A yarn on-line detection and tension closed-loop control system

Abstract

The present application belongs to the field of textile machinery, and discloses a yarn online detection and tension closed-loop control system and method, aiming to solve the technical problems of low tension detection precision, response lag and inability to adapt to working condition changes in the prior art. The system comprises a perception layer, a control layer, an execution layer and an application layer. The perception layer collects yarn dynamic tension values in real time through low-friction yarn guide tension wheels in cooperation with high-precision strain sensors, and simultaneously acquires yarn diameter, hairiness amount and defect features through photoelectric or laser non-contact measurement units. The control layer adopts a main controller to run adaptive PID and feedforward compensation algorithms, thereby realizing tension-velocity cascade closed loop. The execution layer drives servo adjusting wheels to quickly correct tension deviation. The present application can realize multi-parameter fusion online detection of tension, diameter, hairiness and defects, control tension fluctuation within 3%, and response time is less than 20 ms, thereby significantly reducing the breakage rate and improving the forming quality.

Description

Technical Field

[0001] This invention belongs to the field of textile machinery, specifically relating to an online yarn detection and tension closed-loop control system. Background Technology

[0002] Yarn tension is a core process parameter that determines the quality of the finished product, breakage rate, and production efficiency in textile weaving, winding, twisting, and covering processes. During high-speed weaving and winding, the stability of yarn tension directly affects the surface uniformity, strength, and continuous operation capability of the final fabric. With the development of textile equipment towards higher speeds and finer denier yarns, higher requirements are placed on the precision and response speed of tension control.

[0003] Among these, closed-loop tension control is a crucial step in achieving stable yarn feeding. In existing technologies, tension control is mostly achieved using mechanical passive adjustment devices combined with simple sensors. These devices typically consist of a tension disc, spring pressure plates, or magnetic dampers, relying on manual experience to preset pressure values ​​or damping parameters. When the yarn is running, tension fluctuations are passively absorbed through the deformation or resistance changes of the mechanical components, and then fed back to the actuator for adjustment.

[0004] However, in existing implementations, the aforementioned passive adjustment mechanism suffers from fundamental technical limitations. Due to the inherent inertia of mechanical transmission, the accumulated delay in the closed-loop path from the generation of tension deviation to the response of the actuator often exceeds hundreds of milliseconds. Under high-speed conditions, the yarn passes through the detection point at speeds of several meters or even tens of meters per second, and the generation and evolution of tension fluctuations are completed in an extremely short time. The response speed of existing control systems is far from sufficient to suppress tension deviations before they expand. The direct consequence of this response lag is that tension fluctuations are amplified in the closed-loop circuit, leading to frequent overshoot and oscillation phenomena, ultimately manifesting as uneven yarn formation and a persistently high breakage rate. Simultaneously, mechanical adjustment mechanisms experience wear and elasticity decay during long-term operation, further weakening the stability and repeatability of control performance. Summary of the Invention

[0005] The purpose of this invention is to provide an online yarn detection and tension closed-loop control system, which can effectively solve the technical problems in the prior art such as low yarn tension detection accuracy, slow response, inability to adapt to changes in working conditions, lack of multi-parameter linkage detection, and offline data traceability.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: an online yarn detection and tension closed-loop control system, comprising a sensing layer, a control layer, an execution layer, and an application layer; the system realizes non-contact online detection of yarn tension, diameter, hairiness, and defects, as well as tension closed-loop control based on adaptive PID and feedforward compensation.

[0007] S1: The sensing layer uses a low-friction yarn tension wheel in conjunction with a high-precision strain sensor to collect the dynamic tension value of the yarn in real time. At the same time, it uses a photoelectric or laser non-contact measurement unit to simultaneously acquire the yarn diameter, hairiness, and defect characteristics. The control layer uses a main controller to run adaptive PID, fuzzy PID, and feedforward compensation algorithms to achieve a tension-speed-position cascade closed loop. The execution layer drives the servo or stepper adjustment wheel to move according to the control command, or controls the electromagnetic brake and frequency conversion mechanism to quickly correct the tension deviation. The application layer provides a human-machine interface to realize parameter setting, real-time curve display, alarm recording, and data uploading functions.

