Online detection method for plastic film blow molding production quality

Abstract

The invention belongs to the technical field of quality detection, and discloses a plastic film blow molding production quality on-line detection method, which adopts a melt rheology detection device to detect viscosity, shear rate and temperature in real time, combines a mathematical model to calculate a flow index and a viscosity error, and adjusts extrusion temperature or shear rate according to a detection result. The influence of rheological fluctuation on the final thickness is reduced from the melt stage; the temperature and the initial thickness of the melt are measured through an infrared detection device, thickness data of transverse TD and longitudinal MD are obtained in combination with a laser thickness measurement device, the thickness deviation is calculated, the die gap is adjusted, and the initial film thickness is stabilized; the temperature distribution of the film bubbles is obtained through an infrared detection device, and the thickness of the film bubbles is measured by combining a laser thickness measuring device; an ultrasonic thickness measuring device is used for measuring the thickness of the thin film, a CNN (convolutional neural network) is used for extracting thickness features and predicting thickness distribution, the traction rate or air ring air pressure is adjusted in combination with detection data, and the uniformity of the thickness of a final product is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of quality detection, and in particular relates to an online detection method for the production quality of plastic film blow molding. Background Art

[0002] In the blow molding production process of plastic film, optical thickness measurement, X-ray thickness measurement, infrared sensor detection and ultrasonic thickness measurement are mostly used to detect the thickness of the finished product. These detection methods are mainly concentrated in the finished product stage.

[0003] For example, the Chinese utility model patent with patent application number: CN201520675848.4 discloses a production equipment for laser measurement and control of plastic film thickness, including a screw extruder, a screw extrusion motor driving the screw extruder, a film forming device arranged at the outlet end of the screw extruder, a film forming motor driving the film forming device, and a controller; the outlet of the film forming device is sequentially provided with a first stretching roller group, a first oven, a second stretching roller group, a second oven and a film rolling roller; the outlet end of the film forming device is provided with a laser detector for measuring the thickness of the plastic film, the rollers of the first stretching roller group are connected by a pulley mechanism and driven by a first motor, and the rollers of the second stretching roller group are connected by a pulley mechanism and driven by a second motor; the screw extrusion motor, the film forming motor, the first motor, the second motor and the laser detector are electrically connected to the controller.

[0004] The above-mentioned existing plastic film production equipment of this type uses a laser detector to detect the thickness of the plastic film at the outlet end of the film forming device in real time. However, in the actual blow molding production process, uneven thickness, local thinning or thickening may occur. The above-mentioned traditional detection methods can only discover the problem after the fact, but cannot issue an early warning or adjust the process parameters in the early stage of the problem. The production adjustment process has a slow response speed, which affects the consistency of product quality. In addition, when the production parameters are unstable, the existing solutions are difficult to identify the root cause of the problem and can only compensate by adjusting the cooling or extrusion speed, and the adjustment effect is unstable.

[0005] The above problems are particularly prominent in the blow molding production process of biodegradable films, such as PLA and PBAT. Compared with traditional petroleum-based plastics, the melting temperature range of biodegradable plastic film raw materials is narrower, and the rheological properties are more affected by temperature and shear rate. A slight fluctuation may lead to uneven melt flow, thereby affecting the thickness distribution of the film; however, existing online thickness detection methods, such as X-rays and laser thickness gauges, can only measure the thickness of the final product and cannot perceive changes in the melt state in advance. Therefore, when the melt rheology is abnormal, it can only be detected after the film is formed. At this time, it is too late to adjust the production parameters, which not only leads to an increase in the scrap rate, but also increases production costs; therefore, there is an urgent need for an online detection method for the quality of plastic film blow molding production to solve such problems. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide an online detection method for the production quality of plastic film blow molding, which is used to solve the limitations of traditional online detection solutions, namely, detection lag, inability to detect the melt state in real time, lack of early warning and process adjustment capabilities, resulting in difficulty in controlling the thickness fluctuation of the finished plastic film.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions: A method for online detection of plastic film blow molding production quality, comprising the following steps: Step S1, melting and extruding the biodegradable plastic film raw material, detecting the melt rheology data in real time by a melt rheology detection device, and adjusting the extrusion temperature or shear rate according to the detection result; Step S2, conveying the melt to the die adjustment device, using an infrared detection device and a laser thickness measuring device to measure the initial film thickness after the melt is extruded, obtaining thickness detection data, and adjusting the die gap based on the rheological data and the thickness detection data; Step S3, after the melt is extruded through the die to form a film bubble, an infrared detection device and a laser thickness measuring device are used to measure the thickness of the film bubble, and rheological data are compared, and the air ring cooling rate or the pulling rate is adjusted based on the comparison results; Step S4, during the stretching and cooling process of the film, an ultrasonic thickness measuring device is used to measure the thickness of the film, and a thickness distribution analysis model based on a convolutional neural network (CNN) is used to calculate thickness distribution data, and based on the film thickness data measured by the ultrasonic thickness measuring device and the analysis results of the CNN thickness distribution analysis model, the traction rate or the wind ring wind pressure is adjusted; Step S5, after the film is formed, the thickness is measured using an optical thickness measuring device and an infrared detection device, and the extruder temperature, shear rate or pulling rate are adjusted based on the measurement data.

