Bio-based pulp low-sizing intelligent control system and method thereof

Abstract

The application discloses a kind of bio-based pulp low intelligent control system of sizing and method thereof, it is related to textile sizing automation control technical field, the system is by perception unit, model unit, optimization control unit and execution unit constitute, perception unit, model unit, optimization control unit and execution unit are sequentially connected, form closed loop control loop, the four-layer closed loop control architecture of " perception unit-model unit-optimization control unit-execution unit" is constructed in the application, realizes from data acquisition, quality evaluation, parameter identification, optimization decision to the complete closed loop of execution control, avoids the hysteresis of " after discovering quality problem manually adjusting " in traditional experience sizing, aims to build " sizing formula-sizes state-sizing rate-weaving quality" multi-parameter collaborative control framework, realizes accurate, stable and intelligent control of green sizing material without PVA under the condition of low sizing rate, reduces sizing material consumption and desizing pollution load.

Description

Technical Field

[0001] This invention relates to the field of automated control technology for textile sizing, specifically to a bio-based sizing low-sizing intelligent control system and method. Background Technology

[0002] Sizing is a key step in the weaving preparation process. Sizing agents improve the surface structure of warp yarns through penetration, coating and film formation, thereby enhancing yarn strength, abrasion resistance and friction resistance, and reducing weaving breakage rate. Traditional PVA sizing agents have good film-forming and adhesion properties, but their biodegradability is poor and desizing wastewater is difficult to treat, which is not compatible with the requirements of green manufacturing and clean production. Although ordinary starch is widely available, low in cost and biodegradable, it has problems such as large viscosity fluctuations, insufficient adhesion, brittle and hard sizing film and difficulty in yarn separation. When used alone, it is difficult to meet the production requirements of high-count and high-density, polyester-cotton blended and regenerated cellulose fiber fabrics.

[0003] The main development direction of existing green sizing technology is based on modified starch, using techniques such as grafting, etherification, esterification, crosslinking, degradation, and compounding to improve the molecular structure, and combining it with environmentally friendly polymeric auxiliaries, waxy smoothing agents, and antistatic components to form PVA-free or low-pollution sizing systems suitable for different fibers and weaving equipment. However, existing technologies have the following shortcomings:

[0004] First, there is a lack of systematic control methods for low sizing rate processes. Low sizing rate is not simply about reducing the amount of sizing agent. Rather, it requires ensuring yarn reinforcement, abrasion resistance, hair adhesion, and yarn separation performance. This can be achieved by designing sizing agent molecules, regulating sizing agent rheology, and optimizing equipment parameters to allow more sizing agent to participate in effective film formation, thereby reducing ineffective coating and desizing burden. Existing equipment often relies on manual experience for adjustment, making it difficult to achieve precise control.

[0005] Second, the lack of a multi-parameter collaborative optimization mechanism means that the sizing quality is affected by multiple factors such as sizing formula, sizing state (temperature, viscosity, pH value), equipment parameters (sizing pressure, speed) and quality feedback (sizing rate, warp breakage rate). Traditional single-variable control methods are difficult to adapt to complex working conditions.

[0006] Third, there is a lack of model-driven adaptive control capabilities. In the existing sizing process, once the formula and process parameters are set, they are basically fixed and cannot be adjusted in real time according to the differences between yarn batches, the fluctuation of sizing batches, and environmental changes, resulting in large fluctuations in sizing rate and poor quality stability.

[0007] Fourth, there is a lack of evaluation and control systems that match the requirements of green manufacturing. The evaluation of the application effect of green slurry in existing technologies is mostly at the level of qualitative description, lacking quantifiable comprehensive quality evaluation indicators and green benefit calculation models. Summary of the Invention

[0008] This invention provides a bio-based sizing agent low sizing intelligent control system and its control method. The aim is to achieve precise, stable and intelligent control of PVA-free green sizing agents under low sizing conditions by constructing a multi-parameter collaborative control framework of "sizing agent formulation - sizing liquid state - sizing rate - weaving quality", thereby reducing sizing agent consumption and desizing pollution load.

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A bio-based slurry low-sizing intelligent control system includes:

[0011] The sensing unit is used to collect multi-source process parameters during the sizing production process;

[0012] The model unit is used to perform data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on the multi-source process parameters, and outputs a prediction model.

[0013] An optimization control unit is used to solve for the optimal control quantity based on the prediction model according to the set control objectives and constraints, and generate control commands.

[0014] An execution unit, which is used to execute multi-parameter collaborative optimization control actions according to the control instructions;

[0015] The sensing unit, model unit, optimization control unit, and execution unit are connected in sequence to form a closed-loop control circuit, ultimately achieving multi-parameter collaborative optimization and control of the sizing process.

[0016] As a preferred embodiment of the present invention, the multi-source process parameters include slurry state parameters, equipment operating parameters, and quality feedback parameters, wherein:

[0017] The slurry state parameters include slurry temperature, slurry viscosity, pH value, liquid level, and solid content;

[0018] The equipment operating parameters include sizing pressure, drying room temperature, sizing machine speed, and yarn tension;

[0019] The quality feedback parameters include sizing rate, moisture regain rate, and warp breakage rate.

[0020] As a preferred embodiment of the present invention, the model unit includes a sizing rate dry weight correction model, a sizing viscosity stability evaluation model, a sizing yarn comprehensive quality evaluation model, and a sizing rate prediction and recursive parameter identification model, wherein:

[0021] The dry weight correction model for sizing rate achieves data correction.