[0008] S2: The online detection method includes the following steps: a low-friction yarn guide tension wheel collects tension signals, and after hardware filtering and digital filtering, the average tension, fluctuation, peak value and abnormal threshold are calculated; a photoelectric or laser measurement unit simultaneously collects diameter, hair and defect signals, and uses Kalman filtering or moving average to suppress interference and drift, extracts feature parameters and determines abnormalities.

[0009] S3: The tension closed-loop control method includes the following steps: input the target tension value according to the yarn type and process, and calculate the deviation between the measured tension and the target tension in real time; the adaptive PID module adjusts the proportional, integral, and derivative parameters online according to changes in machine speed and roll diameter; the feedforward compensation module outputs control quantities in advance to suppress overshoot based on changes in machine speed, acceleration, and roll diameter; the tension loop as the main loop and the speed loop as the secondary loop form a cascaded closed-loop structure; the drive adjustment wheel rotation angle or braking torque quickly corrects the tension, and triggers protection logic and alarms when the limit is exceeded or yarn breaks.

[0010] Preferably, the sensing layer includes a strain gauge tension sensor, a fiber optic tension sensor, a non-contact photoelectric diameter gauge, a laser feather detector, an encoder, and a roll diameter monitoring unit. Each sensor transmits data to the control layer through a unified interface.

[0011] Preferably, the main controller of the control layer adopts a DSP or PLC, runs an adaptive PID control algorithm, and realizes one-click switching of yarn varieties through online self-learning of yarn characteristics, eliminating the need for frequent manual parameter adjustment.

[0012] Preferably, the adaptive PID algorithm automatically adjusts the proportional coefficient, integral coefficient, and derivative coefficient based on historical operating data to adapt to changes in different yarn materials, counts, machine speeds, and roll diameters.

[0013] Preferably, the feedforward compensation algorithm pre-calculates the compensation amount based on the vehicle speed change rate and acceleration roll diameter change rate, and outputs it after superimposing it with the feedback control amount, effectively suppressing overshoot and oscillation caused by sudden changes in operating conditions.

[0014] Preferably, the execution layer includes a servo-driven active tension adjustment wheel, an electromagnetic brake, an electromagnetic clutch, and a frequency conversion yarn feeding mechanism, which can respond to changes in tension deviation within a predetermined time.

[0015] Preferably, the multi-parameter fusion detection achieves integrated online monitoring of tension, diameter, fuzz, and defects, with the synchronous acquisition frequency reaching a predetermined acquisition frequency, and tension fluctuations controlled within the allowable range.

[0016] Preferably, the application layer supports MODBUS, Profinet, and EtherNet / IP digital interfaces, and can be connected to MES systems and cloud platforms to realize full-process data storage, historical query, report export, and remote operation and maintenance.

[0017] Preferably, the yarn breakage and defect linkage protection function includes: real-time detection of yarn breakage signals and defect over-limit signals, triggering audible and visual alarms and immediately stopping the machine, while recording the fault time and location information for source tracing analysis.

[0018] Compared with existing technologies, the present invention has the following advantages: stable tension, significantly reduced fluctuations in closed-loop control, and a significant decrease in breakage rate; online non-destructive testing, with non-contact and low-friction detection reducing fuzz and yarn damage; high efficiency and intelligence, with adaptive batch change eliminating manual parameter adjustment and significantly improving efficiency; controllable quality, with full-process data traceability and defect early warning improving yield; and compatibility and upgradeability, with modular design adapting to existing equipment upgrades and reducing implementation costs. Attached Figure Description

[0019] Figure 1 This is a schematic diagram of the overall technical architecture of the yarn online detection and tension closed-loop control system proposed in this invention; Figure 2 This is a schematic diagram of the core principle framework of tension closed-loop control based on adaptive PID and feedforward compensation in this invention. Figure 3 This is a flowchart illustrating the logical flow of online detection and feature extraction of yarn parameters in this invention. Figure 4 This is a schematic diagram of the multi-level interaction relationship and data flow between the perception layer, control layer, execution layer and application layer in this invention. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0021] Example 1 This embodiment uses the winding process in a textile weaving production line as an application scenario to explain in detail the system architecture construction and workflow of the yarn online detection and tension closed-loop control system of the present invention.

