Safety net warp knitting machine

Abstract

The invention is suitable for the field of safety net production, and provides a safety net warp knitting machine which comprises a warp knitting machine body, a body integrated controller, a tension detection unit, a needle bed execution unit, a knit defect monitoring unit and an execution driving unit. The core is that an intelligent tension dynamic adjusting module, a mesh specification parameterization adaptation module and a multi-mode knit defect early warning and self-repairing module are arranged in the controller; according to the invention, the intelligent tension dynamic adjustment module corrects control parameters in real time through a self-adaptive PID algorithm, and tension fluctuation is controlled within + / -0.5 N; the mesh specification parameterization adaptation module and a needle bed are subjected to closed-loop positioning through a pre-stored parameter library, and the specification switching time is smaller than or equal to 5 minutes; according to the multi-mode knit fault early warning and self-repairing module, through image + laser sensor multi-source fusion, the knit fault detection rate reaches 100%, and the slight knit fault self-repairing success rate is larger than or equal to 90%. The three-level module cooperates with hardware, safety net weaving precision, production flexibility and intelligent level are remarkably improved, and the system is suitable for efficient batch production of safety nets of multiple specifications.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of safety net production, and particularly relates to a safety net warp knitting machine. BACKGROUND

[0002] In actual industrial production, the existing safety net warp knitting machine has the following three key problems due to design defects, which seriously restrict the product quality stability and production efficiency:

[0003] The tension control algorithm is static and cannot adapt to complex working condition fluctuations: the existing equipment generally uses a "fixed parameter PID algorithm", and the core parameters (Kp, Ki, Kd) are pre-set at the factory and cannot be dynamically adjusted during production. 、 、

[0004] However, the warp tension is dynamically affected by multiple factors:

[0005] Warp material differences: the elastic modulus of polyester and nylon differs by more than 20%, and the stretching deformation is different under the same tension;

[0006] Changes in environmental temperature and humidity: for every 10% increase in workshop humidity, the friction coefficient between the warp and the guide component increases by 5-8%, resulting in increased tension loss;

[0007] Warp wear: after 1000 meters of production, the warp diameter decreases by 0.1-0.2mm due to friction, and the tension bearing capacity decreases by 10-15%. Fixed parameters cannot match these changes, resulting in a tension fluctuation amplitude generally exceeding ±3N, a mesh size deviation exceeding 0.8mm, which does not meet the requirement of "first-class product mesh size deviation ≤0.3mm" in GB5725-2022 "Safety Net", and a safety net impact strength decrease of more than 15%;

[0008] Mesh size switching relies on manual operation, and the parameter matching accuracy is low and time-consuming: safety nets need to be produced in different mesh sizes according to application scenarios, and each size corresponds to a specific "needle spacing-sending speed-weft density" parameter combination, and there is a strict matching relationship between the parameters; the existing equipment needs to be manually completed:

[0009] Parameter calculation: according to the mesh size, the needle spacing is deduced, and the sending speed is calculated according to experience, which takes more than 30 minutes;

[0010] Manual adjustment: use a wrench to disassemble the needle bed bolt to adjust the position, and use an oscilloscope to calibrate the sending motor speed, with an error of ±0.5mm (needle spacing) and ±0.1m / min (sending speed);

[0011] ​Trial weaving verification: after adjustment, 3-5 meters of net cloth need to be trial woven, and after detection of qualified specifications, batch production can be carried out. The total time of single switching is more than 60 minutes, which cannot meet the order demand of multiple varieties and small batches;

[0012] Single detection means of weaving defects, missing decision and self-repairing ability: the existing equipment depends on "single-point contact yarn sensor" for detecting weaving defects, which can only detect the broken yarn fault at the installation position of the sensor, and has significant limitations:

[0013] Narrow detection range: unable to cover the whole net surface, high missing detection rate of non-broken yarn defects such as eyelet deformation and local relaxation;

[0014] Weak decision-making ability: after detecting the broken yarn, only sound and light alarms can be triggered, and the fault position needs to be checked along the warp path after manual shutdown;

[0015] No self-repairing function: even slight local tension abnormalities (which can be corrected by fine-tuning tension) need to be handled after shutdown, resulting in poor production continuity and high scrap rate;

[0016] Therefore, a safety net warp knitting machine is needed to solve the above problems. SUMMARY

[0017] The purpose of the embodiment of the present application is to provide a safety net warp knitting machine to solve the problems raised in the background art.