[0022] The slurry viscosity stability evaluation model and the sizing yarn comprehensive quality evaluation model achieve quality evaluation;

[0023] The sizing rate prediction and recursive parameter identification model realizes parameter identification and sizing rate prediction.

[0024] As a preferred embodiment of the present invention, the sizing rate dry mass correction model is used to uniformly correct the weighed wet mass to dry mass:

[0025]

[0026] in, For dry quality, The wet mass is obtained by weighing. This is the moisture content (mass fraction) on a wet basis, calculated from the corresponding moisture regain R. ;

[0027]

[0028] in, for Dry weight of warp yarns before sizing This refers to the dry weight of the sized yarn after sizing.

[0029] As a preferred embodiment of the present invention, the slurry viscosity stability evaluation model uses viscosity fluctuation rate... Characterizing the stability of the slurry in continuous production:

[0030]

[0031] in, The initial dynamic viscosity is measured at t=0 after the slurry has completed gelatinization or dissolution and reached the set temperature; The dynamic viscosity is the value measured after t hours of continuous heat preservation from the initial moment. When the concentration is ≤15%, the slurry is deemed to have thermal stability and continuous use stability.

[0032] As a preferred embodiment of the present invention, the sizing comprehensive quality evaluation model uses a comprehensive quality index. Evaluating sizing quality:

[0033]

[0034] in, For sizing breaking strength, For wear resistance cycles, The feather index, The rate of warp breakage per unit length, subscript Indicates the standard process or raw yarn condition. As weight.

[0035] As a preferred embodiment of the present invention, the sizing rate prediction and recursive parameter identification model achieves online identification of the production process through a recursive least squares parameter identification formula with a forgetting factor:

[0036]

[0037] in, Output as sizing rate or overall quality index. This refers to the sizing concentration. The viscosity of the slurry. The grouting pressure, For the temperature of the slurry, The speed of the sizing machine. For moisture regain, For the parameters to be identified, Noise term;

[0038] The recursive least squares parameter identification formula is updated as follows:

[0039]

[0040]

[0041]

[0042] in, The forgetting factor has a value range of 0.95 to 0.99.

[0043] In a preferred embodiment of the present invention, the optimization control unit employs a model predictive control algorithm to solve for the optimal control quantity, with the objective function being:

[0044]

[0045] in, To predict the time domain, To control the time domain, To predict the sizing rate, To achieve the target sizing rate, To control the change in quantity, and This is the weight matrix;

[0046] The control targets include slurry viscosity, slurry pressure, drying room temperature, and sizing machine speed. The constraints include slurry viscosity range, slurry temperature range, upper limit of slurry pressure, upper and lower limits of sizing machine speed, and moisture regain range.

[0047] As a preferred embodiment of the present invention, the system further includes a green benefit calculation model, which is used to estimate the overall benefits of the low sizing rate process. :

[0048]

[0049] in, and These are the sizing rates for the baseline process and the low sizing rate process, respectively (converted from percentages to decimals during calculation). For warp yarn production, For the unit cost of slurry, , This refers to the total energy consumption for the same accounting period and corresponding to the same warp yarn output. For energy unit price, , For the same accounting period and corresponding warp yarn production, the total water consumption for desizing and wastewater treatment is calculated. For the unit cost of water treatment, The cost of adding new additives or intelligent control systems is amortized.

[0050] A method for intelligent control of low sizing in bio-based slurry includes the following steps:

[0051] S1. The sensing unit collects multi-source process parameters related to the sizing state, equipment operation, and sizing quality during the sizing process.

[0052] S2. The model unit performs data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on the collected multi-source process parameters, and outputs a prediction model.

[0053] S3. The optimization control unit calculates the optimal control quantity based on the prediction model according to the set control objectives and constraints, and generates control commands.

[0054] S4. The execution unit coordinates the sizing feed amount, water replenishment amount, steam valve opening, sizing roller pressure, sizing machine speed and drying room temperature according to the control instructions and the optimal control amount, and performs control actions according to the preset upper and lower limits, single-cycle change limit and safety interlock conditions.

[0055] S5. Repeat steps S1 to S4 to form a closed-loop control, thereby achieving multi-parameter collaborative optimization and control of the sizing process.

[0056] This invention provides a smart control system for low sizing of bio-based slurries. It has the following beneficial effects:

[0057] 1. This invention constructs a four-layer closed-loop control architecture consisting of a "perception unit, model unit, optimization control unit, and execution unit," realizing a complete closed loop from data acquisition, quality evaluation, parameter identification, optimization decision-making to execution control, thus avoiding the lag of "manual adjustment after discovering quality problems" in traditional experience-based pulping.

[0058] 2. This invention proposes a dry mass correction model for sizing rate. By uniformly correcting the wet mass to dry mass, the influence of moisture content and moisture regain on the weighing results is eliminated, thereby improving the accuracy of sizing rate detection.

[0059] 3. This invention proposes a slurry viscosity stability evaluation model and a sizing yarn comprehensive quality evaluation model, which realizes the quantitative evaluation of slurry thermal stability and continuous use stability, as well as the comprehensive evaluation of sizing yarn quality, and provides a clear objective function for intelligent control;

[0060] 4. This invention adopts a recursive least squares parameter identification model with a forgetting factor, which can identify system parameters online and automatically adapt to model parameter drift caused by changes in yarn batches, sizing batches and equipment status, thereby improving the adaptive capability of the control system.

[0061] 5. This invention employs a model predictive control algorithm to solve for the optimal control quantity based on the sizing rate prediction model. This allows for the prediction of sizing rate deviations in advance and the implementation of compensatory control, effectively suppressing sizing rate fluctuations.

