Spinning cake roll changing automatic splicing device and method based on intelligent algorithm and precise control
Through intelligent algorithms and precise control technology, the entire process of automatic roll changing and twisting of yarn cakes in the production of chemical fiber filaments is automated, which solves the problem of low automation in existing technologies, improves production efficiency and equipment stability, and reduces scrap rate and maintenance costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-29
- Publication Date
- 2026-03-13
AI Technical Summary
The lack of an automated and precise mechanism for joining the ends of yarn cakes in the production of chemical fiber filaments results in low success rate, unstable joint tension, insufficient coordination between modules, and long processing time, making it difficult to meet the needs of continuous and efficient production.
Employing intelligent algorithms and precise control technology, the system achieves automatic yarn cake changing and splicing through the coordinated operation of components such as intelligent suction nozzles, adaptive gripper assemblies, micro-force feedback limit grippers, intelligent shearing mechanisms, intelligent splicers, and a central intelligent control module. Combined with algorithms such as A* path planning, fuzzy PID, and Kalman filtering, it realizes fully automated operation throughout the entire process.
It achieves highly efficient automation of the yarn cake changing process, increasing production efficiency by 40%, reducing the joint scrap rate by 25%, extending the continuous operation time of the equipment by 50%, reducing maintenance costs by 20%, and ensuring the stability of winding quality and a significant reduction in production costs.
Smart Images
Figure CN121651179A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of chemical fiber winding equipment technology, specifically to an automatic twisting device and method for changing yarn cakes based on intelligent algorithms and precise control. Background Technology
[0002] In the production of synthetic fiber filaments, the knotting and splicing of the yarn ends between yarn cakes is a crucial step in ensuring continuous production. Currently, this process suffers from the following problems: yarn end finding relies on manual positioning, lacking an automated and precise mechanism; manual fixing of the yarn during splicing leads to low splicing success rates, unstable joint tension, insufficient coordination between modules, and long roll changes. Some systems even lack splicing modules, requiring machine shutdown and manual connection. These problems result in excessive manual intervention, unstable quality, and poor equipment coordination, leading to low efficiency and high scrap rates, making it difficult to meet the demands of continuous and efficient production.
[0003] Although some automation attempts have been made in the industry, most of them simply add mechanical actions and fail to fundamentally solve core problems such as precise control, intelligent decision-making, and signal interference. In response to the above problems, this invention introduces advanced technologies such as intelligent algorithms, micro-force control, and filtering algorithms, and achieves a major breakthrough in automatic yarn cake changing and splicing technology. Therefore, it is necessary to invent an automatic yarn cake changing and splicing device and method based on intelligent algorithms and precise control to solve the above problems. Summary of the Invention
[0004] (I) Technical Solution To achieve the aforementioned breakthrough in automatic yarn cake changing and splicing technology through the introduction of advanced technologies such as intelligent algorithms, micro-force control, and filtering algorithms, this invention provides the following technical solution: An automatic twisting device for changing yarn cakes based on intelligent algorithms and precise control includes an intelligent suction nozzle, an adaptive gripper assembly, a micro-force feedback limiting gripper, an intelligent shearing mechanism, an intelligent twister, a precision holding roller, a central intelligent control module, and a leather roller. The intelligent shearing mechanism includes a first shear and a second shear.
[0005] Preferably, the intelligent suction nozzle is used to adsorb the ends of the silk cake, is connected to an air compressor, and is equipped with a displacement sensor and an intelligent algorithm chip (using an improved A* path planning algorithm, with a path evaluation function f(n) = g(n) + 0.8|x n -x_goal|+0.2|y n -y_goal|, positioning accuracy ±0.5mm).
[0006] Preferably, the adaptive gripper assembly is a two-finger structure with a built-in pressure sensor and a microprocessor (using a fuzzy PID adaptive control algorithm, Kp=0.9-1.2, Ki=0.8-1.1, Kd=0.6-0.9, gripping force 0.5-5N, accuracy ±0.05N).
[0007] Preferably, the micro-force feedback limiting claw has an arc-shaped contact surface, a built-in micro-force feedback device (strain gauge force sensor, accuracy 0.1 grade, sampling frequency 1kHz), and adopts a position-force hybrid control algorithm (position accuracy ±5μm, force accuracy ±0.02N).