[0018] To achieve the above purpose, the present application provides the following technical solutions:

[0019] A safety net warp knitting machine, comprising a warp knitting machine main body, the warp knitting machine main body is provided with a controller, a tension detection unit, a needle bed execution unit, a weaving defect monitoring unit and an execution driving unit;

[0020] The controller is built-in with three-stage progressive software modules, which are intelligent tension dynamic adjustment module, mesh specification parameterization adaptation module and multi-modal weaving defect early warning and self-repairing module;

[0021] The intelligent tension dynamic adjustment module is signal connected with the tension detection unit and the execution driving unit, for receiving the warp tension signal and outputting the tension adjustment instruction;

[0022] The mesh specification parameterization adaptation module is bidirectional data interactive with the needle bed execution unit and the intelligent tension dynamic adjustment module, pre-stores a plurality of mesh specification associated parameter libraries and can call the adaptive parameters;

[0023] The multi-modal weaving defect early warning and self-repairing module is linked with the weaving defect monitoring unit, the execution driving unit and the previous two software modules, for receiving the weaving defect detection signal and executing the early warning or self-repairing instruction;

[0024] The three-level software modules achieve real-time data interaction through the controller's internal data bus, forming a complete control system from basic control to intelligent decision-making.

[0025] Through a progressive design of three modules, the shortcomings of existing equipment's "independent control of each link" are overcome. The intelligent tension dynamic adjustment module provides a stable warp yarn foundation for weaving, the mesh specification parameterization adaptation module achieves precise switching based on stable tension, and the multimodal weaving defect early warning and self-repair module relies on the parameter benchmarks of the former two to achieve accurate detection. The three form a closed loop of "basic control → flexible production → quality assurance", reducing manual intervention by more than 90% and significantly improving the automation level of the equipment.

[0026] A further technical solution is that the intelligent tension dynamic adjustment module integrates an adaptive PID algorithm submodule, which can adjust the tension based on the deviation between the actual warp tension and the target tension. Real-time correction of PID control parameters (proportional coefficient) Integral Time Differential time The corrected formula is as follows: ,

[0027] The definitions and physical meanings of the variables are as follows:

[0028] Kp: Proportional coefficient before correction (unitless), a core parameter in PID control for responding to tension deviation, determining the initial adjustment force (e.g., initial adjustment of polyester warp yarn). );

[0029] : Corrected proportionality coefficient (unitless), depending on Dynamic adjustment to enhance or weaken the proportion adjustment effect (e.g.) (Increases when positive, decreases when negative);

[0030] Ti: Integral time before correction (unit: s), used to eliminate steady-state error of the system. The longer the time, the weaker the integral effect (e.g., initial time). );

[0031] Corrected integration time (in seconds), as Reverse adjustments can accelerate error elimination when shortening the length and prevent system oscillation when extending it.

[0032] Td: Differential time before correction (unit: s), used to predict the trend of deviation change and suppress overshoot (such as initial time). );

[0033] The corrected differential time (in seconds) varies with... Reverse adjustment shortens the forecasting time, reducing its predictive effect and avoiding overshoot;

[0034] : Tension deviation (unit: N), calculated by the formula Actual tension value - Target tension value , positive value indicating that the actual tension is greater than the target value, and negative value indicating that the actual tension is less than the target value;

[0035] : Algorithm inherent correction coefficient (unitless, all constants greater than 0), used to limit the parameter correction amplitude and avoid system shock (such as polyester warp , , , nylon warp , , );

[0036] Compared with the traditional fixed parameter PID, this algorithm can match the working condition changes in real time. When (tension deviation) is large, (increase proportional adjustment, quickly reduce tension), (shorten integration, accelerate error elimination), (weaken the differential, avoid overshoot);

[0037] When (tension deviation) is small, the parameters are corrected in the opposite direction to ensure that the tension rises smoothly. The final tension fluctuation can be controlled within ±0.5N, and the mesh size deviation is ≤0.2mm, fully meeting the requirements of GB5725-2022 first-class products, and without the need for manual parameter adjustment, suitable for polyester, nylon and other warp materials.

[0038] Further technical solutions, the parameter library of the mesh specification parameterization adaptation module contains multiple sets of associated parameters of mesh specifications, at least covering needle spacing (unit: mm), yarn feeding speed (unit: m / min), weft density (unit: root / 10cm) and corresponding target tension value (unit: N), and supporting manual entry of custom specification parameters through human-computer interaction devices (such as touch screens). When entering, automatically check the parameter rationality (such as needle spacing ≤20mm, yarn feeding speed ≥0.3m / min), and after passing the verification, automatically store and associate with the existing parameter library for subsequent calling;

[0039] The parameter library pre-stores the correlation between "needle spacing - yarn feed speed - weft density - target tension" (as shown in Table 1), eliminating the need for manual calculation and trial knitting. When calling a 10cm×10cm specification, the module automatically matches "needle spacing 10mm, yarn feed speed 0.8m / min, weft density 18 threads / 10cm, target tension 22N". The parameter matching error is ≤0.02mm (needle spacing) and ≤0.01m / min (yarn feed speed), which is far superior to the accuracy of manual adjustment. The custom input function supports the production of special specifications. After input, the parameters are automatically stored, and the next call only takes 1 second. The total time for specification switching is reduced from 60 minutes to less than 5 minutes, and the equipment flexibility is improved by 12 times.