[0008] The following is a further optimization of the above technical solution by the present invention: The rheological data include viscosity, shear rate and temperature.

[0009] Further optimization: In step S1, the steps of real-time detection of melt rheology data and adjustment of extrusion temperature or shear rate are as follows: The melt rheology detection device is used to measure the melt flow characteristics in real time. Collect the shear rate of the melt and viscosity , and combined with the temperature measured by the temperature sensor Perform data analysis and calculate melt flow index and consistency index , characterizing the rheological properties: , in, represents the melt viscosity, Represents the consistency index, which measures the flow resistance of the molten material. is the shear rate, represents the flow index; Calculate the deviation of rheological data : , in, Indicates the set target viscosity, Indicates viscosity error; like , increase the extrusion temperature To reduce viscosity, adjust the formula to: , like , reduce the extrusion temperature Or reduce the shear rate , the adjustment formula is: , in, is the adjusted temperature, is the adjusted shear rate, is the temperature adjustment coefficient, is the shear rate adjustment factor, calculated as: , in, is the maximum allowable temperature, is the minimum allowable temperature, Indicates the maximum acceptable viscosity, Indicates the minimum acceptable viscosity, represents the maximum acceptable shear rate, Indicates the minimum acceptable shear rate.

[0010] Further optimization: The steps of measuring the initial film thickness after melt extrusion using an infrared detection device and a laser thickness measuring device are as follows: Using infrared detection device to measure melt temperature and melt thickness , and record the measurement data, The laser thickness measuring device is used to scan the extruded initial film to obtain the thickness data in the transverse direction TD and the longitudinal direction MD: , in, For horizontal position and vertical position The thickness of is the thickness deviation at this position, and are the coordinates of the measurement point on the two-dimensional plane.

[0011] Further optimization: The steps of adjusting the die gap based on the rheological data and thickness detection data are as follows: Calculate the thickness error: , like , adjust the die gap , the adjustment formula is: , in, Indicates the allowable tolerance, is the adjusted die gap, is the gap adjustment factor.

[0012] Further optimization: The step of adjusting the air ring cooling rate or traction rate based on the comparison result is: Use infrared detection device to obtain the surface temperature of the film bubble and its temperature distribution data, Use laser thickness measuring device to obtain film bubble thickness , measuring different positions Thickness at: , in, Indicates that the membrane bubble is in position The thickness of is the initial bubble thickness, is the thickness deviation at this position; Calculate the bubble thickness error using the formula: , in, is the bubble thickness error, is the target thickness; Calculate the rate of change of bubble thickness: , in, Indicates the rate of change of film bubble thickness, is the thickness adjustment coefficient; like , increase the air ring cooling rate , the adjustment formula is: , like , reduce the air ring cooling rate or increase the traction rate , the adjustment formula is: , in, is the adjusted air ring cooling rate, is the adjusted traction rate, is the cooling rate adjustment factor, is the traction rate adjustment factor.