[0008] Preferably, the first and second shears are each equipped with an intelligent control unit and a filtering algorithm (Kalman filter Q=0.01, R=0.1; moving average window 5) to control the shearing action according to the tension threshold.
[0009] Preferably, the intelligent splicer is a knotting mechanism, which embeds an intelligent algorithm module (adjusting splicing parameters based on thread parameters) and a tension sensor (accuracy ±0.5cN).
[0010] Preferably, the precision bearing roller has a double roller pressing structure and adopts a closed-loop control system (incremental PID, Kp=1.2, Ki=0.05, Kd=0.3, pressure 1-10N, accuracy ±0.05N).
[0011] Preferably, the central intelligent control module adopts a hierarchical control architecture, controls the collaborative work of various components through an EtherCAT bus (cycle ≤ 1ms), integrates multiple intelligent algorithms, and supports redundant and fault-degraded operation.
[0012] A splicing method for an automatic splicing device for yarn cake changing based on intelligent algorithms and precise control includes the following steps: S1, First Silk Pancake Launch Phase The intelligent suction nozzle moves along the axis of the first yarn cake according to the path planned by the A* algorithm (speed 50mm / s) to pick up the yarn end, and passes through the opened adaptive gripper assembly (initial gripping force 0.5N) and the micro-force feedback limiting claw (pre-tightening force 0.2N) in sequence. During the process, the micro-force feedback limiting claw adjusts the pressure in real time (sampling period 10ms) to ensure the stability of the yarn. After receiving the sensor signal, the central intelligent control module instructs the adaptive gripper assembly and the micro-force feedback limiting claw to close synchronously (the gripping force is adjusted according to the yarn diameter). Then, the first scissors eliminate signal interference based on the filtering algorithm (wavelet filtering 3 layers db4) and accurately cut the yarn according to the set tension threshold. Finally, the intelligent suction nozzle sends the yarn end into the roller inlet and starts the winding. S2, the stage of connecting the old and new silk cakes. After the second yarn cake is in place, the intelligent suction nozzle picks up its new yarn end (positioning accuracy ±0.3mm) and sucks the yarn end of the first yarn cake into the gripper. The two yarns are then passed sequentially through a precision holding roller (pressure 3.5-5.0N, adjusted according to yarn material) to apply pressure, and an intelligent splicer (knotting according to algorithm: polyester 0.2mm: splicing speed 100r / min, 5 turns; nylon 0.4mm: speed 80r / min, 6 turns) to complete automatic splicing. During the process, a tension sensor monitors in real time (data refresh rate 1kHz). After splicing, the second scissors precisely trim the remaining yarn end (length ≤0.5mm), and all actuators reset (time ≤1s) to prepare for the next round of operation. S3, Sensor signal filtering A Kalman filter + moving average combined algorithm is used to process tension, displacement and force sensor signals, filtering mechanical vibrations of 10-100Hz and power frequency interference of 50Hz, with signal delay ≤1ms and waveform distortion ≤3%. S4, Control Signal Filtering Wavelet filtering (3-layer dB4, soft threshold λ=σ√(21nN)) is used to process the control signals of gripper, limit gripper and clamping roller, improving the signal-to-noise ratio by ≥20dB and ensuring control accuracy. S5, Multi-source signal fusion The adaptive weighted fusion algorithm (weights based on signal-to-noise ratio) fuses data from multiple sensors, improving data reliability by 40% and system stability by 30%.
[0013] Preferably, in step S1, the complete parameter configurations for two typical application scenarios are specified, where the silk cake diameter is 200-250mm, the thickness is 50-60mm, and the target tension is 8-15cN; the A* path planning weights are α=0.8 and β=0.2, the fuzzy PID parameter Kp is 0.9-1.2; the Kalman filter Q value is 0.008-0.01 and the R value is 0.08-0.1, and the wavelet filter uses 3 layers of db4. These parameters provide a standardized benchmark for subsequent processes, ensuring technical repeatability.
[0014] Preferably, in step S2, the entire process from wire end positioning to winding start is fully automated. The suction nozzle moves and positions along path A* with an accuracy of ±0.3-0.4mm. The limiting claw adjusts the pressure in real time in 0.01N increments, ensuring tension fluctuation ≤±0.2cN, fixed-length cutting accuracy ±0.1mm, and cut surface flatness ≤0.05mm. These are quantified accuracy indicators for the entire process.