[0040] A further technical solution is that the multimodal fabric defect early warning and self-repair module integrates an image recognition submodule and a signal fusion submodule; the image recognition submodule uses the HOG (Histogram of Oriented Gradients) feature extraction algorithm + SVM (Support Vector Machine) classifier to extract fabric defect features from the mesh image, resulting in high fabric defect recognition accuracy;

[0041] The signal fusion submodule employs DS evidence theory to fuse and determine the image recognition results (global fabric defect information, confidence level 0.8-1.0) with the sensor signals from the fabric defect monitoring unit (local yarn condition, confidence level 0.7-0.9), outputting the fabric defect type (broken yarn, mesh deformity, local tension abnormality, etc.) and severity level (slight, moderate, severe).

[0042] The image recognition submodule extracts the gradient direction features of the mesh surface using the HOG algorithm. The SVM classifier is trained based on 10,000+ images of fabric defects and can identify three main types of fabric defects with high accuracy.

[0043] The signal fusion submodule addresses single detection defects—for example, if the image identifies "mesh shrinkage," and the laser sensor detects "vibration amplitude exceeding 0.5mm," it can be determined as "local tension abnormality" (minor weaving defect) rather than yarn breakage, thus avoiding misjudgment.

[0044] If the image identifies a "blank area" and the sensor detects that "the warp yarn is not present", it is judged as "broken yarn" (serious weaving defect); the rate of missed detection of weaving defects is reduced, providing an accurate basis for graded processing.

[0045] A further technical solution includes a tension detection unit comprising at least two tension sensors arranged at intervals (300-500mm) along the warp yarn transport path, which respectively collect tension signals of the warp yarn from the output end of the warp beam and the input end of the weaving area (denoted as...). , This forms a "dual-point monitoring" system, providing tension data for the entire warp yarn feeding process to the intelligent tension dynamic adjustment module, thus avoiding the limitation of single-point monitoring in capturing tension loss.

[0046] Dual-point monitoring can calculate tension loss. (like , Then it will be damaged. ), if the loss The system detects wear on the yarn guide components and sends a maintenance reminder to the touchscreen to prevent tension loss due to component wear (traditional single-point monitoring cannot detect this problem and easily leads to batches of defective products); simultaneously, the data from two points are averaged. Using this as the actual tension value reduces the error of a single sensor and improves the accuracy of tension detection.

[0047] A further technical solution includes a needle bed execution unit comprising an electric positioning component and a high-precision position feedback element. The electric positioning component uses a servo motor to drive a ball screw, achieving linear movement of the needle bed. The position feedback element uses a grating ruler to collect the actual position signal of the needle bed in real time and feed it back to the mesh specification parameterization adaptation module, forming a positioning closed-loop control with a positioning accuracy of ≤±0.02mm, ensuring precise matching between the needle bed position and the needle spacing parameter.

[0048] The servo motor and ball screw drive system achieves a positioning accuracy of ±0.02mm, far superior to the ±0.5mm accuracy of manual adjustment. The grating ruler feedback frequency is 200kHz, which can capture minute deviations in needle bed movement in real time. The module uses a position loop PID algorithm to fine-tune the motor speed to avoid overshoot or lag. For example, when switching to a 10cm specification, the needle bed needs to move 150mm, and the actual movement distance is 150.018mm, with a deviation of only 0.018mm. The needles and warp yarns are precisely aligned, and there is no problem with uneven mesh size.

[0049] A further technical solution includes an industrial image acquisition device, a non-contact sensor, and a supplementary lighting device. The industrial image acquisition device uses an area array camera with the lens facing the mesh surface of the weaving area (500mm away from the mesh surface), and works in conjunction with an LED supplementary lighting device (wavelength 5500K) to acquire a clear image of the mesh surface. The non-contact sensor uses at least two laser sensors (such as Keyence LV-H62, response time 10μs), evenly distributed along the warp yarn transport path (200mm spacing), to detect the presence and vibration amplitude of individual warp yarns in real time (vibration exceeding 0.5mm is considered abnormal).

[0050] A 20-megapixel camera can capture details of the mesh surface, and a supplementary lighting device avoids image blurring caused by uneven lighting in the workshop; the laser sensor has a response time of 10μs, which can instantly detect broken yarns (traditional contact sensors have a response time of 1ms, a 100-fold lag), and the non-contact design avoids wear on the warp yarns; the combination of the two achieves dual coverage of "global mesh surface monitoring + local yarn monitoring", with no blind spots, ensuring that even minor weaving defects can be detected.

[0051] A further technical solution includes a tension adjusting roller and a second servo motor. The tension adjusting roller is made of 45# steel as the base material, with a roller diameter of 80-100mm, covered with a 3-5mm thick silicone layer, and the surface is processed with diamond-shaped anti-slip texture. The second servo motor is a 1.5-2kW permanent magnet synchronous servo motor, equipped with a high-precision servo driver. The second servo motor is coaxially connected to the tension adjusting roller through a flexible coupling, with a coaxiality error ≤0.02mm. The execution drive unit receives PWM control commands from the controller's built-in intelligent tension dynamic adjustment module, and drives the tension adjusting roller to rotate by adjusting the speed of the second servo motor, thereby changing the warp yarn winding angle and conveying speed, and realizing real-time dynamic adjustment of the warp yarn tension.

