Intelligent control method and system for melt viscosity of regenerative parallel composite polyester fibers based on big data

Abstract

The invention relates to an intelligent production method of waste polyester, in particular to an intelligent control method and system for melt viscosity of regenerated parallel composite polyester fibers based on big data. According to an actual production process, influence factors, namely the screw temperature and the drying temperature, for regulating and controlling the viscosity of the polyester are determined, a database is established according to the influence of the screw temperature and the drying temperature on the viscosity of the polyester to guide viscosity regulation and control, and meanwhile, a viscosity control system is established according to actual production, so that the viscosity of the polyester is controlled. The melt viscosity can be regulated and controlled in a pre-judging, ordered and rapid manner, and the fluctuation of the melt viscosity is controlled to be + / -0.002 dl / g.

Description

Technical Field

[0001] The invention relates to an intelligent production method for waste polyester, and in particular to an intelligent control method and system for melt viscosity of regenerated parallel composite polyester fibers based on big data. Background Art

[0002] The curling structure of general chemical fibers is achieved by mechanically deforming the fibers by certain means, and then fixing the deformed curling structure by heat setting. However, the curling structure of parallel composite fibers is achieved by distributing two components with different heat shrinkage rates in parallel in the same cross section. The different heat shrinkage forces of the two components generate torque, thereby obtaining a permanent and stable curling structure. For two-component parallel composite polyester fibers, a common process is to use two different types of polyester components for parallel composite, such as PET / PBT, PET / PTT parallel composite fibers, etc. Another common process is to use high- and low-viscosity polyesters for parallel composite, and use the difference in polyester viscosity to produce a curling structure.

[0003] The raw materials of recycled parallel composite polyester fibers are mainly polyester bottle flakes and polyester foams. The difference in viscosity between polyester bottle flakes and polyester foams is used to produce curling. However, the sources of recycled materials are wide, and the viscosity of raw materials varies greatly between different batches, resulting in large fluctuations in product curling. At present, the regulation of viscosity still relies mainly on manual control, which is adjusted by adjusting the screw temperature and drying temperature. It is necessary to set up a separate central control personnel, and manual control often has a lag, resulting in product volatility. Therefore, how to control the viscosity fluctuation of polyester melt and improve the stability of product quality has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to provide a method for intelligently controlling melt viscosity of recycled parallel composite polyester fibers based on big data, aiming to solve the technical problem of recycling waste polyester. According to the actual production process, the present invention determines the influencing factors for regulating polyester viscosity: screw temperature and drying temperature, and establishes a database to guide viscosity regulation based on the influence of screw temperature and drying temperature on polyester viscosity. At the same time, according to the actual production, the present invention establishes a viscosity control system, which has the ability to predict, order and quickly regulate melt viscosity, and the melt viscosity fluctuation is controlled within ±0.002dl / g.

[0005] In order to achieve the above-mentioned purpose, the present invention adopts the following technical solutions: A method for intelligently controlling melt viscosity of recycled parallel composite polyester fibers based on big data comprises the following steps: 1) Data collection: Collect data from multiple data sources in the production process, including raw material data, extruder screw temperature, raw material drying temperature, real-time viscosity data and historical production data; 2) Data processing and preprocessing: Clean, standardize and extract features from the collected data to generate structured data in order to establish a multidimensional data model of the temperature-viscosity-raw material relationship; 3) Establish a big database: store the processed structured data in the big database, which includes raw material information table, screw temperature-viscosity relationship data table, drying temperature-viscosity relationship data table and historical production data table; 4) Data analysis and modeling: Based on the data in the big database, a viscosity prediction model is established using a machine learning algorithm to generate viscosity prediction results and analyze the effects of different production parameters on melt viscosity; 5) Real-time monitoring and feedback: Data from the production process is collected in real time through sensors and online viscometers, and the real-time data is compared with the results of the viscosity prediction model to calculate the difference between the actual viscosity and the target viscosity; 6) Intelligent control: According to the difference in real-time monitoring feedback, the screw temperature is used as the main control factor, and the drying temperature and raw materials are used as secondary control factors to adjust the melt viscosity to ensure that the viscosity is stable within the preset range.