[0013] Further optimization: The steps of calculating thickness distribution data by the thickness distribution analysis model based on convolutional neural network CNN are: Ultrasonic thickness gauge to measure film thickness : , in, For location The film thickness measured at is the base film thickness, is the thickness deviation at this position; The sliding window method is used to extract local thickness data: , in, For the A local thickness data window, and is the window size, and is the starting position of the window; A three-layer convolutional CNN is used to calculate the thickness feature map: , , , in, , , is the layer-by-layer thickness feature extracted by CNN. , , is the convolution kernel of each layer of CNN, , , is the bias for each layer, is the activation function; The final thickness distribution prediction is calculated using the fully connected layer: , in, is the film thickness distribution predicted by CNN, is the weight matrix of the fully connected layer, is the bias term of the fully connected layer.

[0014] Further optimization: Based on the film thickness data measured by the ultrasonic thickness measuring device and the analysis results of the CNN thickness distribution analysis model, the steps of adjusting the traction rate or the wind ring wind pressure are as follows: Film thickness measured with ultrasonic thickness gauge And CNN predicts thickness , calculate the thickness error: , in, is the thickness error, For location The film thickness measured at The film thickness distribution predicted by CNN; Calculate the local average thickness error: , in, For Window The average thickness error at is the number of measurement points within the window; Calculate the adjusted pulling rate: like , increase the pulling rate , the adjustment formula is: , like , reduce the traction rate, and the adjustment formula is: , in, is the adjusted traction rate, is the traction rate adjustment factor.

[0015] Further optimization: the step of adjusting the traction rate or the wind ring wind pressure also includes: Calculate the adjusted wind ring pressure: like , reduce the wind pressure of the wind ring , the adjustment formula is: , like , increase the wind ring wind pressure, the adjustment formula is: , in, is the adjusted wind ring wind pressure, is the wind pressure adjustment coefficient of the wind ring.

[0016] Further optimization: The steps of measuring the thickness using an optical thickness measuring device and an infrared detection device, and adjusting the extruder temperature, shear rate or pulling rate based on the measurement data are as follows: The final thickness of the film is measured by an optical thickness measuring device, and the uniformity data of the film thickness is obtained; Use infrared detection device to detect the surface temperature distribution of the film to determine whether there is an abnormal temperature area; Calculate the uniformity of thickness data and make adjustments if the thickness deviation exceeds the set tolerance range; Combined with the data from the infrared detection device, the temperature distribution of the film is analyzed to determine the correlation between temperature and thickness; If the overall thickness of the film is too large or too small, adjust the extruder temperature according to the temperature data of the infrared detection device; If the film thickness varies greatly in different areas, adjust the shear rate to optimize melt flow; If there is a systematic error in the film thickness in the longitudinal direction, adjust the pulling rate.

[0017] The present invention adopts the above technical solution and has the following beneficial effects: 1. The present invention adopts a melt rheology detection device to detect the viscosity, shear rate and temperature of the melt state of the degradable plastic film raw material in real time, and calculates the flow index and viscosity error in combination with a mathematical model, adjusts the extrusion temperature or shear rate according to the detection results, and reduces the influence of rheological fluctuations on the thickness of the final product from the melt stage; measures the melt temperature and initial thickness through an infrared detection device, and obtains the thickness data of the transverse TD and longitudinal MD in combination with a laser thickness measuring device, calculates the thickness deviation and adjusts the die gap to stabilize the initial film thickness; obtains the bubble temperature distribution through an infrared detection device, measures the bubble thickness in combination with a laser thickness measuring device, and adjusts the air ring cooling rate or traction rate based on the calculated thickness error to maintain the stability of the bubble morphology and thickness.

[0018] 2. The present invention adopts an ultrasonic thickness measuring device to measure the thickness of the film, and uses a convolutional neural network (CNN) to extract thickness features, predict thickness distribution, and adjust the traction rate or wind ring pressure in combination with the detection data to improve the uniformity of product thickness; an optical thickness measuring device and an infrared detection device are combined to measure the final thickness, and the extruder temperature, shear rate or traction rate are adjusted based on the measurement data.