[0052] Higher tension adjustment precision: The silicone layer of the tension adjustment roller can increase the friction with the warp yarn, avoiding tension loss due to warp yarn slippage, and the diamond anti-slip texture further improves the stability of friction; with the high-precision servo motor, the speed control accuracy can reach ±1rpm, which can quickly respond to the tension adjustment command and keep the tension fluctuation within ±0.5N, meeting the first-grade product requirement of mesh size deviation ≤0.2mm;

[0053] The equipment operates more stably: the tension regulating roller made of 45# steel has high strength and small deformation, and is not prone to bending after long-term use; the flexible coupling can buffer the impact force when the motor starts and brakes, reduce mechanical wear on the roller and motor, reduce the frequency of equipment failure, and extend service life.

[0054] Adaptable to various warp materials: The flexible properties of the silicone layer can adapt to warp yarns with different elastic moduli such as polyester and nylon, avoiding warp wear caused by hard contact; The wide speed range (500-3000rpm) of the servo motor can match different yarn feeding speed requirements, and can realize the weaving of warp yarns of multiple materials and specifications without changing the drive components, thus improving the versatility of the equipment.

[0055] A further technical solution involves using an industrial-grade PLC that supports independent online upgrades of three software modules. During the upgrade process, the controller employs a dual-program area design (the main program area runs the current production module, while the backup program area writes the new module program). After the program is written, seamless switching occurs (switching time ≤ 1 second), ensuring uninterrupted production flow and retaining historical production data and parameter libraries after the upgrade.

[0056] The independent upgrade function allows for the optimization of a single module, reducing upgrade risks. The dual-program area design avoids the downtime issues associated with traditional upgrades—for example, when upgrading the multimodal fabric defect warning and self-repair module, the main program area continues to run tension adjustment and specification adaptation, while the backup program area simultaneously writes the new algorithm (approximately 30 seconds), taking effect immediately after the switch. Production interruption time is ≤1 second, reducing annual downtime losses by over 50,000 yuan. Parameter libraries and historical data are not lost after the upgrade, eliminating the need for re-debugging and ensuring production continuity.

[0057] Compared with the prior art, the beneficial effects of the present invention are:

[0058] This invention solves the problem of poor adaptability in tension control and improves mesh uniformity and product strength: The intelligent tension dynamic adjustment module, through adaptive PID algorithm and dual-point monitoring, reduces tension fluctuation from the traditional ±3N to within ±0.5N, and mesh size deviation from 0.8mm to below 0.2mm, fully meeting the requirements of GB5725-2022 Grade 1 product; uniform tension ensures consistent stress on the warp yarns, improves the impact resistance of the safety net, avoids mesh tearing caused by excessive local tension, and improves product qualification rate;

[0059] This invention solves the problem of low efficiency in specification switching, improving production flexibility and equipment utilization: The mesh specification parameterized adaptation module, through parameter library and closed-loop positioning of the needle bed, reduces the total time for specification switching from 60 minutes to less than 5 minutes, thus improving switching efficiency; automatic parameter adaptation avoids human error, improves the first meter of mesh fabric qualification rate, and eliminates the need for trial knitting waste; it supports more than 5 conventional specifications and custom specifications, significantly improving equipment flexibility, enabling rapid response to small batch orders, increasing equipment utilization, and reducing labor costs.

[0060] This invention addresses the problem of insufficient defect handling capabilities, reducing scrap rates and production costs: the multimodal defect early warning and self-repair module uses "image + laser" multi-source fusion to reduce the defect miss rate and can identify more than 8 types of defects; the self-repair success rate of minor defects (such as local tension abnormalities) is high, reducing downtime; the response time for serious defects is short, reducing the scrap rate.

[0061] To more clearly illustrate the structural features and effects of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0062] Fig. 1 This is a diagram of the overall system architecture of the present invention;

[0063] Fig. 2 This is a parameter call diagram for the mesh specification parameterization adaptation module of the present invention.

[0064] In the diagram: 1. Warp knitting machine body; 2. Controller; 3. Three-level software module; 31. Intelligent tension dynamic adjustment module; 32. Mesh specification parameterization adaptation module; 33. Multimodal fabric defect early warning and self-repair module; 4. Tension detection unit; 41. Warp beam end tension sensor; 42. Knitting zone end tension sensor; 5. Needle bed execution unit; 51. Servo motor one; 52. Ball screw; 53. Grating ruler; 6. Fabric defect monitoring unit; 61. Area array camera; 62. Complementary lighting device; 63. Laser sensor; 7. Execution drive unit; 71. Tension adjustment roller; 72. Servo motor two; 8. Human-machine interaction device. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0066] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0067] Example 1

[0068] like Figs. 1-2 As shown, this embodiment of the invention provides a safety net warp knitting machine, including a warp knitting machine body 1. The frame of the warp knitting machine body 1 is welded from Q235 steel plate (12mm thick) to ensure operational stability.

[0069] The main body 1 is equipped with a controller 2, a tension detection unit 4, and an execution drive unit 7. Each unit is connected to the main body terminal block via an aviation plug for easy maintenance.