[0006] Preferably, the data collection step further comprises: collecting data in real time from sensors and online viscometers in the production workshop to obtain current screw temperature, drying temperature and melt viscosity.

[0007] Preferably, the data processing and preprocessing steps include: 1) Clean the data to remove outliers and noise data; 2) Standardize data from different sources; 3) Extract key features from the raw data, including the effect of temperature on viscosity and the effect of different raw material batches on product quality.

[0008] Preferably, the data analysis and modeling step includes using association rule analysis, regression analysis and neural network algorithm to establish a viscosity prediction model, and finding out the relationship between temperature, raw material ratio and viscosity through data mining.

[0009] Preferably, the real-time monitoring and feedback step includes utilizing a stream computing platform to process the data stream in real time, and comparing the real-time viscosity data with the output of the prediction model to generate a feedback signal.

[0010] Preferably, the intelligent control step comprises: 1) Automatically adjust the screw temperature according to the feedback signal through the control system; 2) When temperature regulation is not enough to control viscosity, viscosity can be further controlled by adjusting drying temperature and raw material ratio; 3) Based on the results of big data analysis, optimize the production parameter combination and improve the stability of product quality.

[0011] Preferably, the adjustment range of each screw temperature zone is: 295-320°C for the feeding section, 295-345°C for the melting section, 280-330°C for the homogenization section, and 280-300°C for the mixing section. The temperature is controlled by increasing or decreasing the temperature in each zone at the same time. The drying temperature adjustment range is 115-145°C.

[0012] Preferably, the screw temperature-viscosity relationship control table includes a viscosity-time variation graph after temperature change and a viscosity-temperature variation graph, and the temperature-time-viscosity variation relationship control table is established by a linear regression statistical method.

[0013] Preferably, the viscosity is controlled within ±0.002 dl / g to ensure high quality production of the regenerated side-by-side composite polyester fibers.

[0014] Furthermore, the present invention also discloses an intelligent control system for melt viscosity of recycled parallel composite polyester fibers, which adopts the method described above and comprises the following components: A large database for collecting and storing data in the production process, including raw material data, production parameter data, environmental data, real-time viscosity data and historical data; Data processing and preprocessing modules for cleaning, standardization, and feature extraction to generate structured data; A data analysis and modeling module, based on the data in the big database, uses a machine learning algorithm to establish a viscosity prediction model and generate a viscosity prediction result; Real-time monitoring and feedback module, used to collect data from the production process in real time, compare the real-time data with the results of the viscosity prediction model, and calculate the difference between the actual viscosity and the target viscosity; Intelligent control center, based on feedback data and viscosity prediction results, automatically adjusts production parameters to stabilize viscosity; The control system, including the screw temperature control system, the drying temperature control system and the intelligent distribution and mixing system, adjusts the screw temperature, drying temperature and raw material ratio according to the instructions of the intelligent control center.

[0015] The present invention has the following technical effects due to the adoption of the above technical solution: 1. Improve the stability of the production process The present invention can accurately obtain the screw temperature, drying temperature, raw material ratio, environmental conditions and real-time viscosity data during the production process through real-time data collection and monitoring. These data are integrated and analyzed through a big data platform to form a full range of monitoring of the production process. Through these monitoring, any viscosity fluctuations beyond the normal range can be quickly discovered and responded to, thereby maintaining the stability of the production process.

[0016] 2. Enhance the consistency of product quality By establishing a large database and viscosity prediction model, the present invention can accurately predict and control melt viscosity according to different raw material batches, temperature conditions and other production parameters. By adjusting the screw temperature, drying temperature and raw material ratio in real time, it ensures that the product viscosity is stable within the set range (for example, controlled within ±0.002dl / g), thereby significantly improving the consistency of the quality of the recycled parallel composite polyester fiber product.

[0017] 3. Reduce manual intervention and regulatory lag The traditional production process relies on manual monitoring and adjustment of temperature and raw material ratio, which not only requires specialized central control personnel, but is also prone to lag and inaccuracy. By adopting an intelligent control center and an automatic control system, the present invention can automatically respond and adjust the viscosity at the moment of detecting the change in viscosity. Through an intelligent algorithm based on big data, the viscosity control process is automatically completed, greatly reducing manual intervention and improving production efficiency.

