refers to data series recorded in chronological order. It focuses on the trend of a period of time. Compared with data at a specific point in time, calculating time-series data can more accurately reflect the periodic or quasi-periodic changes of the data. Based on the periodic or quasi-periodic nature, it can make more accurate and real-time monitoring and predictions, and the calculated control quantities are more precise, better meeting actual requirements. The analysis time is shorter and the feedback is faster.

[0060] The calendering control system of this invention (employing the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA)) is embedded as a control module within a PID controller. The control quantity obtained through the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) serves as the input vector for the PID controller. The output quantity calculated by the PID controller is then output to the corresponding actuator to adjust the relevant components. By embedding the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) into the PID controller as a control module, this invention centralizes the equipment and system, making it more intelligent and autonomous. Compared to traditional centralized computing models, the embedded algorithm model embeds AI algorithm capabilities into the equipment itself, enabling it to process data and make decisions locally without relying on cloud servers. This allows for faster and more real-time data analysis and response, reduces network dependence, improves system efficiency and performance, and enables rapid understanding and response to user needs, automatically adapting to environmental changes and providing more intelligent and personalized services.

[0061] In this invention, a discrete-time system is used in the PID controller. Discrete-time systems can suppress noise and interference through techniques such as digital filtering, and are more robust than continuous-time systems, making them suitable for real-time control applications. Complex control strategies can be implemented easily using algorithms. In this invention, the PID controller uses an Euler transform discretization control algorithm: where P is the proportional band; the smaller P is, the greater the amplification value K of the error signal (e(t)). PThe larger the value, the greater the value; I represents the integral part of the control algorithm, which is to eliminate the "residual error". Without this part, even in a steady state, the setpoint and the process value may be different; finally, D represents the differential part of the algorithm. This part has two objectives: it makes the system more flexible to sudden changes or disturbances in the setpoint, but most importantly, it helps to stabilize the system.

[0062] The transfer function of a PID controller is:

[0063] In the formula, s is the Laplace variable (s=α+jω), K p It is the magnification factor, T I It is the integration time, T D It is the time for differentiation, {K p T I T D} represents the parameter set of the PID controller, which depends on the physical properties of the process.

[0064] The PID tuning function is:

[0065] The integral of e(t) can be approximated as a rectangular sum, i.e.

[0066] Where e'(t) can be approximated as a straight line, i.e.

[0067] Therefore, when using the Euler transform discrete control algorithm, the PID controller yields the following expression:

[0068] like Figure 10 As shown, this invention introduces the Zebra algorithm to optimize the neural network model based on the Euler variation discrete control algorithm, thereby improving the PID input parameter set {K} through the neural network model. p T I T D To improve accuracy, three fundamental characteristics of the dynamic response of neural network training are introduced: maximum overshoot (Mp), settling time (ts), and steady-state error (ess).

[0069] Before being embedded in a control system, the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm requires prior training. The training learns the mapping function X→Y, where X is the input vector and Y is the output vector. The recurrent neural network model is trained using historical data, including parameters such as roll gap width, diaphragm thickness, motor speed, diaphragm tension, roller angle, position, and other key timing parameters required by each sub-control system. Information collected by sensors is preprocessed before being input into the recurrent neural network model. Preprocessing includes noise removal and data normalization, which improves the training effect of the neural network model. The recurrent neural network model can also be periodically updated and adjusted, retraining when operating conditions change to adapt to changes in the working environment and system evolution. The recurrent neural network model can also store multi-process data and perform self-learning to achieve flexible self-adjustment capabilities. During training, the Zebra Optimization algorithm is used to continuously adjust the weights and biases of the neural network to minimize the error between the predicted output and the actual output, thereby improving the accuracy and stability of the predicted output.

[0070] Figure 8 The Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm model optimization process is illustrated. In optimizing an RNN model, the parameter matrix needs to be optimized for five variables: number of network layers, number of neurons, learning rate, epochs, and batch size. Upper and lower limits need to be set for each variable.