[0070] Controller 2: Uses Siemens S7-1500 PLC (model 1511C-1PN), with built-in dual program areas (main program area 16MB, spare program area 16MB), and supports Ethernet communication (baud rate 100Mbps).

[0071] The built-in intelligent tension dynamic adjustment module 31 is implemented through the PLC's OB35 interrupt program (interrupt cycle 5ms) to ensure the real-time performance of tension adjustment.

[0072] The adaptive PID algorithm of the intelligent tension dynamic adjustment module 31 is written in ST language and stored in the FC1 function block. , , Preset in DB1 data block;

[0073] Tension detection unit 4: includes two LSZ-A type tension sensors (range 0-50N, accuracy 0.1%FS), which are mounted on the warp beam output end 41 and the knitting area input end 42 via brackets, with a spacing of 500mm;

[0074] The sensor communicates with controller 2 via a 485 bus (Modbus-RTU protocol), uploading data every 5ms. , data;

[0075] The drive unit 7 includes a tension adjusting roller 71 (45# steel, 80mm in diameter, covered with a 3mm silicone layer, with diamond anti-slip texture) and a 1.5kW servo motor 72 (Yaskawa SGMAH-01AAA41, equipped with SGDV-1R6A01A driver).

[0076] The motor is connected to the adjusting roller via a flexible coupling (LK45), with a coaxiality error ≤0.02mm;

[0077] This embodiment is used to produce a 5cm×5cm safety net made of 1600D polyester warp yarn. The specific operation process is as follows:

[0078] Parameter initialization: Set the target tension via the touchscreen. PID initial parameters , , After the equipment is started, the warp beam releases the warp yarns at a rate of 0.6 m / min;

[0079] Normal tension adjustment: Sensor 41 collects data. Sensor 42 collects data. The intelligent tension dynamic adjustment module 31 calculates... , Maintaining the original parameters, motor 72 rotates at 1500 rpm, and adjusting roller 71 steadily feeds the warp yarn;

[0080] Adjustment of operating conditions: When the humidity in the workshop increases from 50% to 70%, the coefficient of friction of the warp yarns increases. Reduced to 17.2N, Rising to 18.3N, , ;

[0081] The intelligent tension dynamic adjustment module 31 corrects the parameters according to the formula: After correction, the output PWM signal (duty cycle reduced from 50% to 48%), the motor speed decreased to 1470 rpm, the regulating roller speed decreased, the warp yarn winding angle increased, and the tension rebounded; after 0.6 seconds, , The parameters have stopped being corrected.

[0082] In this embodiment, the tension fluctuation is stable within ±0.3N, the actual mesh size is 5.0±0.15mm, which meets the requirements of GB5725-2022 Grade I product; after continuous production of 100 meters, the mesh uniformity is 99.9%, the warp breaking strength retention rate is 98%, and there are no quality problems caused by tension fluctuation.

[0083] Example 2

[0084] The difference between this embodiment and Embodiment 1 is that the warp knitting machine body 1 is equipped with a needle bed execution unit 5 and a human-machine interaction device 8, and the controller 2 is equipped with a mesh specification parameterization adaptation module 32, which supports independent online upgrades. Specific details are as follows:

[0085] The needle bed execution unit 5 includes a Yaskawa SGMAH-02AAA41 servo motor 51, a THKBF1510 ball screw 52 (lead 10mm), and a Keyence LC-2450 grating ruler 53;

[0086] The motor is connected to the lead screw via a coupling, and the lead screw nut is rigidly connected to the needle bed mounting base (bolt torque 8 N·m).

[0087] The grating ruler reading head is fixed to the frame, and the main body is fixed to the needle bed, providing real-time position feedback. Mesh specification parameterization adapter module 32:

[0088] The parameter library is stored in DB2 data blocks and contains 5 sets of common parameters in Table 1. It also supports manual input of custom parameters via touchscreen.

[0089] The module uses the FC2 function block to call and validate parameters. If the validation fails, the touch screen will display a "Parameter exceeded" message.

[0090] Human-computer interaction device 8: Weintek MT8102iE touch screen, with an interface including "Parameter Settings", "Status Monitoring" and "Fault Alarm" pages;

[0091] The "Status Monitoring" page displays F, needle bed position, and yarn feed speed in real time, with a refresh rate of 1Hz;

[0092] Online upgrade function: The optimized parameter adaptation algorithm (new version of FC2) is imported into the PLC's spare program area via USB flash drive. After the import is completed, the PLC executes the "STOP→Switch program area→START" instruction, with a switching time of 0.8 seconds and no production interruption.

[0093] This example is used for switching between 5cm×5cm and 10cm×10cm sizes, and the process is as follows:

[0094] Switching trigger: Select "10cm×10cm" on the "Parameter Settings" page of the touch screen, click "Confirm Switching", and the mesh specification parameterization adaptation module 32 sends a speed reduction command to the yarn feeding motor, reducing the speed from 0.6m / min to 0.3m / min;

[0095] Needle bed positioning: The mesh specification parameterization adaptation module 32 calls the parameter "needle spacing 10mm", calculates that the needle bed needs to move 150mm (initial position 50mm → target position 200mm), and sends a positioning command to the motor 51;

[0096] The motor drives the lead screw to rotate, and the grating ruler provides feedback on the position (e.g., 50mm→60mm→…→199.982mm).