[0015] Preferably, in step S4, this step integrates intelligent algorithms and micro-force control to achieve a coaxiality error of ≤0.1mm for double thread end capture, a coaxial pressure control accuracy of ±0.05N and a temperature compensation of 0.02% / ℃; for polyester, the twisting is performed at 100r / min for 5 turns, and for nylon, the twisting is performed at 80r / min for 6 turns, and the residual thread end after fine finishing is ≤0.4-0.5mm.
[0016] Preferably, in step S5, the system's rapid reset time is ≤1s, the entire process parameters are recorded with 1ms accuracy and a database is built, and parameters such as the number of splice turns can be automatically optimized, coupled with a graded alarm fault diagnosis mechanism of early warning-calibration-stop.
[0017] (ii) Beneficial effects Compared with the prior art, the present invention provides an automatic twisting device and method for yarn cake changing based on intelligent algorithms and precise control, which has the following beneficial effects: 1. This automatic yarn cake changing and splicing device and method based on intelligent algorithms and precise control achieves in-depth linkage adjustment of multiple components such as intelligent nozzle displacement (±0.5mm), adaptive gripper pressure (±0.05N), and intelligent splicer speed through a hierarchical architecture and multiple intelligent algorithms (A* path planning, fuzzy PID, distributed MPC) in the central intelligent control module. The collaborative error is ≤0.1mm, which improves efficiency by 40% compared with traditional systems. Through the intelligent algorithm module (including improved A* path planning algorithm and fuzzy PID adaptive control algorithm), the entire process from yarn end picking, cutting, splicing to trimming is automated, completely eliminating manual intervention and greatly improving production efficiency. The path planning accuracy reaches ±0.5mm and the planning cycle is ≤100ms.
[0018] 2. This automatic twisting device and method for yarn cake changing based on intelligent algorithms and precise control introduces knotting micro-force control technology (position-force hybrid control) and applies it to micro-force feedback limiting claws, adaptive clamps, and precision clamping rollers. It accurately senses and controls the minute force of the yarn (±0.02N), ensuring stable yarn transmission (tension fluctuation ±0.5cN) and precise splicing. The splice scrap rate is reduced by 25%, and the quality stability is improved by 30%. By adopting knotting micro-force control technology (position-force hybrid control algorithm), the twisting force is precisely adjusted (control accuracy ±0.05N), ensuring the splicing accuracy and stability of new and old yarns, significantly reducing the splice scrap rate, ensuring winding quality, and reducing the tension fluctuation range to ±0.5cN.
[0019] 3. This automatic twisting device and method for yarn cake changing based on intelligent algorithms and precise control effectively filters noise signals such as electromagnetic interference (50Hz) and mechanical vibration (10-100Hz) in the production environment by using a combined filtering algorithm (Kalman filter + moving average filter + wavelet filter) in the intelligent shearing mechanism and various sensors. The data acquisition error rate is reduced to below 0.1%, the continuous operation time of the equipment is increased by 50%, and the maintenance cost is reduced by 20%. The use of filtering algorithms (Kalman filter + moving average filter + wavelet filter) to process equipment signals effectively reduces 10-100Hz mechanical vibration and 50Hz power frequency interference, improves the signal-to-noise ratio by ≥20dB, shortens the changing interval, improves the continuous operation efficiency of the equipment, and at the same time reduces labor costs and material losses, resulting in a significant reduction in production costs. Attached Figure Description
[0020] Figure 1 This is a schematic diagram of the automatic splicing device of the present invention.
[0021] In the diagram: 1. Intelligent suction nozzle; 2. Adaptive gripper assembly; 3. Micro-force feedback limiting gripper; 4. Intelligent shearing mechanism; 41. First shears; 42. Second shears; 5. Intelligent splicer; 6. Precision clamping roller; 7. Central intelligent control module; 8. Leather roller. Detailed Implementation
[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] Please see Figure 1 The present invention provides a technical solution: An automatic twisting device for changing yarn cakes based on intelligent algorithms and precise control includes an intelligent suction nozzle 1, an adaptive gripper assembly 2, a micro-force feedback limiting gripper 3, an intelligent shearing mechanism 4, an intelligent twister 5, a precision holding roller 6, a central intelligent control module 7, and a leather roller 8. The intelligent shearing mechanism 4 includes a first shear 41 and a second shear 42.