[0018] 4. Optimize production efficiency and energy consumption Through precise data analysis and intelligent regulation, the present invention can optimize the control of screw temperature and drying temperature while ensuring product quality, thereby minimizing energy consumption. In addition, through reasonable raw material ratio adjustment, the present invention reduces resource waste caused by uneven raw materials, thereby improving raw material utilization and reducing production costs.

[0019] 5. Improve the adaptability and flexibility of production systems The present invention continuously learns and updates the model of historical data and real-time data through machine learning algorithms, so that the system can adapt to different production conditions and raw material changes. This adaptive ability enables the production system to maintain efficient operation and stable product quality when facing different raw material batches and production requirements.

[0020] 6. Realize predictive maintenance and failure warning Through big data analysis, the present invention can not only predict the change of viscosity during the production process, but also detect abnormalities in the operation status of the equipment in advance. By analyzing the equipment operation data, possible equipment failures can be predicted and maintenance can be performed in advance, thereby avoiding production interruptions and quality problems caused by equipment failures.

[0021] 7. Increase the breadth of technology applications The intelligent control method and system of the present invention are not only applicable to the production of recycled parallel composite polyester fibers, but can also be extended to other similar chemical fiber material production processes and other industrial production fields that require precise control of physical property parameters. This wide adaptability makes the present invention have high commercial application value.

[0022] In summary, the present invention realizes precise control of melt viscosity of recycled parallel composite polyester fiber through intelligent control driven by big data, which not only significantly improves the stability of the production process and the consistency of product quality, but also reduces production costs and optimizes resource utilization. Its adaptive ability and intelligent characteristics give it significant technical advantages and broad application prospects in modern production. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a system framework diagram of the present invention. DETAILED DESCRIPTION

[0024] The following is a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Example 1: Construction of viscosity intelligent control system based on big data like Figure 1 A method for intelligently controlling melt viscosity of recycled parallel composite polyester fibers based on big data is shown, comprising the following steps: 1) Data collection: Collect data from multiple data sources in the production process, including raw material data, extruder screw temperature, raw material drying temperature, real-time viscosity data and historical production data; 2) Data processing and preprocessing: Clean, standardize and extract features from the collected data to generate structured data in order to establish a multidimensional data model of the temperature-viscosity-raw material relationship; 3) Establish a big database: store the processed structured data in the big database, which includes raw material information table, screw temperature-viscosity relationship data table, drying temperature-viscosity relationship data table and historical production data table; 4) Data analysis and modeling: Based on the data in the big database, a viscosity prediction model is established using a machine learning algorithm to generate viscosity prediction results and analyze the effects of different production parameters on melt viscosity; 5) Real-time monitoring and feedback: Data from the production process is collected in real time through sensors and online viscometers, and the real-time data is compared with the results of the viscosity prediction model to calculate the difference between the actual viscosity and the target viscosity; 6) Intelligent control: According to the difference in real-time monitoring feedback, the screw temperature is used as the main control factor, and the drying temperature and raw materials are used as secondary control factors to adjust the melt viscosity to ensure that the viscosity is stable within the preset range.

[0026] 1. Data Collection In this embodiment, the data collection system collects data from the following data sources: Raw material data: including the type, ratio, supplier information, actual viscosity of raw materials, etc. The specific raw material ratio is 70% for polyester bottle flakes and 30% for polyester foam.

[0027] Screw temperature: The screw temperature data is collected by temperature sensors installed in different sections of the extruder, including the feeding section, melting section, homogenization section, and mixing section. The temperature range is 295°C to 345°C.

[0028] Drying temperature: The raw material drying temperature data is obtained through the temperature sensor of the drying system, and the temperature range is 115°C to 145°C.

[0029] Real-time viscosity data: Get melt viscosity data in real time by installing an online viscometer.

[0030] Historical production data: Contains all relevant data from previous production processes, including records of successful cases and abnormal situations.

[0031] In the data collection step, data is collected from different data sources, which can be represented as: Raw material data: R i Represents the ratio of the i-th raw material.

