Intelligent control method for traction speed of solar glass tube drawing machine

Abstract

The present application relates to the technical field of industrial automation control, in particular to a method for intelligent control of traction speed of a solar glass tube drawing machine. The method comprises: collecting actual speed sequence of a traction motor, actual diameter sequence of a laser diameter measuring instrument and electrode current data of an electric smelting furnace and preprocessing to obtain a time sequence alignment sequence containing a plurality of continuous sampling points; performing linear regression analysis on the time sequence alignment sequence based on a preset sliding window to calculate actual control sensitivity, and calculating a geometric sensitivity benchmark according to a preset target diameter to further obtain a thermal speed coupling efficiency factor, and implementing lower limit limiting on the thermal speed coupling efficiency factor; using the thermal speed coupling efficiency factor to dynamically compensate the PID gain and superimpose a current feedforward amount to generate an instruction to drive the traction motor to make relevant adjustments. The present application effectively suppresses control oscillation caused by thermal speed reverse coupling effect and realizes safe, self-adaptive and precise control of traction speed.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control technology, specifically to an intelligent control method for the traction speed of a solar glass tube drawing machine. Background Technology

[0002] In the production of glass tubes, the Dana process or Vero process is usually used for forming. In this process, the traction motor is the key actuator that controls the geometric dimensions of the glass tube. It directly controls the final diameter of the glass tube by changing the traction speed. According to the law of conservation of volumetric flow rate, an increase in traction speed will lead to a decrease in the diameter of the glass tube, and vice versa.

[0003] However, existing control technologies typically employ standard PID control algorithms, which presuppose a linear and stable inverse relationship between speed and diameter, with PID parameters usually set to fixed values. However, in actual production, there is a significant thermal-speed inverse coupling effect. Specifically, when the controller commands the traction motor to accelerate in order to reduce the diameter of the glass tube, the accelerated glass flow will rapidly carry the high-temperature molten glass out of the electric melting furnace, causing the temperature in the forming zone to rise, the viscosity of the molten glass to decrease and soften, and the softened molten glass will become excessively thin under the action of gravity. Furthermore, the decrease in the viscosity of the molten glass will also lead to a decrease in the transmission efficiency of the traction force of the traction motor.

[0004] The dynamic changes in these physical properties lead to problems with the existing technology, such as action backlash and stiffness inconsistency. That is, the temperature rise caused by the acceleration command cancels out the expected thinning effect and may even cause system oscillation. In addition, fixed PID parameters are difficult to adapt to changes in the thermal environment caused by long-term production and can only make hysteretic adjustments to the diameter error that has already occurred, and cannot detect potential interference from changes in enthalpy inside the electric melting furnace. Summary of the Invention

[0005] To address the problems of control failure and system oscillation caused by neglecting the thermal-velocity reverse coupling effect in existing technologies, this invention provides an intelligent control method for the traction speed of a solar glass tube drawing machine, the method comprising: The actual speed sequence of the traction motor, the measured diameter sequence of the glass tube, and the electrode current data of the electric melting furnace are collected in real time by a preset multi-dimensional data acquisition unit and preprocessed to obtain a time-aligned sequence containing several continuous sampling points. A sliding window is preset, and linear regression analysis is performed on the time-aligned sequence based on the sliding window to calculate the actual control sensitivity. At the same time, a geometric sensitivity benchmark is calculated based on the preset target diameter. Based on the actual control sensitivity and the geometric sensitivity benchmark, the thermal-rate coupling efficiency factor is calculated, and the rate of change of the electrode current data is calculated. A feedforward compensation amount is constructed by combining the pre-calibrated thermal-rate conversion coefficient. The PID control gain is dynamically compensated based on the thermal-rate coupling efficiency factor, and an adaptive speed control command is generated by combining the feedforward compensation amount, thereby driving the traction motor to make relevant adjustments.