[0024] Furthermore, the intelligent suction nozzle 1 is used to pick up the ends of the silk cake, connects to an air compressor, and is equipped with a displacement sensor and an intelligent algorithm chip (using an improved A* path planning algorithm, path evaluation function f(n)=g(n)+0.8|x n -x_goal|+0.2|y n -y_goal|, positioning accuracy ±0.5mm).
[0025] Furthermore, the adaptive gripper assembly 2 has a two-finger structure and a built-in pressure sensor and microprocessor (using a fuzzy PID adaptive control algorithm, Kp=0.9-1.2, Ki=0.8-1.1, Kd=0.6-0.9, gripping force 0.5-5N, accuracy ±0.05N).
[0026] Furthermore, the micro-force feedback limiting claw 3 has an arc-shaped contact surface and a built-in micro-force feedback device (strain gauge force sensor, accuracy class 0.1, sampling frequency 1kHz), and adopts a position-force hybrid control algorithm (position accuracy ±5μm, force accuracy ±0.02N).
[0027] Furthermore, the first shears 41 and the second shears 42 are each equipped with an intelligent control unit and a filtering algorithm (Kalman filter Q=0.01, R=0.1; moving average window 5) to control the shearing action according to the tension threshold.
[0028] Furthermore, the intelligent splicer 5 is a knotting mechanism that incorporates an intelligent algorithm module (adjusting splicing parameters based on thread parameters) and a tension sensor (accuracy ±0.5cN).
[0029] Furthermore, the precision bearing roller 6 has a double roller pressing structure and adopts a closed-loop control system (incremental PID, Kp=1.2, Ki=0.05, Kd=0.3, pressure 1-10N, accuracy ±0.05N).
[0030] Furthermore, the central intelligent control module 7 adopts a hierarchical control architecture, controls the collaborative work of various components through the EtherCAT bus (cycle ≤1ms), integrates multiple intelligent algorithms, and supports redundant and fault-degraded operation.
[0031] The splicing method of this automatic splicing device for yarn cake changing based on intelligent algorithms and precise control includes the following steps: S1, First Silk Pancake Launch Phase The intelligent suction nozzle 1 moves along the axis of the first yarn cake according to the path planned by the A* algorithm (speed 50mm / s) to absorb the yarn end, which then passes sequentially through the open adaptive gripper assembly 2 (initial gripping force 0.5N) and the micro-force feedback limiting claw 3 (pre-tightening force 0.2N). During the process, the micro-force feedback limiting claw 3 adjusts its pressure in real time (sampling period 10ms) to ensure yarn stability. After receiving the sensor signal, the central intelligent control module 7 instructs the adaptive gripper assembly 2 and the micro-force feedback limiting claw 3 to close synchronously (the gripping force is adjusted according to the yarn diameter). Subsequently, the first scissors 41 eliminates signal interference based on a filtering algorithm (wavelet filtering 3 layers db4) and accurately cuts the yarn according to the set tension threshold. Finally, the intelligent suction nozzle 1 feeds the yarn end into the inlet of the roller 8 and starts winding. The complete parameter configuration for two typical application scenarios is defined, where the yarn cake diameter is 200-250mm, the thickness is 50-60mm, and the target tension is 8-15cN; the A* path planning weights α=0.8, β=0.2, and the fuzzy PID parameter Kp is 0.9- 1.2; Kalman filter Q value 0.008 -0.01, R value 0.08 -0.1, wavelet filter using 3 layers db4. These parameters provide a standardized benchmark for subsequent processes, ensuring technical repeatability; S2, the stage of connecting the old and new silk cakes. After the second yarn package is in place, the intelligent suction nozzle 1 picks up its new yarn end (positioning accuracy ±0.3mm) and sucks the yarn end of the first yarn package into the gripper. The two yarns are then automatically joined by the precision holding roller 6 (pressure 3.5-5.0N, adjusted according to yarn material) and the intelligent splicer 5 (knotting according to algorithm: polyester 0.2mm: splicing speed 100r / min, 5 turns; nylon 0.4mm: speed 80r / min, 6 turns). During the process, the tension sensor monitors in real time (data refresh rate 1kHz). After joining, the second scissors 42 precisely trims the remaining yarn end (length ≤0.5mm), and all actuators reset (time ≤1s) to prepare for the next round of operation. This achieves full automation from yarn end positioning to winding start. The suction nozzle moves and positions along path A* with an accuracy of ±0.3-0.4mm; the limiting gripper adjusts the pressure in real time in 0.01N steps to ensure tension fluctuation ≤±0.2cN and fixed-length cutting accuracy ±0.1mm. Furthermore, the flatness of the cut surface is ≤0.05mm. This represents a fully quantified precision indicator across the entire process. S3, Sensor signal filtering A Kalman filter + moving average combined algorithm is used to process tension, displacement and force sensor signals, filtering mechanical vibrations of 10-100Hz and power frequency interference of 50Hz, with signal delay ≤1ms and waveform distortion ≤3%. S4, Control Signal Filtering Wavelet filtering (3 layers, db4, soft threshold λ=σ√(21nN)) is used to process the control signals of the gripper, limiter, and clamping roller, improving the signal-to-noise ratio by ≥20dB and ensuring control accuracy. This step integrates intelligent algorithms and micro-force control to achieve a coaxiality error of ≤0.1mm for dual-thread head capture, and a coaxial pressure control accuracy of ±0.05N with a temperature compensation of 0.02% / ℃. For polyester, the splicing is done at 100r / min for 5 turns, and for nylon, it is done at 80r / min for 6 turns. After fine finishing, the residual thread ends are ≤0.4-0.5mm. S5, Multi-source signal fusion The adaptive weighted fusion algorithm (weights based on signal-to-noise ratio) fuses data from multiple sensors, improving data reliability by 40% and system stability by 30%. The system's rapid reset time is ≤1s. It records all process parameters with 1ms accuracy and builds a database. It can also automatically optimize parameters such as the number of splice turns, and is equipped with a graded alarm fault diagnosis mechanism of early warning, calibration, and shutdown.
[0032] This automatic yarn feeding and splicing device and method based on intelligent algorithms and precise control involves the following steps: In the first yarn feeding stage, an intelligent suction nozzle moves along the axis of the first yarn feed according to the path planned by the A* algorithm (speed 50mm / s) to absorb the yarn end, which then passes sequentially through an open adaptive gripper (initial clamping force 0.5N) and a micro-force feedback limiting gripper (pre-tightening force 0.2N). During this process, the limiting gripper adjusts its pressure in real time (sampling period 10ms) to ensure yarn stability. After receiving sensor signals, the central control module instructs the gripper and limiting gripper to close synchronously (clamping force adjusted according to yarn diameter). Subsequently, the scissors 1, based on a filtering algorithm (wavelet filtering 3 layers db4), eliminates signal interference and precisely cuts the yarn according to the set tension threshold. Finally... The suction nozzle feeds the thread end into the roller inlet, initiating winding. During the new and old yarn cake joining stage: after the second yarn cake arrives, the intelligent suction nozzle picks up its new thread end (positioning accuracy ±0.3mm) and sucks the thread end of the first yarn cake into the gripper. The two yarns sequentially pass through a precision holding roller (pressure 3.5-5.0N, adjusted according to yarn material) to apply pressure, and an intelligent splicer (knotting according to an algorithm: polyester 0.2mm: splicing speed 100r / min, 5 turns; nylon 0.4mm: speed 80r / min, 6 turns) to complete automatic joining. During the process, a tension sensor monitors in real time (data refresh rate 1kHz). After joining, scissors 2 precisely trim any remaining thread ends (length ≤0.5mm). Each actuator resets (time ≤ 1s), preparing for the next round of operation. Through the hierarchical architecture and multiple intelligent algorithms (A* path planning, fuzzy PID, distributed MPC) in the central intelligent control module, the system achieves deep linkage adjustment of multiple components such as intelligent nozzle displacement (±0.5mm), adaptive gripper pressure (±0.05N), and intelligent splicer speed, with a coordination error ≤ 0.1mm. Compared with the traditional system, the efficiency is improved by 40%. Through the intelligent algorithm module (including improved A* path planning algorithm and fuzzy PID adaptive control algorithm), the system realizes fully automated operation from thread end picking, cutting, splicing to trimming, completely eliminating manual intervention and significantly improving production efficiency. The path planning accuracy reaches ±0.5mm, and the planning cycle is ≤100ms. By introducing knotting micro-force control technology (position-force