[0022] In the winding process, yarn is drawn from the spinning frame or twisting machine and enters the winding mechanism of the winding machine via a yarn guide rod. During this process, yarn tension directly affects the quality of the package formation and the breakage rate. Traditional technology uses mechanical passive adjustment devices, which have large response delays and low control accuracy. The system provided in this embodiment achieves non-contact online detection of yarn tension, diameter, hairiness, and defects through the collaborative work of the sensing layer, control layer, execution layer, and application layer, and realizes closed-loop tension control based on adaptive PID and feedforward compensation algorithms.

[0023] System Architecture Construction The system in this embodiment includes four layers: perception layer, control layer, execution layer, and application layer. Data interaction and functional collaboration between the layers are achieved through standardized communication interfaces.

[0024] The sensing layer is positioned at several key locations along the yarn's path to collect multi-dimensional process parameters of the yarn in real time. The core components of the sensing layer include a low-friction yarn tension wheel, a high-precision strain sensor, a non-contact photoelectric diameter gauge, a laser hairiness detector, a rotary encoder, and a roll diameter monitoring unit. The low-friction yarn tension wheel is positioned on the path before the yarn enters the winding mechanism. Its shaft is connected to the high-precision strain sensor, which senses changes in the force on the tension wheel and outputs an electrical signal proportional to the yarn tension. The non-contact photoelectric diameter gauge is positioned downstream of the low-friction yarn tension wheel and measures the yarn diameter using the photoelectric beam principle, achieving a measurement accuracy of 0.01 mm. The laser hairiness detector is positioned adjacent to the non-contact photoelectric diameter gauge and uses the laser scattering principle to detect the hair distribution and length characteristics on the yarn surface. The rotary encoder is located at the drive shaft end of the yarn tension wheel and measures the yarn's running speed and cumulative travel. The roll diameter monitoring unit is positioned next to the winding mechanism and monitors changes in the roll diameter of the yarn on the drum in real time using ultrasonic or contact distance sensors. Each sensor transmits its collected data to the control layer through a unified data interface, with a sampling frequency of no less than 1,000 times per second.

[0025] The control layer uses a main controller as the core computing unit, which can be either a DSP chip or a PLC programmable logic controller. The main controller integrates three core algorithm modules: an adaptive PID control module, a fuzzy PID control module, and a feedforward compensation control module. The adaptive PID control module tunes the proportional coefficient KP, integral coefficient KI, and derivative coefficient KD online based on real-time collected operating parameters such as yarn speed and yarn diameter change rate, ensuring the control parameters always adapt to the current yarn characteristics. The fuzzy PID control module further optimizes traditional PID parameters using fuzzy logic. When the system detects nonlinear tension fluctuations, the fuzzy PID module outputs a correction value, which is then added to the output of the adaptive PID. The feedforward compensation control module pre-calculates the compensation value based on the yarn speed change rate and acceleration / yarn diameter change rate. This compensation value is then superimposed with the feedback control value to form the final control command. The main controller also features a cascaded closed-loop structure consisting of a tension loop and a speed loop. The tension loop acts as the main loop, responsible for closed-loop control of tension deviation, while the speed loop acts as the secondary loop, responsible for the speed control of the actuator. The two loops work together to improve system response speed and stability.