[0097] When the deviation is ≤0.02mm, the motor brake engages, and the positioning time is 2.3 seconds.

[0098] Parameter adaptation: The mesh specification parameterization adaptation module 32 pushes parameters to the tension intelligent tension dynamic adjustment module 31. Intelligent tension dynamic adjustment module 31 updated The calculation baseline is set; at the same time, the mesh specification parameterization adaptation module 32 sends an instruction to the weft insertion mechanism to reduce the density from 25 threads / 10cm to 18 threads / 10cm, and sends an instruction to the yarn feeding motor to increase the speed to 0.8m / min; after 1.5 seconds, the switching is completed.

[0099] In this embodiment, the total switching time is 4.8 minutes, the first meter of mesh fabric has a specification of 10.0±0.2mm, and the pass rate is 100%; after custom input of the 8cm specification, the next call only takes 1 second, and the equipment flexibility is significantly improved.

[0100] Example 3

[0101] The difference between this embodiment and embodiment 2 is that the warp knitting machine body 1 is equipped with a fabric defect monitoring unit 6, and the controller 2 is equipped with a multimodal fabric defect early warning and self-repair module 33, as detailed below:

[0102] Defect Monitoring Unit 6: Hikvision MV-CE200-10GM area array camera 61 (20 million pixels, 30fps), Computar M1614-MP2 lens (focal length 16mm), mounted 500mm above the knitting area;

[0103] Four LED supplementary lighting units (62 LEDs, 5500K, 10W) ​​are symmetrically arranged;

[0104] Four Keyence LV-H62 laser sensors 63, with a warp spacing of 200mm and a detection end distance of 0.8mm from the warp;

[0105] Multimodal fabric defect early warning and self-repair module 33: Image recognition submodule (HOG+SVM) is written in C language and stored in FC3; Signal fusion submodule (DS evidence theory) is stored in FC4;

[0106] The warning device consists of a TL-70 audible and visual alarm (90dB, flashing red light) and a touchscreen pop-up window;

[0107] This example is used for production of 10cm×10cm specifications. The defect handling scenario is as follows:

[0108] Scenario 1: Minor weaving defect (localized tension abnormality) Weaving defect occurs: The friction between the 7th warp yarn and the guide needle increases, the tension rises from 22N to 25N, and the mesh shrinks to 8mm;

[0109] Detection and Judgment: Sensor 63 (the second one) detected a vibration amplitude of 0.8mm (exceeding the threshold of 0.5mm) and sent an "abnormal" signal;

[0110] Camera 61 captures images, and the submodule extracts HOG features to determine "mesh deformity";

[0111] The fusion submodule determines "minor fabric defect (localized tension abnormality)";

[0112] Self-repair: The multimodal fabric defect early warning and self-repair module 33 sends a temporary command to the tension intelligent tension dynamic adjustment module 31. Reduced to 20N; Intelligent tension dynamic adjustment module 31 calculates Correct parameters: ,

[0113] After 0.5 seconds, the tension dropped to 20.5N, the mesh size recovered by 10mm, and the self-repair was successful;

[0114] Scenario 2: Severe Weaving Defect (Broken Yarn) The defect occurs when the third warp yarn breaks, resulting in a 20mm missing yarn section.

[0115] Detection and Judgment: Sensor 63 (first one) cannot detect warp yarns and sends a "broken yarn" signal;

[0116] When camera 61 captures a blank area, the submodule determines that the yarn is "broken".

[0117] The fusion submodule determines "serious weaving defects (broken yarn)";

[0118] Warning and shutdown: The multimodal fabric defect warning and self-repair module 33 triggers the alarm, and a pop-up window displays "Broken yarn, X=180mm, Y=150mm";

[0119] The yarn feeding, needle bed, and take-up motor stop within 0.2 seconds, with a waste fabric length of 2cm; production resumes after manual yarn splicing.

[0120] In this embodiment, the defective rate is 0.5%, the scrap rate is 2.8%, and the production continuity is improved by 40%.

[0121] Working principle and usage process of this invention:

[0122] Centered on a three-tiered, progressive software module system, with hardware units working collaboratively, it covers the entire process of "initialization - tension adjustment - specification switching - defect handling - finished product output." The specific logic is as follows:

[0123] I. System initialization phase (power-on preparation, ensuring equipment readiness);

[0124] Hardware self-test: After power-on, controller 2 performs self-tests on each unit: the tension sensor collects the no-load tension (should be ≤1N), and an alarm is triggered if the threshold is exceeded; the needle bed moves to the mechanical origin (the grating ruler detects the origin signal), and an alarm is triggered if the timeout is exceeded; the camera captures a blank image, and the sensor detects no warp yarn, and an alarm is triggered if any abnormality occurs;

[0125] Module Loading and Parameter Preset: After self-testing, load the following three modules: Intelligent Tension Dynamic Adjustment Module 31 reads the warp material (manual input / scanning) and initializes PID parameters; Specification Mesh Specification Parameterized Adaptation Module 32 defaults to the last specification and drives needle bed positioning; Multimodal Weaving Defect Early Warning and Self-Repair Module 33 loads the SVM model and sets the sensor threshold (vibration 0.5mm).