hybrid control) and applying it to micro-force feedback limit claws, adaptive grippers, and precision clamping rollers, it accurately senses and controls the minute force of the yarn (±0.02N), ensuring stable yarn transmission (tension fluctuation ±0.5cN) and precise splicing. The splice scrap rate is reduced by 25%, and the quality stability is improved by 30%. By adopting knotting micro-force control technology (position-force hybrid control algorithm), the splicing force is precisely adjusted (control accuracy ±0.05N), ensuring the splicing accuracy and stability of new and old yarns, significantly reducing the splice scrap rate, ensuring winding quality, and reducing the tension fluctuation range to ±0.The 5cN system effectively filters out noise signals such as electromagnetic interference (50Hz) and mechanical vibration (10-100Hz) in the production environment by employing a combined filtering algorithm (Kalman filter + moving average filter + wavelet filter) in the intelligent shearing mechanism and various sensors. This reduces the data acquisition error rate to below 0.1%, increases continuous equipment operating time by 50%, and reduces maintenance costs by 20%. The filtering algorithm (Kalman filter + moving average filter + wavelet filter) effectively reduces 10-100Hz mechanical vibration and 50Hz power frequency interference, improving the signal-to-noise ratio by ≥20dB, shortening roll change intervals, and increasing continuous equipment operating efficiency. Simultaneously, it reduces labor costs and material waste, resulting in a significant decrease in production costs.
[0033] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. An automatic twisting device for yarn cake changing based on intelligent algorithms and precise control, comprising an intelligent suction nozzle (1), an adaptive gripper assembly (2), a micro-force feedback limiting gripper (3), an intelligent shearing mechanism (4), an intelligent twister (5), a precision holding roller (6), a central intelligent control module (7), and a leather roller (8), characterized in that: The intelligent cutting mechanism (4) includes a first pair of scissors (41) and a second pair of scissors (42).
2. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The intelligent suction nozzle (1) is used to adsorb the ends of the silk cake, is connected to an air compressor, and is equipped with a displacement sensor and an intelligent algorithm chip. It adopts an improved A* path planning algorithm, and the path evaluation function is f(n)=g(n)+0.8|x n -x_goal|+0.2|y n -y_goal|, positioning accuracy ±0.5mm.
3. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The adaptive gripper assembly (2) is a two-finger structure with a built-in pressure sensor and microprocessor. It adopts a fuzzy PID adaptive control algorithm with Kp=0.9-1.2, Ki=0.8-1.1, Kd=0.6-0.9, a gripping force of 0.5-5N, and an accuracy of ±0.05N.
4. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The micro-force feedback limiting claw (3) has an arc-shaped contact surface, a built-in micro-force feedback device, a strain gauge force sensor with an accuracy of 0.1 grade and a sampling frequency of 1kHz. It adopts a position-force hybrid control algorithm with a position accuracy of ±5μm and a force accuracy of ±0.02N.
5. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The first shears (41) and the second shears (42) are each equipped with an intelligent control unit and a filtering algorithm, with Kalman filter Q=0.01, R=0.1, and sliding average window 5, and the shearing action is controlled according to the tension threshold.
6. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The intelligent splicer (5) is a knotting mechanism with an embedded intelligent algorithm module. It adjusts the splicing parameters and tension sensor based on the thread parameters, with an accuracy of ±0.5cN.
7. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The precision bearing roller (6) is a double roller pressing structure, adopts a closed-loop control system, incremental PID, Kp=1.2, Ki=0.05, Kd=0.3, pressure 1-10N, accuracy ±0.05N.
8. The automatic twisting and splicing device for yarn cake changing based on intelligent algorithms and precise control according to claim 1, characterized in that: The central intelligent control module (7) adopts a hierarchical control architecture. Through the EtherCAT bus, with a cycle of ≤1ms, it controls the collaborative work of various components, integrates multiple intelligent algorithms, and supports redundant and fault-degraded operation.