[0071] The objective function to be minimized is:

[0072] t represents the training time of the neural network, and loss represents the loss value of the neural network. m and n represent the weights, which are user-defined parameters that allow users to selectively focus on optimizing the target.

[0073] After optimization begins, the first step is to divide the dataset into a training set and a test set.

[0074] Next, we will build the RNN model. The variables that need to be optimized will not be set yet and will be determined later.

[0075] Randomly initialize n sets of hyperparameters (n is a positive integer). These n sets of hyperparameters are called a population. Substitute the population into the RNN model and train the model. After training, the objective function value can be obtained.

[0076] Next, the hyperparameters are optimized and updated according to the Zebra optimization algorithm. The optimization objective is to minimize the objective function, and the parameter matrix to be optimized is abstracted as the location of the zebra.

[0077] In the Zebra Optimization Algorithm, the first stage is the foraging stage. The individual in the population that forages the most (i.e., the set of hyperparameters that minimizes the objective function) is the lead zebra, and other individuals will move towards the position of the lead zebra. The mathematical model is as follows:

[0078] Where r is a random number between [0,1], and I is a random value belonging to the set {1,2}.

[0079] The second stage is the predator encounter stage. In this stage, the individual will choose to disperse or gather based on the predator's strength. Dispersing or gathering updates the individual's position. If the individual's new position reduces the objective function, then that individual's position will be updated. The mathematical model is as follows:

[0080] Where t is the number of iterations, T is the maximum number of iterations, R is a constant of 0.01, Ps is the switching probability between the two strategies, and its value is a random number between [0,1]. AZ is the state of the attacked zebra.

[0081] The optimization process is a loop iteration, and it is necessary to set the iteration termination condition for the algorithm. The iteration termination condition can be one or more combinations of the following conditions: reaching the upper limit of the loop iteration; the objective function value obtained by the optimal parameters is less than the ideal value; the optimal parameters have not been updated after m iterations, etc.

[0082] After iteration, find the set of parameters that minimizes the objective function; this set of parameters is the optimal solution. The RNN model corresponding to the optimal solution is the optimal RNN model. Substituting the test set into the optimal RNN model yields the model's accuracy on the test set.

[0083] like Figure 9 As shown, the model selected by the Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) is embedded into the control system. The specific control method of the system's control center is as follows: 1) The Zebra Optimized Recurrent Neural Network Algorithm (RNN-ZOA) starts running; 2) Initialization of the actuators of each control subsystem: The first roll gap control actuator 13 and the second roll gap control actuator 28 of the roll gap control subsystem are initialized to their initial positions; the servo driver of the belt speed control subsystem is initialized to a constant speed; the swing roller motor 6 of the micro-tension control subsystem is initialized to a constant torque; the correction actuator 35 of the correction subsystem is initialized to its initial position; the initial positions of each actuator are preset during installation. 3) The Zebra Optimized Recurrent Neural Network (RNN-ZOA) model predicts the initial control quantity of the actuator based on the difference between the initial actual parameter value and the set parameter value of the actuator of each control subsystem; 4) The actual parameter values ​​detected in real time by the detection components of each control subsystem are used as lightweight input vectors to input the Zebra Optimized Recurrent Neural Network (RNN-ZOA) model; For the roll gap control subsystem, the input vector X1 = [W1_roll gap, W2_roll gap, T1_film, T2_film, timestamp] (W1 is the roll gap width between the two film forming rollers 12, W2 is the roll gap width between the two thinning rollers 22, T1 is the thickness of the film 400 before calendering (data detected by the first thickness gauge 27), T2 is the thickness of the film 400 after calendering (data detected by the second thickness gauge 33), and timestamp is the timing sequence). For the belt speed control subsystem, the input vector X2 = [n1_motor] (n1 is the speed of each servo motor); For the micro-tension control subsystem, the input vector X3=[θ_actual, θ_set, v, a] (where θ_actual is the actual angle of the swing roller 1, θ_set is the set angle of the swing roller 1, v is the speed of the swing roller motor 6, and a is the acceleration of the swing roller motor 6). For the correction subsystem, the input vector X4 = [ΔL] (ΔL is the displacement of the diaphragm by 40°). 5) The Zebra Optimized Recurrent Neural Network (RNN-ZOA) model optimizes the data and calculates the predicted real-time control quantity, which is then used as a lightweight output vector. For the roll gap control subsystem, the output vector Y1=[P] (P is the pressure execution command of the hydraulic cylinder of the first roll gap control actuator 13 and the second roll gap control actuator 28); For the belt speed control subsystem, the output vector Y2=[n_motor] (n is the speed command of the motor of the reference roller); For the micro-tension control subsystem, the output vector Y3 = [y_1, y_2, ...] (where y_1 is the speed execution command of the swing roller motor 6, and y_2 is the acceleration execution command of the swing roller motor 6). For the correction subsystem, the output vector Y4 = [α] (α is the position execution command of the actuator of the correction actuator 35). 6) The lightweight output vector of the Zebra Optimized Recurrent Neural Network (RNN-ZOA) model is input to the PID controller, and the PID controller adjusts the output in real time according to the output vector (time-sequence predictive control quantity) of the neural network model. The adjustment function of a PID controller is: u'(t) = u(t) + f(Y) (where u(t) is usually a function based on the error e(t), including proportional, integral and derivative terms, and the error e(t) refers to the difference between the actual parameter value and the set parameter value; f(Y) is an empirical adjustment function).