[0006] This invention constructs a comparison model between actual control sensitivity and geometric sensitivity benchmarks to evaluate in real time the impact of the traction motor's traction speed on the glass tube diameter under the current process environment, especially when the viscosity of the molten glass changes. Based on this, the PID gain is dynamically adjusted to ensure that the control intensity can be automatically adjusted regardless of how the viscosity of the molten glass changes.

[0007] Furthermore, the actual control sensitivity satisfies the following relationship:

[0008] in, The actual control sensitivity; The length of the sliding window; The index of the sampling point within the sliding window ranges from 1 to... ; For the first in the sliding window The traction speed of each of the sampling points; For the first in the sliding window The measured diameter of each of the aforementioned sampling points; The arithmetic mean of the traction velocities of all the sampling points within the sliding window; It is the arithmetic mean of the measured diameters of all the sampling points within the sliding window.

[0009] This invention employs linear regression analysis based on a sliding window to calculate adaptive speed control commands. Compared to simple differential calculation, this invention can effectively suppress minor fluctuations in the readings of the multidimensional data acquisition unit and high-frequency random noise. It can accurately determine the true response trend between the traction motor and the diameter of the glass tube from the interference-filled industrial field data, avoiding calculation distortion caused by errors at individual sampling points and ensuring the stability and reliability of subsequent calculations.

[0010] Furthermore, the geometric sensitivity benchmark is calculated based on the law of conservation of volumetric flow rate, combined with the arithmetic mean of the target diameter and the traction velocity of all the sampling points within the sliding window.

[0011] Furthermore, the adaptive speed control command satisfies the following relationship:

[0012] in, This refers to the adaptive speed control command; As the reference speed; This refers to the thermal coupling efficiency factor; , These are the preset proportional parameters and integral parameters, respectively; The deviation between the measured diameter of the current sampling point and the target diameter in the time-aligned sequence; The term represents the historical cumulative sum of the aforementioned deviations; The heat conversion rate coefficient is... The rate of change of the electrode current data; The term refers to the feedforward compensation amount.

[0013] This invention utilizes the thermal-rate coupling efficiency factor as a dynamic gain adjustment coefficient to achieve real-time correction of the proportional and integral terms of the PID controller. When changes in the process environment cause a decrease in the traction efficiency of the traction machine on the glass tube, the output force of the PID controller is automatically increased to compensate for the decrease in control efficiency caused by the decrease in the viscosity of the molten glass, effectively improving the anti-interference capability.

[0014] Furthermore, the thermal conversion rate coefficient was comprehensively calibrated through an open-loop step experiment.

[0015] Furthermore, the preprocessing includes: unifying the length measurement units of the collected data, and performing time-series alignment between the actual velocity sequence and the measured diameter sequence to obtain the time-series aligned sequence.

[0016] Furthermore, acquiring the actual speed sequence of the traction motor includes: reading the real-time operating frequency of the traction motor through the communication interface of the inverter driving the traction motor, and converting the real-time operating frequency into linear speed.

[0017] Furthermore, the multi-dimensional data acquisition unit includes: a traction motor encoder for acquiring the actual speed sequence, a laser diameter gauge for acquiring the measured diameter sequence, and a current transformer for acquiring the electrode current data.

[0018] Furthermore, the length of the sliding window is determined based on the transmission lag time of the traction motor during the traction process.

[0019] This invention establishes a strict correlation between the length of the sliding window and the transmission lag time of the traction motor during the traction process, which ensures that the actual speed sequence and the measured diameter sequence are physically strictly corresponding. This avoids the problems of incomplete capture of effective information due to an excessively short sliding window or diluted data correlation due to an excessively long sliding window.

[0020] Furthermore, the intelligent control method for the traction speed of the solar glass tube drawing machine also includes: setting a preset safety lower limit threshold, determining whether the thermal coupling efficiency factor is less than the safety lower limit threshold, if so, assigning the thermal coupling efficiency factor to the safety lower limit threshold, if not, then keeping the value of the thermal coupling efficiency factor unchanged.