[0026] The execution layer is located at the adjustment mechanism along the yarn path and is used to adjust the yarn tension according to the instructions from the control layer. The core components of the execution layer include a servo-driven active tension adjustment wheel, an electromagnetic brake, an electromagnetic clutch, and a frequency-controlled yarn feeding mechanism. The servo-driven active tension adjustment wheel is positioned on the yarn path between the low-friction guide tension wheel and the winding mechanism. Its shaft is driven by a servo motor, and the controller actively adjusts the yarn tension by adjusting the servo motor's angle and speed, with a response time controllable within 20 milliseconds. The electromagnetic brake is located at the braking end of the guide tension wheel. When rapid tension reduction is needed, the electromagnetic brake applies braking torque to quickly decelerate the tension wheel. The electromagnetic clutch is located between the tension adjustment wheel and the drive motor, used for rapid engagement and disengagement of power. The frequency-controlled yarn feeding mechanism is located at the yarn source and achieves precise control of the yarn feeding speed by adjusting the speed of the frequency-controlled motor. Each execution mechanism is connected to the main controller via a drive circuit, receiving PWM pulse width modulation signals or analog control signals output by the main controller.

[0027] The application layer, located in the operator's control room, provides a human-machine interface for parameter setting, real-time curve display, alarm logging, and data uploading. The hardware for the application layer can be a touchscreen HMI or an industrial panel PC. The application layer software includes a parameter configuration module, a real-time display module, a historical storage module, and a communication upload module. The parameter configuration module provides an input interface for process parameters such as yarn type selection, target tension setting, machine speed setting, and roll diameter setting. The real-time display module shows the changing trends of parameters such as tension, speed, diameter, and hairiness in real-time curve form, and can switch to numerical display mode. The historical storage module automatically stores operating data to a local database, supporting historical queries and report export based on time period, equipment number, and anomaly type. The communication upload module supports industrial communication protocols such as MODBUS, Profinet, and EficatNet / IP, enabling data upload to the MES (Manufacturing Execution System) or cloud platform for data sharing and remote operation and maintenance.

[0028] Workflow Deduction After the system is powered on, it first executes a self-test program. The main controller performs communication tests on each sensor in the sensing layer. After confirming that the sensors are working properly, it sends initialization signals to the servo drivers and electromagnetic brakes in the execution layer, causing the actuators to reset to their initial positions. The human-machine interface in the application layer displays the system's ready status, waiting for the operator to input process parameters.

[0029] The operator selects the yarn type in the parameter configuration module of the application layer, including yarn material such as cotton yarn, polyester yarn, nylon yarn, etc., and yarn count such as 20 count, 40 count, 60 count, etc. The system automatically matches a preset target tension range based on the yarn type. The operator can set a specific target tension value within this range; for example, for 40 count cotton yarn, the target tension is set to 80 to 120 grams. Simultaneously, the operator sets the target speed of the winding machine and the expected maximum drum diameter. After the parameters are set, the system enters a standby state.

[0030] The operator guides the yarn onto the path between the low-friction guide tension wheel and the tension adjusting wheel, and then presses the start button. The system enters online monitoring and closed-loop control mode.

[0031] During the online detection phase, the low-friction yarn tension wheel of the sensing layer rotates with the yarn movement, and the rotating shaft transmits the yarn tension to the high-precision strain sensor. The output signal of the strain sensor undergoes primary filtering by a hardware filtering circuit to remove high-frequency electromagnetic interference before entering the A / D conversion channel of the main controller. The main controller further performs Kalman filtering on the digital signal to eliminate random noise and sensor drift effects, and calculates the mean, fluctuation, peak value, and abnormal threshold of the current tension in real time. Simultaneously, a non-contact photoelectric diameter gauge synchronously acquires yarn diameter data, and a laser hairiness detector acquires hairiness characteristic data. The rotary encoder outputs pulse signals, and the main controller calculates the yarn running speed by counting the pulse frequency. The roll diameter monitoring unit monitors the roll diameter change in real time and sends the roll diameter data to the main controller. The sampling frequency of each parameter is no less than 1000 times per second to ensure real-time detection.

[0032] In the feature extraction and anomaly detection stage, the main controller preprocesses the collected raw data and extracts representative features of each parameter. For tension signals, the average tension over the past second is calculated as the current representative tension value, and the standard deviation of tension is calculated as a fluctuation index. The maximum peak value and minimum trough value are recorded for analyzing tension impact. For diameter signals, the average diameter is calculated and compared with the standard value to determine whether there are thick or thin defects. For fuzz signals, the total amount of fuzz and the proportion of excessively long fuzz are calculated to determine whether the fuzz exceeds the standard. For defect signals, anomalies such as neps, oil stains, and broken yarns are detected. The main controller compares each feature parameter with preset thresholds, and triggers the corresponding alarm flag when any parameter exceeds the threshold range.