[0062] 6. This invention proposes anomaly identification indicators and a conservative control mode, which avoids abnormal sensor data from directly entering the controller and causing malfunctions, thereby improving the safety and reliability of the system.

[0063] 7. This invention proposes a green benefit calculation model, which can quantitatively evaluate the comprehensive benefits of low sizing rate processes in terms of sizing consumption, energy consumption, water consumption, and desizing treatment costs, providing data support for enterprise process decisions.

[0064] 8. The intelligent control system and control method for low sizing rate of bio-based sizing agents provided by this invention can be widely applied to the green and intelligent transformation of textile sizing processes. By constructing a multi-parameter collaborative control framework and a model-driven adaptive control mechanism, the system achieves precise and stable control of PVA-free green sizing agents under low sizing rate conditions, effectively reducing sizing agent consumption and desizing pollution load, and has significant environmental and economic benefits. It is suitable for sizing production of various fabrics such as pure cotton, polyester-cotton blends, pure polyester, and regenerated cellulose fibers. Attached Figure Description

[0065] Figure 1 This is an overall framework diagram of a bio-based slurry low-sizing intelligent control system according to the present invention; Detailed Implementation

[0066] 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 a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0067] Example 1:

[0068] Please see Figure 1 This invention provides a bio-based sizing low-sizing intelligent control system. The system comprises a sensing unit, a model unit, an optimization control unit, and an execution unit. These units are sequentially connected to form a closed-loop control circuit, ultimately achieving multi-parameter collaborative optimization control of the sizing process. The details are as follows:

[0069] A. Regarding the sensory unit:

[0070] The sensing unit collects multi-source process parameters during the sizing process, including sizing state parameters, equipment operating parameters, and quality feedback parameters. Specifically:

[0071] The slurry state parameters include slurry temperature, slurry viscosity, pH value, liquid level, and solids content. The solids content is also known as the slurry concentration, which is the ratio of solids content to subsequent slurry concentration. Considered synonyms;

[0072] Equipment operating parameters include sizing pressure, drying room temperature, sizing machine speed, and yarn tension;

[0073] Quality feedback parameters include sizing rate, moisture regain, and loom warp breakage rate. It should be noted that, due to the complexity of the formulas and models in this invention, and for the sake of accuracy in formula definition, the moisture regain rate involved in the formulas of this invention... Both moisture regain R and moisture regain are different symbols / descriptions of the same parameter;

[0074] The key monitoring parameters and control targets for intelligent sizing with low sizing rates are provided in Table 1 below:

[0075] Table 1: Key Monitoring Parameters and Control Targets for Intelligent Sizing with Low Sizing Rate

[0076] slurry state Temperature T / ℃ 95~115, fluctuation ≤±2 Slurry pot temperature sensor Affects gelatinization, solubility and viscosity stability slurry state Viscosity η / mPa·s 3-25, 3-hour fluctuation ≤15% Online viscometer / offline sampling inspection Determine the ratio of penetration to coverage slurry state pH value 6.5～7.5 pH sensor / random inspection Prevent corrosion of fibers and equipment Equipment parameters Grouting pressure P / kN Set according to yarn count and density Grouting roller pressure sensor Adjusting the sizing rate and penetration depth Equipment parameters Vehicle speed v / m·min⁻¹ Matching drying capacity Encoder / PLC Affects yarn moisture regain and sizing film formation Quality feedback Sizing rate S / % Pure cotton: 5.0–7.0; Polyester-cotton blend: 4.5–6.5 Weighing method / near-infrared estimation Key indicators for controlling low sizing rate Quality feedback menopause rate E Not higher than the baseline process Loom Statistics / Downtime Records Verify process stability

[0077] The above data is collected by the PLC controller and sensors at 1-60 second intervals, including slurry temperature T, slurry viscosity η, pH value, and solids content / slurry concentration. The sizing pressure P, sizing machine speed v, moisture regain W, and warp breakage rate E form multi-source process parameters related to the sizing solution state, equipment operation, and sizing quality, among which:

[0078] in, The slurry state parameters collected by the sensing unit represent the solids content or slurry concentration at the k-th sampling time, in percentage (%). During online operation, the slurry concentration meter outputs the values ​​directly and calibrates the solids content using the drying method for slurry with the same formulation before startup. For offline calibration, a quantitative slurry sample is dried to a constant weight, and the result is calculated as the ratio of the dried solids mass to the sampled slurry mass × 100%. ;

[0079] B. Regarding model units:

[0080] The model unit performs data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on multi-source process parameters, and outputs a prediction model. Data cleaning employs moving median filtering, Hampel outlier removal, and missing value imputation to obtain valid samples. The model unit consists of a sizing rate dry mass correction model, a sizing viscosity stability evaluation model, a yarn sizing comprehensive quality evaluation model, and a sizing rate prediction and recursive parameter identification model, as detailed below:

[0081] B1. Regarding the correction model for sizing rate and dry weight:

[0082] The sizing rate dry weight correction model uniformly corrects the measured wet weight to dry weight, thereby achieving data correction. Sizing rate is a core indicator for evaluating low sizing rate processes. In actual weighing, the moisture content and moisture regain of the warp yarn will affect the weighing results. If the wet weight is used directly for calculation, it is easy to overestimate or underestimate the actual amount of sizing agent adhering. Therefore, it is necessary to uniformly correct the wet weight of the original yarn and the wet weight of the sizing yarn to dry weight.