[0021] This invention, by setting a preset safety lower limit threshold, can ensure that the thermal coupling efficiency factor is forcibly locked under extreme operating conditions, preventing system oscillations and equipment overload caused by over-adjustment, and improving the safety of the entire production process.

[0022] The present invention has the following technical effects: This invention addresses the problem of oscillations caused by temperature fluctuations and the influence of glass melt viscosity changes on the stretching effect, which is often overlooked in traditional techniques. By constructing a thermal-rate coupling efficiency factor, this invention can sense glass melt viscosity changes in real time and automatically adjust the control intensity, effectively suppressing the bamboo-like pattern in the pipe diameter caused by blind adjustment.

[0023] This invention analyzes the problem that traditional PID control can only adjust after the diameter of the glass tube deviates, and establishes a feedforward adjustment mechanism based on energy changes. Before the diameter of the glass tube deviates, the traction speed of the traction motor is adjusted in advance according to the current change of the electric melting furnace, thus ensuring the accuracy of the product. The method of this invention does not require complex parameter tuning based on human experience, and can adapt to various working conditions from cold start-up to stable production. It effectively reduces the dependence on the skills of operators and can bring high economic benefits. Attached Figure Description

[0024] Figure 1 This is a flowchart of an intelligent control method for the traction speed of a solar glass tube drawing machine provided in an embodiment of the present invention; Figure 2 This is a comparison chart of the diameter control effect of the method provided in this embodiment of the invention and the existing PID control technology when facing thermal interference. Figure 3 This is a graph showing the linkage change between the thermal coupling efficiency factor and the traction speed command provided in an embodiment of the present invention. Detailed Implementation

[0025] This invention provides an intelligent control method for the traction speed of a solar glass tube drawing machine, referring to... Figure 1 This includes steps S1-S4: S1: Data Acquisition and Preprocessing.

[0026] Specifically, the actual speed sequence of the traction motor, the measured diameter sequence of the glass tube, and the electrode current data of the electric melting furnace are collected in real time through a preset multi-dimensional data acquisition unit. The collected data are preprocessed to obtain a time-aligned sequence containing several consecutive sampling points.

[0027] The hardware configuration used in this embodiment is as follows: a Siemens S7-1500 PLC as the main controller, a YW90S-4 variable frequency traction motor with encoder (hereinafter referred to as the traction motor) as the actuator, a Keyence LS-9000 series laser diameter gauge as the detection unit, and a current transformer installed at the flow channel of the electric melting furnace. The above hardware together constitutes a multi-dimensional data acquisition unit, which collects data with a sampling period of 100 milliseconds. The set of discrete data, including traction speed, glass tube diameter, electrode current data, etc., collected at each sampling moment is defined as an independent sampling point.

[0028] This step is the cornerstone of the entire control logic, aiming to obtain raw datasets reflecting the causal relationships in the production process. Due to the complexity of the industrial environment and the different communication protocols of various devices, refined data processing is required. The specific operations are as follows: 1. Acquisition of multi-source heterogeneous data and unification of units (1) Actual speed sequence: The real-time operating frequency of the traction motor is read by the encoder of the traction motor. The number of pole pairs, reduction ratio parameters, and traction wheel diameter of the traction motor are obtained by referring to the technical specifications of the traction motor. The real-time operating frequency is converted into linear speed by combining the linear speed conversion formula, which is defined as the traction speed. The unit is uniformly m / min. The real-time collected traction speeds are combined into a set to obtain the actual speed sequence, denoted as . ; (2) Measured Diameter Sequence: The outer diameter of the glass tube is scanned in real time using a laser diameter gauge. A digital signal is output via RS485 communication. This signal is then converted from digital to digital (A / D) and linearly mapped according to the measurement range. The signal is converted into a physical length value, which is the diameter of the glass tube, defined as the measured diameter. In this embodiment, the measurement range is selected as 0-100 mm, with the unit uniformly in millimeters. The real-time collected measured diameters are combined into a set to obtain the measured diameter sequence, denoted as... ; (3) Electrode current data: The real-time operating current of the electrodes at the flow channel of the electric melting furnace is collected by a current transformer and defined as electrode current data, denoted as . Because molten glass exhibits ionic conductivity at high temperatures, its resistivity decreases with increasing temperature. Therefore, the collected... It can be used as a proxy variable to reflect changes in the viscosity and enthalpy of molten glass, and can be used for subsequent feedforward compensation calculations; (4) Physical transmission distance: The physical path length between the traction motor and the laser diameter measuring instrument is measured as the physical transmission distance, and the unit is uniformly meters.