[0033] During the tension closed-loop control phase, the main controller calculates the deviation between the measured tension and the target tension in real time. This deviation is input to the adaptive PID control module, which tunes the KP, KI, and KD parameters online based on the current vehicle speed and the rate of change of roll diameter. Specifically, as the vehicle speed increases, the system automatically increases the proportional and derivative coefficients to accelerate the response; as the roll diameter increases, the system appropriately increases the integral coefficient to enhance steady-state accuracy. The feedforward compensation control module pre-calculates the compensation amount based on the rate of change of vehicle speed and the rate of change of roll diameter due to acceleration. This compensation amount is superimposed on the output of the adaptive PID to form a feedforward-feedback composite control quantity. The main controller decomposes the composite control quantity into a tension loop control quantity and a speed loop control quantity. The tension loop control quantity directly controls the rotational position of the tension adjusting wheel, while the speed loop control quantity controls the rotational speed of the tension adjusting wheel.

[0034] After receiving control commands from the main controller, the execution layer drives the servo motor to move the tension adjusting wheel. When the measured tension is higher than the target tension, the control system drives the tension adjusting wheel to rotate in the loosening direction, increasing the wrap angle of the yarn on the adjusting wheel and reducing the tension; when the measured tension is lower than the target tension, it drives the tension adjusting wheel to rotate in the tensioning direction, decreasing the wrap angle and increasing the tension. The electromagnetic brake provides auxiliary braking when the speed of the tension adjusting wheel changes too much, accelerating the response speed. The frequency conversion yarn feeding mechanism adjusts the yarn feeding speed according to the control commands of the speed loop to maintain matching with the winding speed.

[0035] In the protection logic, the main controller monitors the tension value in real time to ensure it does not exceed the safe range. When the tension exceeds the upper threshold (e.g., 150% of the target tension) or falls below the lower threshold (e.g., 50% of the target tension), it is determined to be an abnormal tension, and the system immediately triggers protection actions: sending a stop command to the execution layer, cutting off the power output of the variable frequency yarn feeding mechanism, activating the braking function of the tension adjusting wheel, and simultaneously sending an alarm signal to the application layer. After receiving the alarm signal, the application layer displays the abnormal tension alarm information on the interface and provides audible and visual alerts to the operator. When a yarn breakage signal is detected, the system also triggers emergency stop protection and records the time and location of the yarn breakage for subsequent traceability analysis.

[0036] In the data management phase, the application layer's real-time display module continuously updates real-time curves of parameters such as tension, speed, diameter, and hairiness, allowing operators to intuitively monitor production status through the interface. The historical storage module automatically stores process parameters every minute and second to a local database, including timestamps, equipment numbers, yarn types, target tension, measured tension, speed, roll diameter, and anomaly flags. Operators can retrieve historical data at any time for querying and analysis, or export it as Excel reports for quality traceability. When the application layer is configured with network communication capabilities, the system can upload real-time and historical data to the upper-level MES system or cloud platform via MODBUS, Profinet, or EtherNet / IP protocols, enabling centralized management and remote operation and maintenance of production data.

[0037] This embodiment achieves an organic integration of simultaneous online detection and closed-loop tension control of yarn tension, diameter, hairiness, and defects. Real-time yarn tension information is acquired through a low-friction yarn guide tension wheel and a high-precision strain sensor. Diameter, hairiness, and defect characteristics are obtained through photoelectric and laser non-contact measurements. Fast and stable closed-loop tension control is achieved through adaptive PID and feedforward compensation algorithms. Human-machine interaction and data management are implemented through the application layer. The collaborative work of each layer gives the system technical characteristics such as fast response speed, high control accuracy, strong adaptability, and data traceability.

[0038] Example 2 This embodiment uses the doubling process in a textile weaving production line as an application scenario to illustrate the optimization of the system configuration and control method of the present invention in a multi-spindle batch production environment.