9. A splicing method for an automatic splicing device for yarn cake changing based on intelligent algorithms and precise control, as described in any one of claims 1-8, characterized in that, Includes the following steps: S1, First Silk Pancake Launch Phase The intelligent suction nozzle (1) moves along the axis of the first yarn cake according to the path planned by the A* algorithm at a speed of 50 mm / s, adsorbs the yarn end, and passes through the opened adaptive gripper assembly (2) in sequence. The initial clamping force is 0.5 N, and the micro-force feedback limiting claw (3) has a pre-tightening force of 0.2 N. During the process, the micro-force feedback limiting claw (3) adjusts the pressure in real time with a sampling period of 10 ms to ensure the stability of the yarn. After receiving the sensor signal, the central intelligent control module (7) instructs the adaptive gripper assembly (2) and the micro-force feedback limiting claw (3) to close synchronously. The clamping force is adjusted according to the diameter of the yarn. Then, the first scissors (41) uses a filtering algorithm and wavelet filtering with 3 layers of db4 to eliminate signal interference and accurately cuts the yarn according to the set tension threshold. Finally, the intelligent suction nozzle (1) sends the yarn end into the inlet of the roller (8) and starts winding. S2, the stage of connecting the old and new silk cakes. After the second yarn cake is in place, the intelligent suction nozzle (1) picks up its new yarn end with a positioning accuracy of ±0.3mm and sucks the yarn end of the first yarn cake into the clamp. The two yarns pass through the precision holding roller (6) in sequence, with a pressure of 3.5-5.0N. The pressure is adjusted according to the yarn material. The intelligent splicer (5) knots the yarn according to the algorithm. For polyester 0.2mm, the splicing speed is 100r / min and the number of turns is 5. For nylon 0.4mm, the speed is 80r / min and the number of turns is 6. Automatic docking is completed. During the process, the tension sensor monitors in real time with a data refresh rate of 1kHz. After docking is completed, the second scissors (42) precisely trims the remaining yarn end with a length of ≤0.5mm. Each actuator is reset with a time of ≤1s to prepare for the next round of operation. S3, Sensor signal filtering A Kalman filter + moving average combined algorithm is used to process tension, displacement and force sensor signals, filtering mechanical vibrations of 10-100Hz and power frequency interference of 50Hz, with signal delay ≤1ms and waveform distortion ≤3%. S4, Control Signal Filtering Wavelet filtering, 3-layer dB4, soft threshold λ=σ√21nN, to process control signals of gripper, limit gripper and clamping roller, improve signal-to-noise ratio by ≥20dB and ensure control accuracy; S5, Multi-source signal fusion The adaptive weighted fusion algorithm, with weights based on the signal-to-noise ratio, fuses data from multiple sensors, improving data reliability by 40% and system stability by 30%.
10. The automatic twisting method for yarn cake changing based on intelligent algorithms and precise control according to claim 5, characterized in that, In step S1, the complete parameter configurations for two typical application scenarios are defined, with a yarn cake diameter of 200-250mm, a thickness of 50-60mm, and a target tension of 8-15cN; A* path planning weights α=0.8, β=0.2, and fuzzy PID parameter Kp of 0.9-1.2; Kalman filter Q value of 0.008-0.01, R value of 0.08-0.1, and wavelet filtering using 3 layers of db4. These parameters provide a standardized benchmark for subsequent processes, ensuring technical repeatability. In step S2, full automation is achieved from yarn head positioning to winding start. The suction nozzle moves and positions along the A* path with an accuracy of ±0.3-0.4mm; the limiting claw adjusts the pressure in real time with a step size of 0.01N, ensuring tension fluctuation ≤±0.2cN, fixed-length cutting accuracy ±0.1mm, and cut surface flatness ≤0.05mm. The precision indicators for the entire process are quantified. In step S4, this step integrates intelligent algorithms and micro-force control to achieve a coaxiality error of ≤0.1mm for double thread head capture, a coaxial pressure control accuracy of ±0.05N with a temperature compensation of 0.02% / ℃. For polyester, 5 twists are made at 100r / min, and for nylon, 6 twists are made at 80r / min. After fine finishing, the residual thread head is ≤0.4-0.5mm. In step S5, the system's rapid reset time is ≤1s. The entire process parameters are recorded with 1ms accuracy and a database is built. It can also automatically optimize parameters such as the number of twists and turns, and is equipped with a graded alarm fault diagnosis mechanism of early warning-calibration-stop.