[0029] It should be noted that the electrode current data is only one embodiment for characterizing the enthalpy state and viscosity of the molten glass in the electric melting furnace. In other embodiments, physical quantities such as electrode power, direct temperature measurement data of the molten glass, and furnace pressure can also be used as feedforward variables, all of which are within the scope of protection of this invention.

[0030] 2. Timing alignment based on transmission lag Since the production of glass tubes is a continuous process, and there is a certain physical transmission distance between the traction motor and the laser diameter gauge, this means that the measured diameter collected at the current sampling moment is actually determined by the traction speed of the traction motor over a previous period. Therefore, timing alignment is required, as follows: (1) Determine the transmission lag time: The main controller reads the physical transmission distance collected by the multi-dimensional data acquisition unit, presets a time window, and in this embodiment, takes the past 1.5 seconds. Calculate the arithmetic mean of all data in the actual speed sequence within the time window, divide the physical transmission distance by the arithmetic mean to obtain the physical transmission lag time, divide it by the sampling period and round it to obtain the number of sampling points included in the physical transmission lag time, which is recorded as the lag step number; (2) Set up a circular buffer in memory to store historical data and store the traction speed in chronological order; perform the following operations at each sampling moment: store the traction speed collected at that moment into the head of the circular buffer. As new data continues to flow in, the old data will move to the depth of the queue in sequence; for the measured diameter collected at that moment, find the traction speed that is lagging behind the head of the circular buffer at the depth of the queue, and form a data pair with the measured diameter collected at that moment. (3) The traction speed and measured diameter of each sampling point are time-aligned. As the production process continues, these sampling points, which are arranged in chronological order and generated continuously, form a set, which is named the time-aligned sequence.

[0031] 3. Sliding window construction To perform continuous statistical analysis on time-aligned sequences, a dynamic data queue, i.e., a sliding window, needs to be constructed within it by the main controller, following a first-in-first-out principle. The details are as follows: The sliding window is configured as a dedicated container for storing sampling points. After each data acquisition is completed, the main controller immediately performs a stack push operation, pushing the latest acquired sampling point into the entrance of the sliding window. At the same time, it checks the total number of sampling points in the current sliding window. If the number has reached the set length limit, the data at the exit of the sliding window will be automatically removed from memory and discarded. Through this mechanism, the sliding window always locks and retains the latest complete process data with a fixed time span on the production line, ensuring that subsequent calculations are always based on the current production status.

[0032] It should be noted that the length of the sliding window, i.e., the total number of sampling points that can be accommodated within the sliding window, is not an empirically fixed constant, but is determined based on the number of lag steps. In specific implementation, the main controller sets the length of the sliding window to 1.5 to 2 times the number of lag steps. If the length of the sliding window is too short, such as less than 1 times the number of lag steps, the data within the sliding window can only reflect a local segment of the production process, and the calculated results are easily affected by random noise. If the length of the sliding window is too long, too much historical data will be introduced, causing the calculated results to be averaged and unable to accurately capture minute changes in the viscosity of the molten glass. In this embodiment, the length of the sliding window is set to 2 times the number of lag steps, denoted as... .

[0033] This step, through the standardization of multi-source heterogeneous data and the timing alignment based on physical transmission lag, combined with the construction of an adaptive-length sliding window, provides a stable data foundation for subsequent calculations and improves computational accuracy.

[0034] S2: Sensitivity dual analysis.