[0032] Screw temperature: T s,j Indicates the screw temperature of the jth section (feeding section, melting section, homogenization section, mixing section).

[0033] Drying temperature: T d Indicates the drying temperature.

[0034] Real-time viscosity data: V t Represents the viscosity at time t.

[0035] Environmental data: E t (Temperature), H t (humidity).

[0036] Historical production data: D hist Represents all past production data records.

[0037] 2. Data Processing and Preprocessing The collected data is processed and preprocessed through the following steps: Data cleaning: Remove outliers and noise from the data. For example, remove data that exceeds the detection range of the temperature sensor or the error range of the viscometer.

[0038] The processed data is represented in the form of structured data X: X=\left [ {{T}_{s,1},{T}_{s,2},...,{T}_{s,n},{T}_{d},{R}_{1},{R}_{2},...,{R}_{m},{E}_{t},{H}_{t}} \right ] ; Where n is the number of screw temperature zones and m is the number of raw material types.

[0039] Data cleaning: remove outliers X clean ={x∈X|x is valid} Data standardization: Normalize data from different sources so that they can be compared on the same scale. The standardization formula is: X norm =( X - μ X ) / σ X

[0040] Among them, X is the original data, μ X and σ X are the mean and standard deviation of the data respectively.

[0041] Feature extraction: Extract key features F from the data, such as the effect of temperature on viscosity and the effect of different raw material batches on product quality. Use the principal component analysis (PCA) method to reduce the dimension of the data, F=PCA(X norm ); in order to conduct subsequent analysis more efficiently.

[0042] 3. Building a large database The processed structured data is stored in a large database, which contains the following tables: Raw material information table: record the raw material type, ratio, supplier information and actual viscosity value; R={R 1 ,R 2 ,…,R m}.

[0043] Screw temperature-viscosity relationship data table: record the relationship between different screw temperature sections and melt viscosity; T s →V.

[0044] Drying temperature-viscosity relationship data table: record the effect of drying temperature on viscosity; T d →V.

[0045] Historical production data table: contains all historical production process parameters and corresponding viscosity data; D hist .

[0046] 4. Data Analysis and Modeling Based on the data in the big database, the viscosity prediction model was established using the following machine learning algorithm: Association rule analysis: Discover the influence of screw temperature, drying temperature and raw material ratio on viscosity. Through data mining technology, extract key association rules for subsequent prediction and control.

[0047] Linear regression model: predicts melt viscosity. The formula for the linear regression model is: ;

[0048] in, is the predicted viscosity value; β 0 : intercept (constant term) of the model, reflecting the base viscosity when all other variables are taken as zero; T s,j : The temperature of the jth section of the screw; n : Number of screw sections; β j : The regression coefficient of the j-th section temperature of the screw, which indicates the weight of the influence of the j-th section temperature on the melt viscosity; T d : Drying temperature of raw materials; β n+1 : Regression coefficient of drying temperature, indicating the influence weight of drying temperature on melt viscosity; R i : The ratio of the i-th raw material; m : Number of raw material types; β n+1+i : The regression coefficient of the i-th raw material ratio, indicating the influence weight of this raw material ratio on melt viscosity.

[0049] Neural network model: used for modeling complex nonlinear relationships. The neural network uses a multi-layer perceptron structure to process the nonlinear effects of raw material ratio, temperature and environmental conditions on viscosity.

[0050] 5. Real-time monitoring and feedback Install sensors and online viscometers to monitor screw temperature, drying temperature and melt viscosity data in real time during the production process.

[0051] Real-time data is processed through a stream computing platform (such as Apache Kafka), the actual viscosity data is compared with the prediction model results, and the viscosity error ΔV is calculated t : ;

[0052] Generate feedback signals to control the system.

[0053] 6. Intelligent Control Screw temperature control: According to the feedback signal, adjust the temperature of each screw section to mainly control the screw temperature. For example, use a PID controller to adjust the temperature: ; in: ΔT s , temperature adjustment amount, the adjustment value of the screw temperature calculated by the PID controller; k p , proportional control coefficient (Proportional Gain), indicating the current deviation ΔV t Temperature adjustment ΔT s direct contribution of k i , integral control coefficient (Integral Gain), weighing the contribution of cumulative deviation to temperature adjustment; k d , Derivative Gain, weighs the contribution of the deviation change rate to the temperature adjustment amount.