[0039] In the twisting process, multiple spindles operate simultaneously, each with its own independent yarn path and tension control requirements. Traditional single-point control methods are insufficient to meet the needs of centralized management of multiple spindles. This embodiment utilizes a distributed architecture to achieve independent detection and coordinated control of tension across multiple spindles.

[0040] The sensing layer employs a distributed sensor layout. Each spindle is equipped with an independent low-friction yarn guide tension wheel and a high-precision strain sensor, forming an independent tension detection channel. The sensors at each spindle are connected to the main controller via independent A / D conversion channels. The main controller sequentially collects tension data from each spindle using time-division multiplexing. The non-contact photoelectric diameter gauge and laser hairiness detector are arranged in a scanning manner, sequentially detecting yarns from multiple spindles or employing a multi-channel parallel detection scheme. A rotary encoder is positioned at the drive motor shaft of each spindle to measure the actual operating speed of each spindle.

[0041] The control layer employs a two-tier architecture consisting of a master controller and slave controllers. The master controller is responsible for task scheduling, unified parameter distribution, and global data aggregation. Slave controllers are located near each spindle position and independently manage the closed-loop tension control for that spindle. The master controller communicates with each slave controller via a high-speed fieldbus (such as CAN bus or EtherCAT), with a communication cycle controllable to within 1 millisecond. Each slave controller internally runs an adaptive PID algorithm, autonomously adjusting control parameters based on the spindle speed, roll diameter, and yarn characteristics to achieve independent optimized tension control for each spindle position. The master controller periodically collects operational data and alarm information from each slave controller and uploads it to the application layer for centralized display and storage.

[0042] The execution layer is equipped with servo-driven tension adjustment wheels and electromagnetic brakes at each spindle position. The actuator at each spindle position is independently driven and controlled by its corresponding slave controller, enabling independent tension adjustment. When a tension abnormality occurs at a spindle position, the slave controller independently executes a shutdown protection mechanism, without affecting the normal production of other spindle positions. The main controller can be configured to automatically adjust the subsequent winding speed or tension target value of a spindle position when an abnormality is detected, achieving adaptive coordination of the production line.

[0043] The application layer provides a centralized monitoring interface for multiple spindle positions, which can simultaneously display key parameters such as tension, speed, and roll diameter for all spindle positions on the same screen. Operators can set parameters for individual spindle positions or set uniform parameters for all spindle positions in batches. Alarm information is displayed in categories by spindle position for easy and rapid fault location. Historical data is stored separately for each spindle position, supporting quality analysis and comparison by spindle position.

[0044] This embodiment achieves independent detection, independent control, and centralized management of yarn tension across multiple spindles through a distributed sensing, control, and execution architecture, significantly improving the management efficiency and product quality consistency of multi-spindle production lines.

[0045] Example 3 This embodiment uses the coating process in a textile weaving production line as an application scenario to illustrate the special application and extended functions of the present invention in the production of coated yarn.

[0046] In the coating process, the yarn needs to be spirally wrapped under external force, and the tension control accuracy directly affects the uniformity and tightness of the coating layer. This embodiment adds a dedicated coating detection function and control strategy to the basic system configuration.

[0047] The sensing layer, based on the standard configuration, adds a wrapping tension sensor and a wrapping angle detection unit. The wrapping tension sensor is located at the output end of the wrapping device and is used to detect the actual tension of the yarn after wrapping. The wrapping angle detection unit uses a photoelectric encoder to measure the rotation angle of the yarn guide, and is used to calculate the wrapping winding angle and number of turns.

[0048] The control layer adds a dedicated coating control module based on the adaptive PID algorithm. This module establishes a mathematical model of tension, coating speed, and coating angle according to the coating process requirements. When the coating angle changes, the system automatically adjusts the target tension value to ensure the uniformity of the coating layer. When the coating speed changes, the feedforward compensation module adjusts the compensation amount in advance to suppress the impact of sudden speed changes on the tension.