[0035] The core of this step is to analyze the impact of changes in the traction speed of the traction motor on the diameter of the glass tube under the current process conditions at the sampling time. Specifically, based on a preset sliding window, linear regression analysis is performed on the time-aligned sequence to calculate the actual control sensitivity, and the geometric sensitivity benchmark is calculated based on the preset target diameter.

[0036] To perform accurate linear regression analysis, the main controller first needs to decenter the data within the sliding window to eliminate the influence of absolute numerical values ​​on subsequent calculations, as follows: The main controller iterates through all the sampling points stored in the sliding window at the current sampling time, and calculates the arithmetic mean of the traction speed of all sampling points in the sliding window and the arithmetic mean of the measured diameter of all sampling points, which are recorded as the speed mean and diameter mean. The actual control sensitivity relationship is as follows:

[0037] in, The actual control sensitivity; The length of the sliding window; The index of the sampling point within the sliding window, ranging from 1 to... ; For the first in the sliding window The traction speed at each sampling point; For the first in the sliding window The measured diameter at each sampling point; The average speed; This is the average diameter.

[0038] In industrial settings, laser diameter gauge readings can fluctuate at the micrometer level, and traction motor speeds can also vary. This step utilizes the least squares method to analyze the readings from noisy data. From the sampling points, the most statistically significant results were extracted, and the results were obtained. This characterizes the effect of changes in traction speed on the measured diameter at the current sampling time.

[0039] It should be noted that although this embodiment uses linear regression based on the least squares method with a sliding window to calculate the actual control sensitivity, in other embodiments, Kalman filters, recursive least squares methods, and artificial neural networks can also be used to estimate the dynamic gain relationship between traction speed and measured diameter in real time. These equivalent substitution methods are all covered within the protection scope of this invention.

[0040] In order to determine the calculated To assess the reliability of the flow rate equation and obtain the target diameter set in the process formulation, based on the law of conservation of volumetric flow rate, assuming that the cross-section of the glass tube is circular under ideal conditions and that the wall thickness scales proportionally with the diameter, a total differential derivation of the flow rate equation reveals that there exists a relationship between the relative rate of change of diameter and the relative rate of change of velocity. The theoretical coefficient relationship is as follows:

[0041] in, As a geometric sensitivity reference; The target diameter is 58 mm in this embodiment.

[0042] need This represents the theoretical derivative of the change in the diameter of the glass tube relative to the traction speed of the traction machine under ideal conditions where volumetric flow rate is conserved and thermodynamic hysteresis is ignored (i.e., no temperature disturbance and constant glass melt viscosity). It introduces the inherent nonlinear geometric inverse relationship in the glass tube drawing process, providing a standardized theoretical reference for subsequent evaluation.

[0043] It should be noted that, based on the law of conservation of volumetric flow rate, the volumetric flow rate of molten glass flowing through the forming zone is equal to the product of the cross-sectional area of ​​the glass tube and the traction speed, and this product remains constant under ideal steady state. When constructing the geometric model, it is assumed that the cross-sectional area of ​​the glass tube is approximately equal to the product of pi, the outer diameter of the glass tube, and the wall thickness of the glass tube. At the same time, based on the physical characteristics of glass tube stretching and forming, it is assumed that the wall thickness of the glass tube and the outer diameter follow a linear proportional scaling geometric relationship. From this, it can be deduced that the cross-sectional area of ​​the glass tube is proportional to the square of its outer diameter. Based on the above assumptions, the volumetric flow rate conservation equation can be simplified as follows: the product of the square of the outer diameter of the glass tube and the traction speed is equal to a constant. To obtain the response relationship between the outer diameter of the glass tube and the change in traction speed, the natural logarithm of both sides of the simplified equation is taken and the total differential is performed. The resulting mathematical relationship is: the sum of the relative change rate of twice the outer diameter and the relative change rate of the traction speed is zero. Rearranging and simplifying this relationship, we can obtain that the derivative of the outer diameter with respect to the traction speed is numerically equal to negative half multiplied by the ratio of the outer diameter to the traction speed, i.e. .