[0126] Human-machine confirmation: The touchscreen displays "Initialization complete". The operator confirms that the warp yarn is threaded correctly (warp beam → sensor 41 → adjusting roller 71 → sensor 42 → knitting needle), and clicks "Start". The equipment then enters production mode.

[0127] II. Data Acquisition and Filtering during the Dynamic Tension Adjustment Stage (Core Control, Ensuring Warp Stability):

[0128] Sensors 41 and 42 collect data every 5ms. , The intelligent tension dynamic adjustment module 31 performs data moving average filtering (window 5) to remove noise (such as instantaneous peak values);

[0129] Average tension and deviation calculation: The intelligent tension dynamic adjustment module 31 calculates the average tension according to the formula: Recalculate the deviation: ;

[0130] PID parameter correction and execution:

[0131] like Maintain the original parameters, and keep the motor speed unchanged at 72.

[0132] like Corrected according to the adaptive formula Generate PWM commands (1kHz frequency) to drive the motor: Then the speed decreases (tension decreases). Then the growth rate (tension increases);

[0133] Tension loss monitoring: Intelligent tension dynamic adjustment module 31 calculates loss , if the loss Send a "yarn guide component maintenance" reminder to prevent tension loss due to wear;

[0134] III. Mesh specification switching stage (flexible production, adaptable to multiple varieties);

[0135] Switching trigger and preprocessing: Select the target specification on the touch screen, and the mesh specification parameterization adaptation module 32 sends a speed reduction command (to 0.3m / min) to the yarn feeding motor to reduce warp yarn stretching;

[0136] Closed-loop positioning of the needle bed: The mesh specification parameterization adaptation module 32 calls the needle spacing, calculates the moving distance, and sends a command to the motor 51; the motor drives the lead screw, the grating ruler provides feedback on the position, and the mesh specification parameterization adaptation module 32 finely adjusts the speed through the position loop PID, locking when the deviation is ≤0.02mm;

[0137] Multi-parameter collaborative adaptation: The mesh specification parameterization adaptation module 32 pushes parameters to the tension intelligent tension dynamic adjustment module 31. It sends parameter commands to the weft insertion mechanism and yarn feeding motor; after the speed and density stabilize, it sends a "switching complete" signal.

[0138] Custom parameter storage: After switching to a custom specification, the mesh specification parameterization adaptation module 32 automatically stores the parameters to DB2. The next time it is called, only the name needs to be selected.

[0139] IV. Defect Monitoring and Grading Stage (Quality Assurance, Reduction of Waste);

[0140] Multi-source data acquisition: The camera captures images every 33ms, the sensor detects warp yarns every 10μs, and the data is transmitted to the multimodal fabric defect early warning and self-repair module 33;

[0141] Image recognition and fusion judgment: The sub-module preprocesses images, extracts HOG features, and performs SVM classification; the fusion sub-module uses DS theory to fuse images and sensor signals to determine the type and grade of fabric defects;

[0142] Hierarchical processing:

[0143] Minor fabric defects: Send a temporary signal to the tension intelligent tension dynamic adjustment module 31. Self-repair is achieved by adjusting parameters;

[0144] Severe fabric defect: triggers alarm and stops the machine; restarts after manual intervention.

[0145] Fault Log Recording: The multimodal fabric defect early warning and self-repair module 33 records fabric defect information (time, type, processing result), stores it on an SD card, and supports export and analysis;

[0146] V. Finished Product Output and Data Management Stage (Production Closed Loop, Optimization Decision-Making);

[0147] Finished product pulling and winding: The pulling mechanism (rubber roller) pulls the mesh at a uniform speed (speed synchronized with yarn feeding), and the winding mechanism uses a torque motor to wind the mesh (the torque is automatically adjusted when the roll diameter increases) to ensure consistent winding tightness;

[0148] Data Upload and Analysis: Controller 2 uploads production output, tension curve, and number of weaving defects to the MES system every minute. Managers can view the reports and optimize the process (e.g., if a certain specification has many weaving defects, the parameters are adjusted).

[0149] Module upgrades and maintenance: During production breaks, modules can be upgraded via USB flash drive / remotely, retaining parameters and logs to ensure continuous equipment optimization;

[0150] In summary, this equipment addresses the core shortcomings of existing equipment through progressive control of three-level modules and hardware collaboration, enabling high-precision, high-flexibility, and highly intelligent production of safety nets, and providing a technological upgrade solution for the safety net manufacturing industry.