[0049] The execution layer adds a dedicated servo drive mechanism for wrapping, based on the standard configuration. This servo drive mechanism drives the yarn guide of the wrapping device to rotate at a set speed and angle, achieving uniform wrapping and winding of the yarn.

[0050] The application layer provides a configuration interface for coating process parameters, including setting the number of coating rings, the coating angle range, and the coating speed curve. The real-time display interface adds real-time display of the coating angle and number of rings. Historical data includes a coating quality analysis function, which can statistically analyze the coating pass rate and anomaly types by batch.

[0051] This embodiment expands the application scope of the system by adding a dedicated detection and control module for coating, making the system suitable for the special process requirements of the coating process.

[0052] The specific embodiments of the present invention have been described in detail above with reference to preferred embodiments. It should be understood that these descriptions are merely preferred examples of specific embodiments of the present invention, and not exhaustive. Equivalent substitutions, partial improvements, or combinations made based on the technical solutions of the present invention should all be considered within the scope of protection of the present invention.

Claims

1. A yarn online detection and tension closed-loop control system, characterized in that, It comprises a sensing layer, a control layer, an execution layer, and an application layer. The sensing layer uses a low-friction yarn tension wheel in conjunction with a high-precision strain sensor to collect the dynamic tension value of the yarn in real time, and simultaneously acquires the yarn diameter, fuzziness, and defect characteristics through a photoelectric or laser non-contact measurement unit. The control layer uses a main controller to run an adaptive PID algorithm and a feedforward compensation algorithm to achieve a tension-speed-position cascade closed loop. The execution layer drives the servo or stepper adjustment wheel to move according to control commands, or controls the electromagnetic brake and frequency conversion mechanism to quickly correct tension deviations. The application layer provides a human-machine interface to realize parameter setting, real-time curve display, alarm recording, and data uploading functions.

2. The system as described in claim 1, characterized in that, The sensing layer includes strain gauge tension sensors, fiber optic tension sensors, non-contact photoelectric diameter gauges, laser feather detectors, encoders, and roll diameter monitoring units. Each sensor transmits data to the control layer through a unified interface.

3. The system as described in claim 1, characterized in that, The main controller of the control layer adopts DSP or PLC, runs an adaptive PID control algorithm, and realizes one-click switching of yarn characteristics through online self-learning of yarn characteristics.

4. The system as described in claim 3, characterized in that, The adaptive PID algorithm automatically adjusts the proportional coefficient, integral coefficient, and derivative coefficient based on historical operating data to adapt to changes in different yarn materials, counts, machine speeds, and roll diameters.

5. The system as described in claim 1, characterized in that, The feedforward compensation algorithm calculates the compensation amount in advance based on the vehicle speed change rate and the acceleration roll diameter change rate, and outputs it after superimposing it with the feedback control amount, effectively suppressing overshoot and oscillation caused by sudden changes in operating conditions.

6. The system as described in claim 1, characterized in that, The execution layer includes a servo-driven active tension adjustment wheel, an electromagnetic brake, an electromagnetic clutch, and a frequency conversion yarn feeding mechanism, which can respond to changes in tension deviation within a predetermined time.

7. The system as described in claim 1, characterized in that, The multi-parameter fusion detection enables integrated online monitoring of tension, diameter, fuzz, and defects, with synchronous acquisition frequency reaching the predetermined acquisition frequency, and tension fluctuations controlled within the allowable range.

8. The system as described in claim 1, characterized in that, The application layer supports MODBUS, Profinet, and EtherNet / IP digital interfaces, and can be connected to MES systems and cloud platforms to realize full-process data storage, historical query, report export, and remote operation and maintenance.

9. The system as described in claim 1, characterized in that, It also includes a yarn breakage and defect linkage protection function, including: real-time detection of yarn breakage signals and defect over-limit signals, triggering audible and visual alarms and immediately stopping the machine, while recording the fault time and location information for source tracing analysis.

10. The system according to any one of claims 1-9, characterized in that, A distributed architecture is adopted to realize independent detection and coordinated control of tension in multiple spindles. Each spindle is equipped with an independent sensing unit and slave controller, and the master controller is responsible for task scheduling and global data aggregation.

Images