[0044] S3: Construction of thermal coupling efficiency factor and calculation of feedforward compensation amount.

[0045] Specifically, the ratio of the absolute value of the actual control sensitivity to the geometric sensitivity benchmark is calculated to obtain the thermal-rate coupling efficiency factor. At the same time, the rate of change of electrode current data is calculated, and the feedforward compensation is constructed in combination with the pre-calibrated thermal-rate conversion coefficient.

[0046] Thermal coupling efficiency factor ,when When the value approaches 1, it indicates that the current sampling time is consistent with the theoretical reference, the viscosity of the molten glass is moderate, and the control system is in an ideal rigid state; when When the value approaches 0, it indicates that at the current sampling time, there are phenomena such as increased glass melt temperature, decreased viscosity, and slippage of the traction wheel of the traction motor, which require compensation.

[0047] It should be noted that the calculations were performed under extreme operating conditions. It may approach 0, to prevent issues in subsequent steps due to... Too small a threshold can cause system oscillations. Therefore, a safe lower limit threshold is preset to determine... Is it less than the safety lower limit threshold? If so, then... Assign the value to the safety lower limit threshold; otherwise, The value remains unchanged; the lower safety threshold is calibrated through a thermal hysteresis-induced experiment, and the specific steps are as follows: 1. Define standard operating conditions: In this embodiment, a high borosilicate 3.3 glass vacuum collector tube with a nominal outer diameter of 58 mm and a wall thickness of 1.6 mm is selected as the product to be processed; the working temperature of the electric melting furnace flow channel is maintained at 1350℃±2℃, the electrode current data is stable at 1200A±10A and the fluctuation rate is less than 1% within 10 minutes; the traction speed of the traction motor is set to 3.5 m / min, the air blowing pressure in the tube is maintained at 30 to 50 Pa, and there is no mechanical slippage; the standard deviation of the data collected by the laser diameter measuring instrument is less than 0.05 mm, and the process capability index CPK>1.33; 2. After running continuously for 30 minutes under standard operating conditions, when the standard deviation of the glass tube diameter is within ±0.1 mm, calculate the thermal rate coupling efficiency factor at this time. Under standard operating conditions, the value of the thermal rate coupling efficiency factor usually fluctuates between 0.8 and 1.1, indicating that the current actual control sensitivity is highly consistent with the geometric sensitivity reference. 3. Keep the traction speed of the traction motor constant and increase the heating power of the electric melting furnace in a stepwise manner. In this embodiment, the current setting is increased by 2% every 5 minutes. As the temperature of the electric melting furnace rises, the viscosity of the glass melt gradually decreases and the glass tube becomes more flexible. At this time, the stretching effect of the traction motor on the glass tube will decrease, the actual control sensitivity will continue to decrease, and the value of the thermal coupling efficiency factor will continue to decrease. 4. After applying a speed step command of 0.1 m / min to the traction motor, if the absolute value of the change in the diameter of the glass tube collected by the laser diameter measuring instrument is less than 0.05 mm, or the absolute value of the actual control sensitivity is less than 10% of the geometric sensitivity benchmark, the value of the thermal-rate coupling efficiency factor at this time is recorded. In this embodiment, the value of the thermal-rate coupling efficiency factor at this time is 0.12. However, in order to ensure production safety, a safety margin must be reserved. Therefore, the safety lower limit threshold of this embodiment is 0.18.

[0048] It can be seen that by comprehensively calibrating the lower safety threshold through experiments, the loss of adjustment capability due to excessively high values ​​is avoided, while system oscillation under extreme conditions due to excessively low values ​​is also avoided, thus achieving a balance between production safety and equipment performance.

[0049] In the glass tube drawing process, temperature fluctuations in the electric melting furnace are a significant factor affecting the quality of the final product. However, traditional feedback control suffers from severe lag. When the power of the electric melting furnace changes, it takes time for heat to be transferred to the molten glass, for the viscosity of the molten glass to be transmitted to the forming zone, and for the diameter change of the formed glass tube to be detected by the laser diameter gauge, which also requires transmission time. Therefore, it is necessary to utilize the instantaneous nature of electrical signals to immediately generate a reverse correction command when the temperature change has just occurred and before it has caused a change in the diameter of the glass tube, in order to counteract the impact of the temperature change.