[0083]

[0084] in, For dry quality, The wet mass is obtained by weighing. This is the moisture content (mass fraction) on a wet basis, calculated from the corresponding moisture regain R. ;

[0085]

[0086] in, for Dry weight of warp yarns before sizing The dry weight of the sized yarn after sizing;

[0087] The above formula can serve as a unified standard for data across laboratory weighing methods, online near-infrared estimation, and production record sheets. For low sizing processes, the control objective is not to pursue the lowest possible sizing rate, but rather to ensure that weaving breakage rate, hairiness, and abrasion resistance meet the requirements. Stabilize within the target range *nearby;

[0088] This modified model is directly derived from the mass conservation relationship of textile material moisture regain: Moisture regain R = (M w -M d ) / M d ×100%, therefore M d =M w / (1+R / 100), to correspond to the correction factor r in the original model, the moisture regain R is converted to the wet basis moisture content mass fraction r=R / (100+R), thus obtaining M. d =M w (1−r), the two ways of writing are equivalent, where M w R corresponds to the same batch and the same sampling time; R is obtained by an online moisture regain meter, or calculated according to the above definition of moisture regain after sampling and drying to a constant mass. When no other volatile components besides moisture cause significant mass loss during weighing, this correction relationship applies to the unified conversion of the dry mass of warp yarn before sizing and sized yarn after sizing. This is achieved by first converting M separately. 0d M 1d Then, by calculating the sizing rate, the systematic deviation caused by different moisture regain states in weighing results can be eliminated;

[0089] B2. Regarding the evaluation model for slurry viscosity stability:

[0090] The viscosity of green slurry is affected by temperature, concentration, holding time, and shear force. To evaluate the stability of the slurry in continuous production, viscosity fluctuation rate can be used. Characterization, as shown above, indicates that the slurry viscosity stability evaluation model utilizes viscosity fluctuation rate... Characterizing the stability of the slurry in continuous production:

[0091]

[0092] In the formula, The initial dynamic viscosity is measured at t=0 after the slurry has completed gelatinization or dissolution and reached the set temperature; The dynamic viscosity is measured after t hours of continuous heat preservation from the initial moment, and both are in mPa·s; t is the heat preservation time. In this embodiment, under the condition of constant solid content and shear conditions, and 95℃±2℃, samples were taken at 0, 1, 2, and 3 hours, and the maximum absolute deviation from the initial value within 3 hours was taken as the dynamic viscosity. , and Prioritize data collection using the same online viscometer, or alternatively, perform offline sampling using the same detection method to avoid additional bias introduced by differences in instruments and methods. When the concentration is ≤15%, the slurry is deemed to possess thermal stability and continuous use stability. A continuous increase indicates that the slurry has the risk of incomplete gelatinization, excessive degradation, foaming, or sedimentation, requiring adjustment of the holding time, temperature, or auxiliary agent ratio. 15% is the process control threshold for continuous production in this system, not a universal physical constant for slurry materials; this threshold is jointly calibrated using the upper limit of 3-hour viscosity fluctuation of stable production samples of the same variety and weaving quality feedback. When the slurry concentration is ≤15%, and there is no obvious stratification, sedimentation, skinning, or abnormal bubbling, and the weaving breakage rate is not higher than the benchmark process, the slurry is judged to have thermal stability and continuous use stability; when When the value exceeds 15% for two consecutive detection cycles, the system enters the slurry adjustment correction or conservative control phase, and prioritizes checking the slurry temperature, solids content, holding time, and shear state. When different slurry systems are put into production, the threshold can be re-determined using the same calibration method. In this embodiment, the default upper limit is 15%.

[0093] B3. Regarding the comprehensive quality evaluation model for sizing:

[0094] The overall quality of sizing can be evaluated by breaking strength, abrasion resistance, hairiness index, and warp breakage rate. The overall quality evaluation model for sizing uses a comprehensive quality index. Evaluating sizing quality:

[0095]

[0096] in, For sizing breaking strength, For wear resistance cycles, The feather index, The rate of warp breakage per unit length, subscript Indicates the standard process or raw yarn condition. As weight;

[0097] and The larger the value, the better the reinforcement and wear resistance. The larger the value, the more prominently the feathers lie flat. The larger the value, the lower the breakage rate. The weight can be adjusted according to the difficulty of the variety. For example, for high-count, high-density pure cotton varieties, the weight of hairiness and abrasion resistance can be increased, and for polyester-cotton varieties with high polyester content, the weight of adhesion and breakage rate can be increased.

[0098] In summary, this system achieves quality evaluation through a slurry viscosity stability evaluation model and a comprehensive yarn quality evaluation model.

[0099] B4. Regarding the model for predicting sizing rate and identifying recursive parameters:

[0100] The difficulty in controlling low sizing rate lies in the fact that sizing rate is affected by multiple coupled factors, and the yarn batch, sizing agent batch, and equipment status can cause model parameter drift. Therefore, the sizing rate prediction and recursive parameter identification model achieves online identification of the production process through a recursive least squares parameter identification formula with a forgetting factor, thereby realizing parameter identification and sizing rate prediction, and finally outputting a prediction model. Specifically:

[0101]

[0102] in, Output as sizing rate or overall quality index. This refers to the sizing concentration. The viscosity of the slurry. The grouting pressure, For the temperature of the slurry, The speed of the sizing machine. For moisture regain, For the parameters to be identified, Noise term;

[0103] The recursive least squares parameter identification formula is updated as follows:

[0104]

[0105]

[0106]

[0107] in, This is the forgetting factor, with a value range of 0.95 to 0.99. The smaller the value, the more the model focuses on the latest data, making it suitable for scenarios with frequent changes in product variety or pulp batches. The closer the value is to 1, the more stable the model is, making it suitable for continuous production of stable varieties.