[0084] 7) Determine whether the signal is stable based on the values ​​detected by the detection components of each control subsystem: if stable, proceed to the next step and complete the control; if unstable, return to the start of operation and restart until the signal is stable.

[0085] The calendering system of this invention incorporates a Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm combined with a PID controller in each control subsystem. The PID controller automatically adjusts the output in real time based on the real-time predictive control input of the Zebra Optimized Recurrent Neural Network (RNN-ZOA) algorithm, enabling precise control of roll gap, belt speed, micro-tension, and deviation correction. When dealing with nonlinear, time-varying, or multivariable systems, it can accurately handle complex dynamic performance and multivariable coupling, exhibiting better adaptability and robustness. This improves the system's control performance, stability, and accuracy, as well as the calender's production efficiency and product quality.

[0086] The above description is an explanation of the present invention and not a limitation thereof. The present invention can be modified in any form without departing from its spirit.

Claims

1. A method for controlling the roll gap in a calendering system, characterized in that: The roll gap control subsystem adjusts the roll gap distance between the two film-forming rollers (12) and / or the two thinning rollers (22) based on the roll gap signal and the film (400) thickness signal. The control center of the roll gap control subsystem calculates the deviation rate between the actual thickness detected by the first thickness gauge (27) and the calibrated thickness. If the deviation rate is within ±5%, the data is stored in the storage library and no control operation is performed. If the deviation rate is within ±5% to ±10%, the control center uses the first control module to adjust the roll gap of the thinning component (200). If the deviation rate exceeds ±10%, the control center uses the second control module to adjust the roll gap of the film-forming component (100). The control process of the first control module is as follows: based on the current time-series thickness data of the first thickness gauge (27), the system retrieves the 10 time-series thickness data before this timestamp from the storage library and combines them with the current time-series thickness data to form a time-series dataset {T1,T2,…,T... 11 }, the time series dataset {T1,T2,…,T} 11 The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P ,T I ,T D ,tsd,Mpd,ess,e(t)} to PID. After receiving the parameter set, PID performs calculations on e(t) to obtain the position adjustment command to the servo hydraulic proportional valve of the second roll gap control actuator (28) of the thinning component (200). This process takes milliseconds. The servo hydraulic proportional valve controls the second roll gap control actuator (28) to extend or retract to adjust the thinning support roller (21) and thus adjust the roll gap of the thinning component (200).