[0019] The present invention is further described below in conjunction with the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a schematic diagram of the process of the online detection method of the plastic film blow molding production quality in the present invention. DETAILED DESCRIPTION

[0021] Reference Figure 1 , a method for online detection of plastic film blow molding production quality, comprising the following steps: Step S1, melting and extruding the biodegradable plastic film raw material, detecting the melt rheology data in real time by a melt rheology detection device, and adjusting the extrusion temperature or shear rate according to the detection result.

[0022] The rheological data include viscosity, shear rate and temperature.

[0023] In step S1, the steps of detecting melt rheological data in real time and adjusting the extrusion temperature or shear rate are as follows: The melt rheology detection device is used to measure the melt flow characteristics in real time. Collect the shear rate of the melt and viscosity , and combined with the temperature measured by the temperature sensor Perform data analysis and calculate melt flow index and consistency index , characterizing the rheological properties: , in, represents the melt viscosity, Represents the consistency index, which measures the flow resistance of the molten material. is the shear rate, Represents the flow index.

[0024] Calculate the deviation of rheological data : , in, Indicates the set target viscosity, Indicates viscosity error; like , increase the extrusion temperature To reduce viscosity, adjust the formula to: , like , reduce the extrusion temperature Or reduce the shear rate , the adjustment formula is: , in, is the adjusted temperature, is the adjusted shear rate, is the temperature adjustment coefficient, is the shear rate adjustment factor, calculated as: , in, is the maximum allowable temperature, is the minimum allowable temperature, Indicates the maximum acceptable viscosity, Indicates the minimum acceptable viscosity, represents the maximum acceptable shear rate, Indicates the minimum acceptable shear rate.

[0025] Specifically, parameter detection is performed through a melt rheology detection device, and a mathematical model is used to calculate the viscosity error to guide the adjustment of the extrusion temperature and shear rate of the extruder.

[0026] Step S2, conveying the melt to the die adjustment device, using an infrared detection device and a laser thickness measuring device to measure the initial film thickness after the melt is extruded, obtaining thickness detection data, and adjusting the die gap based on the rheological data and thickness detection data.

[0027] The steps of measuring the initial film thickness after melt extrusion using an infrared detection device and a laser thickness measuring device are as follows: Using infrared detection device to measure melt temperature and melt thickness , and record the measurement data, The laser thickness measuring device is used to scan the extruded initial film to obtain the thickness data in the transverse direction TD and the longitudinal direction MD: , in, For horizontal position and vertical position The thickness of is the thickness deviation at this position, and are the coordinates of the measurement point on the two-dimensional plane.

[0028] The steps of adjusting the die gap based on the rheological data and the thickness detection data are as follows: Calculate the thickness error: , like , adjust the die gap , the adjustment formula is: , in, Indicates the allowable tolerance, is the adjusted die gap, is the gap adjustment factor.

[0029] Specifically, an infrared detection device is used to measure the melt temperature, and a laser thickness measuring device is used to obtain the initial thickness data of the extruded film. The film thickness error is calculated, and the die gap is adjusted based on the thickness deviation in the transverse TD and longitudinal MD directions. The initial film thickness is made to meet the target value, providing a stable benchmark for subsequent bubble formation and stretching processes, thereby reducing thickness deviations in production.

[0030] Step S3, the melt is extruded through a die to form a film bubble, and the film bubble thickness is measured by an infrared detection device and a laser thickness measuring device, and the rheological data are compared, and the air ring cooling rate or the pulling rate is adjusted based on the comparison result.

[0031] The steps for adjusting the air ring cooling rate or traction rate based on the comparison results are as follows: Use infrared detection device to obtain the surface temperature of the film bubble and its temperature distribution data, Use laser thickness measuring device to obtain film bubble thickness , measuring different positions Thickness at: , in, Indicates that the membrane bubble is in position The thickness of is the initial bubble thickness, is the thickness deviation at this position; Calculate the bubble thickness error using the formula: , in, is the bubble thickness error, is the target thickness; Calculate the rate of change of bubble thickness: , in, Indicates the rate of change of film bubble thickness, is the thickness adjustment coefficient; like , increase the air ring cooling rate , the adjustment formula is: , like , reduce the air ring cooling rate or increase the traction rate , the adjustment formula is: , in, is the adjusted air ring cooling rate, is the adjusted traction rate, is the cooling rate adjustment factor, is the traction rate adjustment factor.