[0108] Table 2 provides some of the variable symbols and related data sources for the formulas mentioned above, as shown in Table 2:

[0109] Table 2: Variable Symbols and Related Data Sources

[0110] S Sizing rate % Weighing method / Online estimation Low sizing rate is the main control target η Slurry viscosity mPa·s Online viscometer / Encker viscosity Reflects penetration and coverage capabilities C slurry solids content or concentration % Slurry formulation / concentration meter Determines the amount of slurry adhesion P Grouting pressure kN Grouting roller sensor Adjusting the ratio of extrusion desizing to permeation T Slurry or drying room temperature ℃ Temperature sensor Affects gelatinization, drying and film formation v Sizing machine speed m·min⁻¹ PLC / encoder Affecting contact time and moisture regain W Sizing moisture regain % Moisture regain detector Used for dry mass correction E menopause rate times / 10,000 meters Loom downtime statistics Final weaving quality feedback

[0111] To prevent abnormal sensor data from directly entering the controller, anomaly detection indicators can be set at the front end of the model unit. :

[0112]

[0113] In the formula, A k S represents the comprehensive abnormal state index at the k-th sampling time. k η k E k These represent the sizing rate, sizing viscosity, and weaving breakage rate at that moment, respectively; S ref η ref E ref These are the baseline values ​​for sizing rate, slurry viscosity, and warp breakage rate during the stable production stage of the same variety; α, β, and γ are dimensionless weights, satisfying α+β+γ=1, used to reflect the relative sensitivity of the three indicators to anomaly detection. In this embodiment, α=0.4, β=0.3, and γ=0.3 can be used. These values ​​can also be recalibrated based on sensitivity analysis of historical stable samples during production. k Since it is a dimensionless quantity, all input items are normalized to the benchmark value, so process variables with different dimensions can be compared in the same index.

[0114] When|A k -A ref |>A th If the data from any critical sensor exceeds the preset physical / safety hard limit for two consecutive sampling cycles, the system will trigger an alarm and enter a conservative control mode.

[0115] The abnormal threshold is determined by statistical analysis of stable production data, rather than being arbitrarily set by the administrator: no fewer than 30 confirmed normal operating conditions of the same product are continuously collected. k The sample is used to calculate its mean A. ref and standard deviation σ A and take A th =3σ A As a deviation threshold; when |A k -A ref |>A th If two consecutive sampling cycles are detected, it is considered a continuous anomaly. If a single sensor exceeds the device's physical / safety hard limit, conservative control is triggered immediately without waiting for two cycles.

[0116] The difference between conservative control mode and normal mode is that normal mode allows recursive parameter identification and model predictive control to be updated online simultaneously and to perform collaborative optimization for multiple actuators; after entering conservative control, the recursive parameter θ is frozen. k The system updates and calls the most recently validated model parameters, automatically adding grout / water to maintain the most recent safe values, and limits the single-cycle variation of grouting pressure, speed, and temperature to 50% of the normal limit. Only basic control loops such as temperature, liquid level, moisture regain, and equipment safety interlocks are retained, while a manual verification prompt is issued. The system satisfies |A| for three consecutive sampling cycles. k -A ref |≤A thAfter all key sensors have returned to their allowable ranges, the system releases conservative control and resumes normal closed-loop optimization after data integrity verification.

[0117] C. Regarding the optimized control unit:

[0118] After obtaining the sizing rate prediction model, based on the set control objectives and constraints, the model predictive control algorithm is used to solve for the optimal control quantity at the next time step, generating control commands. The control objectives include sizing viscosity, sizing pressure, drying room temperature, and sizing machine speed. The objective is to make the sizing rate approach the target value within the prediction time domain, while suppressing drastic fluctuations in the control quantity. The objective function is:

[0119]

[0120] in, To predict the time domain, To control the time domain, To predict the sizing rate, To achieve the target sizing rate, To control the change in quantity, and This is the weight matrix;

[0121] The constraints include the sizing viscosity range, sizing temperature range, upper limit of sizing pressure, upper and lower limits of sizing machine speed, and moisture regain range. When the predicted sizing rate is lower than the target value and the warp breakage rate increases, the system will preferentially increase the sizing pressure or sizing concentration slightly. When the sizing rate is high and the warp breakage rate is normal, the system will reduce the concentration or sizing pressure and maintain a stable moisture regain by adjusting the drying temperature.

[0122] D. Regarding the execution unit:

[0123] The execution unit performs multi-parameter collaborative optimization control actions according to the control instructions. Specifically, the execution unit performs control actions on sizing and feeding, steam valve opening, sizing roller pressure, sizing machine speed and drying room temperature according to the optimal control quantity. It issues instructions through the PLC controller and sets change limits to avoid drastic fluctuations in the control quantity.