2. The roll gap control method for the calendering system according to claim 1, characterized in that: After the second control module adjusts the roller gap of the film-forming component (100) to a deviation rate within ±10%, it switches to the first control module.

3. The roll gap control method for the calendering system according to claim 1, characterized in that: After the first control module adjusts the roll gap of the thinning component (200), the deviation diaphragm is thinned after being corrected by the thinning component (200). The control center calculates the deviation rate 2 between the actual thickness detected by the second thickness gauge (33) and the rated thickness here. When the deviation rate 2 is within the range of ±6%, the roll gap is not corrected. When the deviation rate 2 exceeds the range of ±6%, the roll gap of the thinning component (200) is adjusted based on the thickness signal of the second thickness gauge (33) until the deviation rate 2 is reduced to within the range of ±6%.

4. The roll gap control method for the calendering system according to claim 3, characterized in that: When the deviation rate exceeds ±6%, the time-series thickness data of the first thickness gauge (27) is blocked. Using the current time-series thickness data of the second thickness gauge (33) as a reference, the time-series thickness data of the 10 timestamps before this time stamp are retrieved from the storage and combined with the current time-series thickness data to form a time-series dataset {T1',T2',…,T... 11 '}, and then the time series dataset {T1',T2',…,T 11 The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P ,T I ,T D ,tsd,Mpd,ess,e(t)} to PID. After receiving the parameter set, PID calculates e(t) to obtain the position adjustment command to the servo hydraulic proportional valve of the second roll gap control actuator (28) of the thinning component (200). This process takes milliseconds. By controlling the extension or retraction of the second roll gap control actuator (28) through the servo hydraulic proportional valve, the roll gap of the thinning component (200) is adjusted to within the range of ±6% of the deviation rate. After the adjustment is completed, the roll gap is detected by the second roll gap detection sensor (29) and fed back to PID. Then the shielding of the first thickness gauge (27) is removed, and the adjustment is performed based on the time sequence thickness signal of the first thickness gauge (27).

5. The roll gap control method for the calendering system according to claim 3, characterized in that: When the deviation rate is within ±6%, the time series thickness data is stored in the storage library.

6. The roll gap control method for the calendering system according to claim 1, characterized in that: While the first control module adjusts the roll gap of the thinning component (200), the zebra-optimized recurrent neural network algorithm for controlling the roll gap of the thinning component (200) is optimized a second time.

7. The roll gap control method for the calendering system according to claim 1, characterized in that: The control process of the second control module is as follows: taking the current time-series thickness data of the first thickness gauge (27) as a reference, it retrieves the 10 time-series thickness data before this timestamp from the storage library and combines them with the current time-series thickness data to form a time-series dataset {T1,T2,…,T...}. 11 }, the time series dataset {T1,T2,…,T} 11 The input zebra-optimized recurrent neural network algorithm calculates the output parameter set {K}. P ,T I ,T D ,tsd,Mpd,ess,e(t)} to PID. After receiving the parameter set, PID performs calculations on e(t) to obtain a position adjustment command to the servo hydraulic proportional valve of the first roll gap control actuator (13) of the film forming component (100). This process takes milliseconds. The position command is issued through the servo hydraulic proportional valve to control the first roll gap control actuator (13) to extend or retract to adjust the film forming support roller (11) and thus adjust the roll gap of the film forming component (100) until the thickness deviation rate of the film (400) produced by the film forming component (100) is within ±10%.

8. The roll gap control method for the calendering system according to claim 1, characterized in that: During the time controlled and adjusted by the second control module, the second thickness gauge (33) monitors and records the timing data in real time, and removes the unqualified film (400) during this period after the winding is completed.

9. The roll gap control method for the calendering system according to claim 1, characterized in that: During the roll gap adjustment process, the first roll gap detection sensor (14) synchronously monitors the actual roll gap of the film forming component (100) and feeds the signal back to the control center. All collected data are time-series data.

Images