[0054] Drying temperature and raw material ratio adjustment: When the screw temperature adjustment is not enough to meet the viscosity requirements, further adjust the drying temperature and raw material ratio. Use optimization algorithms (such as genetic algorithms) to find the best parameter combination so that the viscosity is stable within the set range.

[0055] ;

[0056] Where η is the learning rate and L is the viscosity loss function.

[0057] 7. System optimization and model update Based on the latest production data and feedback, the model parameters in the big database are continuously updated to adapt the model to different production conditions and raw material changes.

[0058] Through continuous learning and optimization, the intelligent control center can adaptively adjust production parameters to cope with different batches of raw materials and environmental changes, with screw temperature as the main control factor, and drying temperature and raw materials as secondary control factors to adjust melt viscosity. The adjustment range of screw temperature in each zone is: 295-320℃ for feeding section, 295-345℃ for melting section, 280-330℃ for homogenization section, and 280-300℃ for mixing section. Temperature control adopts the method of increasing or decreasing in each zone; the adjustment range of drying temperature is 115-145℃; ensure that viscosity is controlled within ±0.002dl / g. When the viscosity fluctuation tends to break through the difference range, the control center will issue a temperature adjustment instruction to the screw temperature control system for viscosity control based on the database information; when the temperature adjustment of the screw temperature control system cannot meet the viscosity requirements, the control center will send the temperature adjustment instruction to the drying temperature control system for temperature adjustment; further, when the temperature adjustment of the drying temperature control system cannot meet the viscosity requirements, the control center will transmit the adjustment instruction to the raw material adjustment system to adjust the melt viscosity through the raw material adjustment method.

[0059] The viscosity feedback system is divided into a high viscosity feedback system and a low viscosity feedback system, which feeds back the real-time difference between the viscosity set value and the actual viscosity value to the control center; The screw temperature control system controls the melt viscosity by simultaneously increasing / decreasing the temperature of each zone of the screw. The control is controlled by the control center. Heating promotes thermal decomposition and reduces melt viscosity, while cooling alleviates thermal degradation and increases melt viscosity. The drying temperature control system controls the melt viscosity by raising / lowering the temperature, and its control is controlled by the control center; raising the temperature can reduce the moisture in the raw materials and avoid hydrolysis, while increasing the solid phase viscosity and increasing the melt viscosity; lowering the temperature can increase the moisture in the raw materials, intensify hydrolysis, slow down the solid phase viscosity and reduce the melt viscosity; the operation of the drying temperature control system starts after the screw temperature control system cannot meet the requirements by raising and lowering the temperature.

[0060] The raw material control system stores the actual viscosity values ​​of each raw material, and includes a polyester recycled material intelligent distribution and mixing system and method (Ningbo Dafa New Materials Co., Ltd. CN 109693905 A). Its operation starts when the viscosity still cannot meet the requirements after the drying temperature control system is adjusted.

[0061] Example 2: Predictive maintenance and fault warning In this embodiment, the system is used not only for real-time viscosity control but also for predictive maintenance and fault warning.

[0062] Equipment operation data collection: Through the sensors installed on the equipment, the operating status data of the extruder and drying system are collected, including equipment vibration, noise, temperature, operating time, etc.

[0063] Data analysis and fault prediction: Use machine learning algorithms (such as decision trees and support vector machines) to analyze equipment operation data, predict possible equipment failures, and generate maintenance recommendations.

[0064] Fault alarm and maintenance: When it is predicted that equipment may fail, the system automatically issues an alarm to remind operators to perform maintenance in advance, reducing production interruptions and quality problems caused by equipment failure.

[0065] The above is a description of the embodiments of the present invention. Through the above description of the disclosed embodiments, professionals and technicians in the field can implement or use the present invention. Various modifications to these embodiments will be apparent to professionals and technicians in the field. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown in this article, but will conform to the widest range consistent with the principles and novelties disclosed herein.