[0032] Specifically, in this embodiment, the thickness of the bubble directly determines the mechanical properties and uniformity of the final film. The bubble thickness is measured in real time using infrared and laser thickness measuring devices, and the thickness error is calculated in combination with the melt rheology data to adjust the air ring cooling rate or the pulling rate. If the bubble thickness is too large, the cooling efficiency will be reduced, so the air ring rate is increased. If the bubble thickness is too small, the mechanical properties of the film may decrease. So the air ring speed is reduced or the pulling rate is increased for optimization. This ensures that the bubble thickness is uniform and stable, thereby improving the quality of the final product.

[0033] Step S4: During the stretching and cooling process of the film, an ultrasonic thickness gauge is used to measure the thickness of the film, and the thickness distribution data is calculated in combination with a thickness distribution analysis model based on a convolutional neural network (CNN). Based on the film thickness data measured by the ultrasonic thickness gauge and the analysis results of the CNN thickness distribution analysis model, the traction rate or the wind ring pressure is adjusted.

[0034] The steps for calculating thickness distribution data based on the thickness distribution analysis model of convolutional neural network CNN are as follows: Ultrasonic thickness gauge to measure film thickness : , in, For location The film thickness measured at is the base film thickness, is the thickness deviation at this position; The sliding window method is used to extract local thickness data: , in, For the A local thickness data window, and is the window size, and is the starting position of the window; A three-layer convolutional CNN is used to calculate the thickness feature map: , , , in, , , is the layer-by-layer thickness feature extracted by CNN. , , is the convolution kernel of each layer of CNN, , , is the bias for each layer, is the activation function; The final thickness distribution prediction is calculated using the fully connected layer: , in, is the film thickness distribution predicted by CNN, is the weight matrix of the fully connected layer, is the bias term of the fully connected layer.

[0035] Specifically, a three-layer convolutional CNN is used to perform thickness distribution analysis to improve the thickness uniformity of the predicted adjustment film. In this step S4, a local thickness distribution data set is constructed based on the thickness data measured by the ultrasonic thickness measuring device, and features are extracted through a sliding window. After the thickness distribution features are extracted using a multi-layer convolutional network, the final thickness prediction is performed in combination with a fully connected layer to effectively reduce the impact of noise.

[0036] Based on the film thickness data measured by the ultrasonic thickness measuring device and the analysis results of the CNN thickness distribution analysis model, the steps for adjusting the traction rate or the wind ring wind pressure are as follows: Film thickness measured with ultrasonic thickness gauge And CNN predicts thickness , calculate the thickness error: , in, is the thickness error, For location The film thickness measured at The film thickness distribution predicted by CNN; Calculate the local average thickness error: , in, For Window The average thickness error at is the number of measurement points within the window; Calculate the adjusted pulling rate: like , increase the pulling rate , the adjustment formula is: , like , reduce the traction rate, and the adjustment formula is: , in, is the adjusted traction rate, is the traction rate adjustment factor.

[0037] The steps of adjusting the traction rate or the wind ring wind pressure also include: Calculate the adjusted wind ring pressure: like , reduce the wind pressure of the wind ring , the adjustment formula is: , like , increase the wind ring wind pressure, the adjustment formula is: , in, is the adjusted wind ring wind pressure, is the wind pressure adjustment coefficient of the wind ring.

[0038] Specifically, the traction rate or wind ring pressure is adjusted based on the ultrasonic thickness measuring device and CNN analysis results. After obtaining the thickness distribution predicted by CNN, the thickness error is calculated, and the traction rate or wind ring pressure is adjusted accordingly. When the thickness is too large, compensation is made by increasing the traction rate or reducing the wind ring pressure. When the thickness is too small, the traction rate is reduced or the wind ring pressure is increased to optimize the thickness uniformity. Traditional ultrasonic measurement data and AI prediction are combined to improve the stability of thickness control and reduce manual intervention.

[0039] Step S5, after the film is formed, the thickness is measured using an optical thickness measuring device and an infrared detection device, and the extruder temperature, shear rate or pulling rate are adjusted based on the measurement data.