[0124] In normal closed-loop mode, the execution unit operates according to the following coordinated logic: ① The PLC receives the optimal control increment for the current cycle output by the model predictive control algorithm and maps it to the slurry / water metering pump, steam regulating valve, grouting roller pressure actuator, main drive frequency converter, and drying room temperature control actuator, respectively; ② First, it checks the slurry temperature, level, moisture regain, and equipment safety interlock status, then executes slurry adjustment and slurry temperature control, then executes grouting pressure and speed adjustment, and finally performs linkage correction on the drying room temperature and speed based on the moisture regain; ③ When the predicted slurry application rate is lower than the target and the warp breakage rate shows an upward trend, it prioritizes increasing the slurry concentration or decreasing the grouting pressure to increase effective slurry application. Quantity; When the predicted sizing rate is higher than the target and the warp breakage rate is normal, prioritize reducing the sizing concentration or increasing the grouting pressure to reduce the sizing amount; when the moisture regain rate is too high, increase the drying room temperature or reduce the machine speed; when the moisture regain rate is too low, adjust in the opposite direction; ④ Each control quantity is constrained by its upper and lower limits and single-cycle change limit. When the influence directions of two control actions conflict, the model predictive control algorithm will re-solve within the constraint domain according to the priority of "quality and safety constraints first, sizing rate tracking second, and energy consumption optimization third"; ⑤ After the action is completed, maintain a sampling period, re-collect S, η, C, P, T, v, W and E, update the prediction model and enter the next closed loop cycle;

[0125] The green slurry system formulation adapted to the above system in this embodiment is as follows:

[0126] 1. C40×C40 regular pure cotton fabric:

[0127] Slurry system: 87.5-100 kg corn starch + 6 kg TP-II + 82 kg WAX 12, solid content 5.8%-6.5%, viscosity 7-10 mPa·s, slurry preparation process: heat treatment at 95℃ for 25-35 min, target slurry rate 5.5%-6.5%;

[0128] 2. High-count, high-density pure cotton fabric:

[0129] Slurry system: CE50 + modified starch + QL89-1 + WAX128, solid content 6.2%~7.0%, viscosity 9~13mPa·s, slurry preparation process is 95℃ for 35~45min, target slurry rate 6.0%~7.0%;

[0130] 3. T / C65 / 35 polyester-cotton blended fabric:

[0131] Slurry system: HCE / HCS+QL-89+WAX128, solids content 5.5%–6.2%, viscosity 6–9 mPa·s, slurry preparation process: 95℃ for 50–60 min or 115℃ for 15–20 min, target slurry application rate 4.8%–6.0%.

[0132] 4. Pure polyester medium-high count yarn fabrics:

[0133] Slurry system: HCS + environmentally friendly acrylic additive + WAX128, solids content 5.2%–6.0%, viscosity 5–8 mPa·s, slurry preparation process: 115℃ for 15–20 min, target slurry application rate 4.5%–5.8%.

[0134] 5. Linen / cotton / regenerated cellulose fabrics:

[0135] Slurry system: modified starch + QL89-1 + softener and smoother, solid content 6.0%~7.0%, viscosity 8~12mPa·s, slurry preparation process is 95℃ for 30~45min, target slurry rate 5.8%~7.0%;

[0136] The formulation and process parameters of the above slurry system are shown in Table 3:

[0137] Table 3: Formulation and Process Parameters for Low Sizing Rate PVA-Free Sizing

[0138] C40×C40 Regular Pure Cotton Corn starch 87.5–100 kg + TP-II 6 kg + WAX128 2 kg 5.8～6.5 7～10 Keep warm at 95℃ for 25–35 minutes. 5.5～6.5 High-count, high-density pure cotton CE50 + Modified Starch + QL89-1 + WAX128 6.2～7.0 9～13 Keep warm at 95℃ for 35–45 minutes 6.0～7.0 T / C 65 / 35 Polyester / Cotton HCE / HCS + QL-89 + WAX128 5.5～6.2 6～9 Hold at 95℃ for 50–60 min or at 115℃ for 15–20 min. 4.8～6.0 High-count pure polyester yarn HCS + Environmentally Friendly Acrylic Additives + WAX128 5.2～6.0 5～8 Keep warm at 115℃ for 15–20 minutes. 4.5～5.8 Linen / Cotton / Regenerated Cellulose Modified starch + QL89-1 + Softening and smoothing agent 6.0～7.0 8～12 Keep warm at 95℃ for 30–45 min 5.8～7.0

[0139] Example 2:

[0140] Based on the system described in Example 1, the present invention provides a method (i.e., process) for intelligent control of low sizing in bio-based slurries, comprising the following steps:

[0141] S1. The sensing unit collects multi-source process parameters related to the sizing state, equipment operation, and sizing quality during the sizing process.

[0142] S2. The model unit performs data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on the collected multi-source process parameters, and outputs a prediction model.

[0143] S3. The optimization control unit calculates the optimal control quantity based on the prediction model according to the set control objectives and constraints, and generates control commands.

[0144] S4. The execution unit coordinates the sizing feed amount, water replenishment amount, steam valve opening, sizing roller pressure, sizing machine speed and drying room temperature according to the optimal control amount based on the control instructions, and performs control actions according to the preset upper and lower limits, single-cycle change limit and safety interlock conditions.

[0145] S5. Repeat steps S1 to S4 to form a closed-loop control and realize multi-parameter collaborative optimization and control of the sizing process.

[0146] To verify the effectiveness of the process after its operation, the third sizing system described in Example 1 (i.e., T / C65 / 35 polyester-cotton blended fabric) was selected as an example. The traditional PVA-containing process and the intelligent control process for low sizing rate without PVA of the present invention were compared, as shown in Table 4:

[0147] To ensure verifiable and comparable comparisons, both sets of experiments used the same batch of T / C65 / 35 warp yarn, the same sizing machine, and the same subsequent loom group, maintaining consistent environmental conditions and total warp tension settings. The traditional process uses fixed process parameters, while the process of this invention starts with the same initial equipment parameters, differing only in the sizing system, the low sizing rate target, and the coordinated adjustments implemented by the model predictive control algorithm within given boundaries during operation. Specific input conditions, control boundaries, and operational results are listed in Table 4.