[0151] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A safety net warp knitting machine, comprising a warp knitting machine body, characterized in that: The warp knitting machine body is equipped with a controller, a tension detection unit, a needle bed execution unit, a knitting defect monitoring unit, and an execution drive unit; The controller has a built-in three-level progressive software module, namely, an intelligent tension dynamic adjustment module, a mesh specification parameterized adaptation module, and a multimodal fabric defect early warning and self-repair module; The intelligent tension dynamic adjustment module is signal-connected to the tension detection unit and the execution drive unit, and is used to receive warp tension signals and output tension adjustment commands. The mesh specification parameterization adaptation module interacts bidirectionally with the needle bed execution unit and the intelligent tension dynamic adjustment module, and pre-stores multiple sets of mesh specification associated parameter libraries and can call adaptation parameters. The multimodal fabric defect early warning and self-repair module is linked with the fabric defect monitoring unit, the execution drive unit and the first two software modules to receive fabric defect detection signals and execute early warning or self-repair instructions; The three-level software modules achieve real-time data interaction through the controller's internal data bus, forming a complete control system from basic control to intelligent decision-making.

2. The safety net warp knitting machine according to claim 1, characterized in that: The intelligent tension dynamic adjustment module integrates an adaptive PID algorithm submodule, which can adjust the tension based on the deviation between the actual warp tension and the target tension. Real-time correction of PID control parameters (proportional coefficient) Integral Time Differential time The formula is as follows: The definitions and physical meanings of the variables are as follows: : The proportional coefficient before correction, the initial adjustment force parameter; : The corrected proportional coefficient, the adjustment strength after dynamic adjustment; The integral time before correction, a parameter used to eliminate steady-state error; Corrected integration time, adjusted error elimination speed; : The derivative time before correction, a parameter to suppress overshoot; Corrected derivative time, adjusted overshoot suppression capability; Tension deviation: The difference between the actual tension and the target tension; The algorithm's inherent correction coefficient limits the magnitude of parameter correction and prevents system oscillations.

3. The safety net warp knitting machine according to claim 1, characterized in that: The parameter library of the mesh specification parameterization adaptation module contains multiple sets of associated parameters for mesh specifications, including at least needle spacing, yarn feed speed, weft insertion density, and target tension value for the corresponding specifications. It also supports manual input of custom specification parameters through human-computer interaction devices. The parameters are automatically verified for rationality during input, and after verification, they are automatically associated with and stored in the existing parameter library for easy subsequent use.

4. The safety net warp knitting machine according to claim 1, characterized in that: The multimodal fabric defect early warning and self-repair module integrates an image recognition submodule and a signal fusion submodule; The image recognition submodule uses the HOG feature extraction algorithm + SVM classifier to extract fabric defect features from the mesh image; The signal fusion submodule uses the DS evidence theory to fuse and determine the image recognition results with the sensor signals of the fabric defect monitoring unit, and outputs the fabric defect type and severity level.

5. The safety net warp knitting machine according to claim 1, characterized in that: The tension detection unit includes at least two tension sensors arranged at intervals along the warp yarn transport path. These sensors collect tension signals from the warp yarn output end and the knitting zone input end, respectively, forming a "dual-point monitoring" system. This provides the intelligent tension dynamic adjustment module with tension data throughout the warp yarn transport process, avoiding the shortcomings of single-point monitoring in capturing tension loss.

6. The safety net warp knitting machine according to claim 1, characterized in that: The needle bed execution unit includes an electric positioning assembly and a high-precision position feedback element; The electric positioning component uses a servo motor to drive a ball screw to achieve linear movement of the needle bed; The position feedback element uses a grating ruler to collect the actual position signal of the needle bed in real time and feed it back to the mesh specification parameterized adaptation module to form a positioning closed-loop control.

7. The safety net warp knitting machine according to claim 1, characterized in that: The fabric defect monitoring unit includes industrial image acquisition equipment, non-contact sensors, and supplementary lighting devices; The industrial image acquisition device uses an area array camera with the lens facing the mesh surface of the woven area, and is equipped with an LED supplementary lighting device to obtain a clear image of the mesh surface. The non-contact sensor uses laser sensors, at least two of which are evenly distributed along the warp yarn conveying path to detect the presence and vibration amplitude of a single warp yarn in real time.

8. The safety net warp knitting machine according to claim 1, characterized in that: The execution drive unit (7) includes a tension adjusting roller (71) and a servo motor (72), which is coaxially connected to the tension adjusting roller (71) via a flexible coupling. The execution drive unit (7) receives the PWM control command of the built-in intelligent tension dynamic adjustment module (31) of the controller (2), and drives the tension adjustment roller (71) to rotate through the speed adjustment of the servo motor (72), thereby changing the warp yarn winding roller angle and conveying speed, and realizing the real-time dynamic adjustment of warp yarn tension.

9. The safety net warp knitting machine according to claim 1, characterized in that: The controller uses an industrial-grade PLC and supports independent online upgrades of three software modules; During the upgrade process, the controller uses a dual-program area design to seamlessly switch operation after the program is written, ensuring that the current production process is not interrupted, and retaining historical production data and parameter library after the upgrade.

Images