[0050] Under the Joule heating principle, changes in electrode current data are the source of heat input. Therefore, this step utilizes the rate of change of electrode current data to predict temperature changes in advance, and introduces a heat rate conversion coefficient to calculate the feedforward compensation amount, specifically: To avoid interference from high-frequency electromagnetic noise in the current signal, instead of directly using two adjacent sampling points for differential analysis, a filter window is set, and the rate of change of the electrode current data within this filter window is calculated. This represents the rate of change of the electrode current data at the current sampling moment; This is the electrode current data at the current sampling time; For the preset filtering window, this embodiment uses 8 sampling points; The sampling period.

[0051] The electrode current data changes the viscosity of the molten glass, while the traction speed of the traction motor changes the geometric stretch ratio of the glass tube. These two physical quantities do not have a direct mathematical relationship. Therefore, it is necessary to construct a thermal rate conversion coefficient to convert the rate of change of the electrode current data into the speed adjustment amount that the traction motor needs to perform. It should be noted that the thermal rate conversion coefficient is comprehensively calibrated through an open-loop step experiment. The specific steps are as follows: 1. Using the standard operating conditions defined in the calibration experiment of the lower safety limit threshold, run continuously for 30 minutes under the standard operating conditions. When the standard deviation of the glass tube diameter is within ±0.1mm, record the electrode current data and traction speed at this time, and record them as the reference current and reference traction speed. 2. Keeping the traction speed constant, modify the power of the electric melting furnace to instantly increase the electrode current data by 50A, denoted as... Continuously observe and record the diameter changes of the glass tube as reported by the laser diameter gauge, and record the amount of diameter change when the diameter of the glass tube stabilizes, denoted as . In this embodiment, the diameter of the glass tube was measured to be 0.8 mm smaller. 3. Restore the production line to the reference current and reference traction speed state, keep the electrode current data constant, and adjust the traction speed until the change in the glass tube diameter is... At this time, record the change in traction speed, denoted as . In this embodiment, the traction speed was measured to have decreased by 0.1 m / min. With The ratio is the value of the heat conversion rate coefficient. In this embodiment, the heat conversion rate coefficient is measured. .

[0052] Thus, the product of the thermal conversion coefficient and the rate of change of the electrode current data is the feedforward compensation amount.

[0053] S4: Instruction generation and closed-loop control.

[0054] Specifically, the thermal-speed coupling efficiency factor is used to dynamically compensate the PID control gain, and at the same time, the feedforward compensation is combined to generate an adaptive speed control command, which in turn drives the traction motor to make relevant adjustments.

[0055] The formula for adaptive speed control commands is as follows:

[0056] in, This is an adaptive speed control command; The reference speed is obtained from the process formulation; in this embodiment, it is taken as... ; This refers to the thermal coupling efficiency factor. , These are preset proportional parameters and integral parameters, respectively. In this embodiment, we take... , ; This represents the deviation between the measured diameter and the target diameter at the sampling point at the current sampling time. Item for Historical accumulation and; The term represents the feedforward compensation amount.

[0057] It should be noted that, , The specific values ​​are merely preferred empirical values ​​for this embodiment. In actual engineering applications, those skilled in the art can adjust and optimize the above parameters accordingly using the critical ratio method based on specific traction motor power characteristics, glass tube specifications, and other data.

[0058] right After S-shaped acceleration and deceleration smoothing, the signal is sent to the frequency converter via the bus. The frequency converter drives the traction motor to change the stretching ratio, ultimately achieving intelligent control of the traction speed of the glass tube drawing machine.