[0148] Table 4: Comparative Experiment Input Conditions, Boundary Conditions, and Intelligent Control Operation for Low Sizing Rate

[0149] Comparison object / benchmark conditions T / C65 / 35 warp yarn from the same batch; from the same sizing machine and the same group of looms. T / C65 / 35 warp yarn from the same batch; from the same sizing machine and the same group of looms. Maintain consistency Production batch and machine records Slurry composition PVA + modified starch + acrylic auxiliary slurry + smoothing agent HCE / HCS+QL-89+WAX128, without PVA Formula variables Slurry formulation and feeding records Initial slurry viscosity / mPa·s 10～12 6～9 Set according to their respective slurry systems Online viscometer, 95℃ conditions Slurry tank temperature / ℃ 95±2 95±2 Initial values ​​consistent Slurry tank temperature sensor Grouting pressure P / kN 10.0±0.5, fixed control Initial value: 10.0 ± 0.5; MPC constraint range: 9.0–11.0 Initial values ​​are consistent, and runtime optimization is constrained. Grouting roller pressure sensor Sizing machine speed v / (m·min⁻¹) 50±2, fixed control Initial value 50±2; MPC constraint range 45~55 Initial values ​​are consistent, and runtime optimization is constrained. PLC / encoder Drying oven temperature / ℃ 125±3, fixed control Initial value: 125±3; MPC constraint range: 120~130 Initial values ​​are consistent, and runtime optimization is constrained. Drying oven temperature sensor Target moisture regain W / % 6.5±0.5 6.5±0.5 Maintain consistency Online moisture regain tester Slurry solids content / % 7.5～8.2 5.8～6.5 reduce Slurry preparation records and concentration testing Average sizing rate / % 7.8～8.6 5.4～6.3 Reduced by approximately 20% to 30%. Dry mass corrected weighing method Viscosity fluctuation rate / % ≤12 ≤15 Meets stability requirements Test at 95℃ for 3 hours Sizing reinforcement rate / % 18～25 16～24 Maintain basic Single yarn strength tester Hairiness reduction rate / % 35～45 32～45 Maintain basic Feather tester <![CDATA[Warp breakage rate in weaving / times per ten thousand meters -1 > 0.35～0.55 0.38～0.58 Not significantly increased Loom downtime statistics <![CDATA[Size consumption / kg per ten thousand meters -1 > Baseline 100 70～82 reduce Feeding and output conversion Desizing contamination load PVA residue risk is high. PVA-free, easy to desizing reduce Desizing wastewater testing

[0150] As shown in Table 4, after adopting the system of this invention, under PVA-free conditions, the solid content of the sizing solution decreased from 7.5%–8.2% to 5.8%–6.5%, and the average sizing rate decreased from 7.8%–8.6% to 5.4%–6.3%, a reduction of approximately 20%–30%. The yarn strengthening rate remained at 16%–24%, the hairiness reduction rate remained at 32%–45%, and the weaving warp breakage rate did not increase significantly (0.38–0.58 times / 10,000 meters). Sizing material consumption was reduced to 70%–82% of the baseline, and there was no PVA residue in the desizing wastewater, resulting in a significant reduction in pollution load.

[0151] Example 3:

[0152] The economic and environmental benefits of low-sizing-rate green weaving can be estimated from four aspects: sizing material consumption, energy consumption, water consumption, and desizing treatment costs. This embodiment, based on Example 1, adds a green benefit calculation model to estimate the comprehensive benefits of the low-sizing-rate process. :

[0153]

[0154] in, and These are the sizing rates for the baseline process and the low sizing rate process, respectively (converted from percentages to decimals during calculation). For warp yarn production, For the unit cost of slurry, , This refers to the total energy consumption for the same accounting period and corresponding to the same warp yarn output. For energy unit price, , For the same accounting period and corresponding warp yarn production, the total water consumption for desizing and wastewater treatment is calculated. For the unit cost of water treatment, To amortize the cost of new additives or intelligent control systems, when G>0 and Q is not lower than the benchmark value, it indicates that the low sizing rate process has promotional value; if G>0 but the warp breakage rate increases, the sizing type and control constraints need to be readjusted, and sizing savings cannot be used as the evaluation criterion alone.

[0155] The comprehensive benefit G is constructed using the incremental cost method for the same accounting period: Slurry savings = (S0 − S1)M y c s Energy savings = (E0 − E1)C e Savings from desizing and wastewater treatment = (W0 − W1)C w After deducting the amortization costs of newly added additives and intelligent control systems, C add S0 and S1 are converted from percentages to decimals before calculation. E0, E1 and W0, W1 are all calculated based on the total output of the same warp yarn and the same accounting cycle. C s C e C w The actual average purchase or settlement unit price for the accounting period is taken respectively. This is based on dimensional conversion and cost coefficients rather than subjective weighting. Therefore, after converting each item to the same monetary unit, they can be directly added or subtracted. This structure allows changes in sizing rate to quantitatively correspond to changes in sizing cost through the first item. Equipment energy saving and desizing load reduction are included in revenue through the second and third items respectively. New costs are included through C. add By deducting these factors, a quantitative relationship is established between G and sizing rate, energy consumption, water consumption, and new inputs.

[0156] Example 3 uses the same statistical boundaries for calculation: both the traditional process and the low sizing rate process are aggregated for sizing, energy, and water treatment data based on the same warp yarn output and the same production cycle; if E and W are the unit output consumption recorded in the company's original ledger, they are first multiplied by the warp yarn output M of that cycle. y Converting the total amount and then substituting it into the formula avoids incomparability caused by different outputs or different accounting cycles, and ensures that the comprehensive savings represented by G>0 have a clear cost caliber.