Claims

1. A method for intelligently controlling melt viscosity of recycled parallel composite polyester fibers based on big data, characterized in that: The following steps are involved: 1) Data collection: Collect data from multiple data sources in the production process, including raw material data, extruder screw temperature, raw material drying temperature, real-time viscosity data and historical production data; 2) Data processing and preprocessing: Clean, standardize and extract features from the collected data to generate structured data in order to establish a multidimensional data model of the temperature-viscosity-raw material relationship; 3) Establish a big database: store the processed structured data in the big database, which includes raw material information table, screw temperature-viscosity relationship data table, drying temperature-viscosity relationship data table and historical production data table; 4) Data analysis and modeling: Based on the data in the big database, a viscosity prediction model is established using a machine learning algorithm to generate viscosity prediction results and analyze the effects of different production parameters on melt viscosity; 5) Real-time monitoring and feedback: Data from the production process is collected in real time through sensors and online viscometers, and the real-time data is compared with the results of the viscosity prediction model to calculate the difference between the actual viscosity and the target viscosity; 6) Intelligent control: According to the difference in real-time monitoring feedback, the screw temperature is used as the main control factor, and the drying temperature and raw materials are used as secondary control factors to adjust the melt viscosity to ensure that the viscosity is stable within the preset range.

2. The method according to claim 1, characterized in that The data collection step further includes: collecting data from sensors and online viscometers in the production workshop in real time to obtain current screw temperature, drying temperature and melt viscosity.

3. The method according to claim 1, characterized in that The data processing and preprocessing steps include: 1) Clean the data to remove outliers and noise data; 2) Standardize data from different sources; 3) Extract key features from the raw data, including the effect of temperature on viscosity and the effect of different raw material batches on product quality.

4. The method according to claim 1, characterized in that The data analysis and modeling steps include using association rule analysis, regression analysis and neural network algorithm to establish a viscosity prediction model, and finding out the relationship between temperature, raw material ratio and viscosity through data mining.

5. The method according to claim 1, characterized in that The real-time monitoring and feedback step includes using a stream computing platform to process the data stream in real time, and comparing the real-time viscosity data with the output of the prediction model to generate a feedback signal.

6. The method according to claim 1, characterized in that The intelligent control step comprises: 1) Automatically adjust the screw temperature according to the feedback signal through the control system; 2) When temperature regulation is not enough to control viscosity, viscosity can be further controlled by adjusting drying temperature and raw material ratio; 3) Based on the results of big data analysis, optimize the production parameter combination and improve the stability of product quality.

7. The method according to claim 1, characterized in that The adjustment range of each screw temperature zone is: 295-320°C for the feeding section, 295-345°C for the melting section, 280-330°C for the homogenization section, and 280-300°C for the mixing section. The temperature is controlled by increasing or decreasing the temperature in each zone at the same time. The drying temperature adjustment range is 115-145°C.

8. The method according to claim 1, characterized in that The screw temperature-viscosity relationship control table includes a viscosity-time change spectrum after temperature change and a viscosity-temperature change spectrum. The temperature-time-viscosity change relationship control table is established by a linear regression statistical method.

9. The method according to claim 1, characterized in that: The viscosity is controlled within ±0.002dl / g.

10. An intelligent control system for melt viscosity of recycled parallel composite polyester fibers, characterized in that: The method according to any one of claims 1 to 9 comprises the following components: A large database for collecting and storing data in the production process, including raw material data, production parameter data, environmental data, real-time viscosity data and historical data; Data processing and preprocessing modules for cleaning, standardization, and feature extraction to generate structured data; A data analysis and modeling module, based on the data in the big database, uses a machine learning algorithm to establish a viscosity prediction model and generate a viscosity prediction result; Real-time monitoring and feedback module, used to collect data from the production process in real time, compare the real-time data with the results of the viscosity prediction model, and calculate the difference between the actual viscosity and the target viscosity; Intelligent control center, based on feedback data and viscosity prediction results, automatically adjusts production parameters to stabilize viscosity; The control system, including the screw temperature control system, the drying temperature control system and the intelligent distribution and mixing system, adjusts the screw temperature, drying temperature and raw material ratio according to the instructions of the intelligent control center.

Images