[0059] Figure 2 This is a comparison chart of the diameter control effect of the method provided in this embodiment of the invention and the existing PID control technology when facing thermal interference. As shown in the figure, the existing technology uses fixed PID parameters, which cannot adapt to changes in glass viscosity, resulting in lag in traction speed adjustment, large fluctuations in the diameter of the glass tube, and long-term deviations from the tolerance range. In contrast, the present invention can quickly suppress fluctuations and rapidly pull the diameter of the glass tube back to near the target value.

[0060] Figure 3The figure shows the linkage change curve between the thermal speed coupling efficiency factor and the traction speed command provided in the embodiment of the present invention. When interference occurs, the value of the thermal speed coupling efficiency factor decreases. At this time, the PID controller reacts and quickly increases the traction speed to compensate, thereby stabilizing the production process.

Claims

1. An intelligent control method for the traction speed of a solar glass tube drawing machine, characterized in that, include: The actual speed sequence of the traction motor, the measured diameter sequence of the glass tube, and the electrode current data of the electric melting furnace are collected in real time by a preset multi-dimensional data acquisition unit and preprocessed to obtain a time-aligned sequence containing several continuous sampling points. A sliding window is preset, and linear regression analysis is performed on the time-aligned sequence based on the sliding window to calculate the actual control sensitivity. At the same time, the geometric sensitivity benchmark is calculated based on the preset target diameter. Based on the actual control sensitivity and the geometric sensitivity benchmark, the thermal rate coupling efficiency factor is calculated, and the rate of change of the electrode current data is calculated. The feedforward compensation amount is constructed by combining the pre-calibrated thermal rate conversion coefficient. The PID control gain is dynamically compensated based on the thermal-speed coupling efficiency factor, and an adaptive speed control command is generated by combining the feedforward compensation amount, thereby driving the traction motor to make relevant adjustments.

2. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The actual control sensitivity satisfies the following relationship: in, The actual control sensitivity; The length of the sliding window; The index of the sampling point within the sliding window ranges from 1 to... ; For the first in the sliding window The traction speed of each of the sampling points; For the first in the sliding window The measured diameter of each of the aforementioned sampling points; The arithmetic mean of the traction velocities of all the sampling points within the sliding window; It is the arithmetic mean of the measured diameters of all the sampling points within the sliding window.

3. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The geometric sensitivity benchmark is calculated based on the law of conservation of volumetric flow rate, combined with the arithmetic mean of the target diameter and the traction velocity of all the sampling points within the sliding window.

4. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The adaptive speed control command satisfies the following relationship: in, This refers to the adaptive speed control command; As the reference speed; This refers to the thermal coupling efficiency factor; , These are the preset proportional parameters and integral parameters, respectively; The deviation between the measured diameter of the current sampling point and the target diameter in the time-aligned sequence; The term represents the historical cumulative sum of the aforementioned deviations; The heat conversion rate coefficient is... The rate of change of the electrode current data; The term refers to the feedforward compensation amount.

5. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The thermal conversion rate coefficient was determined through a comprehensive open-loop step experiment.

6. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The preprocessing includes: unifying the length measurement units of the collected data, and performing time-series alignment between the actual velocity sequence and the measured diameter sequence to obtain the time-series aligned sequence.

7. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The acquisition of the actual speed sequence of the traction motor includes: reading the real-time operating frequency of the traction motor through the communication interface of the inverter driving the traction motor, and converting the real-time operating frequency into linear speed.

8. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The multidimensional data acquisition unit includes: a traction motor encoder for acquiring the actual speed sequence, a laser diameter gauge for acquiring the measured diameter sequence, and a current transformer for acquiring the electrode current data.

9. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The length of the sliding window is determined based on the transmission lag time of the traction motor during the traction process.

10. The intelligent control method for the traction speed of the solar glass tube drawing machine according to claim 1, characterized in that, The method further includes: setting a safety lower limit threshold, determining whether the thermal coupling efficiency factor is less than the safety lower limit threshold, if so, assigning the thermal coupling efficiency factor to the safety lower limit threshold, if not, the value of the thermal coupling efficiency factor remains unchanged.

Images