[0157] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A bio-based slurry low-sizing intelligent control system, characterized in that, include: The sensing unit is used to collect multi-source process parameters during the sizing production process; The model unit is used to perform data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on the multi-source process parameters, and outputs a prediction model. An optimization control unit is used to solve for the optimal control quantity based on the prediction model according to the set control objectives and constraints, and generate control commands. An execution unit, which is used to execute multi-parameter collaborative optimization control actions according to the control instructions; The sensing unit, model unit, optimization control unit and execution unit are connected in sequence to form a closed-loop control loop, which ultimately realizes multi-parameter collaborative optimization and control of the sizing process. The model units include a sizing rate dry weight correction model, a sizing viscosity stability evaluation model, a sizing yarn comprehensive quality evaluation model, and a sizing rate prediction and recursive parameter identification model, wherein: The dry weight correction model for sizing rate achieves data correction. The slurry viscosity stability evaluation model and the sizing yarn comprehensive quality evaluation model achieve quality evaluation; The sizing rate prediction and recursive parameter identification model realizes parameter identification and sizing rate prediction. The sizing rate dry mass correction model is used to uniformly correct the measured wet mass to dry mass: ； in, For dry quality, The wet mass is obtained by weighing. This is the moisture content (mass fraction) on a wet basis, calculated from the corresponding moisture regain R. ; ； Among them, the sizing rate Dry weight of warp yarns before sizing The dry weight of the sized yarn after sizing; The slurry viscosity stability evaluation model uses viscosity fluctuation rate. Characterizing the stability of the slurry in continuous production: ； in, The initial dynamic viscosity is measured at t=0 after the slurry has completed gelatinization or dissolution and reached the set temperature; The dynamic viscosity is the value measured after t hours of continuous heat preservation from the initial moment. When the concentration is ≤15%, the slurry is judged to have thermal stability and continuous use stability. The comprehensive quality evaluation model for sizing is based on a comprehensive quality index. Evaluating sizing quality: ； in, For sizing breaking strength, For wear resistance cycles, The feather index, The rate of warp breakage per unit length, subscript Indicates the standard process or raw yarn condition. As weight; The sizing rate prediction and recursive parameter identification model achieves online identification of the production process through a recursive least squares parameter identification formula with a forgetting factor: ； in, Output as sizing rate or overall quality index. This refers to the sizing concentration. The viscosity of the slurry. The grouting pressure, For the temperature of the slurry, The speed of the sizing machine. For moisture regain, For the parameters to be identified, Noise term; The recursive least squares parameter identification formula is updated as follows: ； ； ； in, The forgetting factor has a value range of 0.95 to 0.

99.

2. The intelligent control system for low sizing of bio-based slurry according to claim 1, characterized in that, The multi-source process parameters include slurry state parameters, equipment operating parameters, and quality feedback parameters, wherein: The slurry state parameters include slurry temperature, slurry viscosity, pH value, liquid level, and solid content; The equipment operating parameters include sizing pressure, drying room temperature, sizing machine speed, and yarn tension; The quality feedback parameters include sizing rate, moisture regain rate, and warp breakage rate.

3. The intelligent control system for low sizing of bio-based slurry according to claim 1, characterized in that, The optimized control unit uses a model predictive control algorithm to solve for the optimal control quantity, and the objective function is: ； in, To predict the time domain, To control the time domain, To predict the sizing rate, To achieve the target sizing rate, To control the change in quantity, and This is the weight matrix; The control targets include slurry viscosity, slurry pressure, drying room temperature, and sizing machine speed. The constraints include slurry viscosity range, slurry temperature range, upper limit of slurry pressure, upper and lower limits of sizing machine speed, and moisture regain range.

4. The intelligent control system for low sizing of bio-based slurry according to claim 1, characterized in that, The system also includes a green benefit calculation model, which is used to estimate the overall benefits of low sizing rates. : ； in, and These represent the sizing rates for the baseline process and the low sizing rate process, respectively. The calculations are converted from percentages to decimals. For warp yarn production, For the unit cost of slurry, , This refers to the total energy consumption for the same accounting period and corresponding to the same warp yarn output. For energy unit price, , For the same accounting period and corresponding warp yarn production, the total water consumption for desizing and wastewater treatment is calculated. For the unit cost of water treatment, The cost of adding new additives or intelligent control systems is amortized.

5. A method for intelligently controlling low sizing of bio-based slurry, comprising using a bio-based slurry low sizing intelligent control system as described in any one of claims 1-4, characterized in that, Includes the following steps: S1. The sensing unit collects multi-source process parameters related to the sizing state, equipment operation, and sizing quality during the sizing process. S2. The model unit performs data cleaning, data correction, quality evaluation, parameter identification, and sizing rate prediction on the collected multi-source process parameters, and outputs a prediction model. S3. The optimization control unit calculates the optimal control quantity based on the prediction model according to the set control objectives and constraints, and generates control commands. S4. The execution unit coordinates the sizing feed amount, water replenishment amount, steam valve opening, sizing roller pressure, sizing machine speed and drying room temperature according to the control instructions and the optimal control amount, and performs control actions according to the preset upper and lower limits, single-cycle change limit and safety interlock conditions. S5. Repeat steps S1 to S4 to form a closed-loop control, thereby achieving multi-parameter collaborative optimization and control of the sizing process.

Images