[0040] The steps of measuring the thickness using an optical thickness measuring device and an infrared detection device and adjusting the extruder temperature, shear rate or pulling rate based on the measured data are as follows: The final thickness of the film is measured by an optical thickness measuring device, and the uniformity data of the film thickness is obtained; Use infrared detection device to detect the surface temperature distribution of the film to determine whether there is an abnormal temperature area; Calculate the uniformity of thickness data and make adjustments if the thickness deviation exceeds the set tolerance range; Combined with the data from the infrared detection device, the temperature distribution of the film is analyzed to determine the correlation between temperature and thickness; If the overall thickness of the film is too large or too small, adjust the extruder temperature according to the temperature data of the infrared detection device; If the film thickness varies greatly in different areas, adjust the shear rate to optimize melt flow; If there is a systematic error in the film thickness in the longitudinal direction, adjust the pulling rate.

[0041] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for online detection of plastic film blow molding production quality, characterized in that: The steps include: Step S1, melting and extruding the biodegradable plastic film raw material, detecting the melt rheology data in real time by a melt rheology detection device, and adjusting the extrusion temperature or shear rate according to the detection result; Step S2, conveying the melt to the die adjustment device, using an infrared detection device and a laser thickness measuring device to measure the initial film thickness after the melt is extruded, obtaining thickness detection data, and adjusting the die gap based on the rheological data and the thickness detection data; Step S3, after the melt is extruded through the die to form a film bubble, an infrared detection device and a laser thickness measuring device are used to measure the thickness of the film bubble, and rheological data are compared, and the air ring cooling rate or the pulling rate is adjusted based on the comparison results; Step S4, during the stretching and cooling process of the film, an ultrasonic thickness measuring device is used to measure the thickness of the film, and a thickness distribution analysis model based on a convolutional neural network (CNN) is used to calculate thickness distribution data, and based on the film thickness data measured by the ultrasonic thickness measuring device and the analysis results of the CNN thickness distribution analysis model, the traction rate or the wind ring wind pressure is adjusted; Step S5, after the film is formed, the thickness is measured using an optical thickness measuring device and an infrared detection device, and the extruder temperature, shear rate or pulling rate are adjusted based on the measurement data.

2. A method for online detection of plastic film blow molding production quality according to claim 1, characterized in that: The rheological data include viscosity, shear rate and temperature.

3. A method for online detection of plastic film blow molding production quality according to claim 2, characterized in that: In step S1, the steps of detecting melt rheological data in real time and adjusting the extrusion temperature or shear rate are as follows: The melt rheology detection device is used to measure the melt flow characteristics in real time. Collect the shear rate of the melt and viscosity , and combined with the temperature measured by the temperature sensor Perform data analysis and calculate melt flow index and consistency index , characterizing the rheological properties: ， in, represents the melt viscosity, Represents the consistency index, which measures the flow resistance of the molten material. is the shear rate, represents the flow index; Calculate the deviation of rheological data : ， in, Indicates the set target viscosity, Indicates viscosity error; like , increase the extrusion temperature To reduce viscosity, adjust the formula to: ， like , reduce the extrusion temperature Or reduce the shear rate , the adjustment formula is: ， in, is the adjusted temperature, is the adjusted shear rate, is the temperature adjustment coefficient, is the shear rate adjustment factor, calculated as: ， in, is the maximum allowable temperature, is the minimum allowable temperature, Indicates the maximum acceptable viscosity, Indicates the minimum acceptable viscosity, represents the maximum acceptable shear rate, Indicates the minimum acceptable shear rate.

4. A method for online detection of plastic film blow molding production quality according to claim 3, characterized in that: The steps of measuring the initial film thickness after melt extrusion by using an infrared detection device and a laser thickness measuring device are as follows: Using infrared detection device to measure melt temperature and melt thickness , and record the measurement data, The laser thickness measuring device is used to scan the extruded initial film to obtain the thickness data in the transverse direction TD and the longitudinal direction MD: ， in, For horizontal position and vertical position The thickness of is the thickness deviation at this position, and are the coordinates of the measurement point on the two-dimensional plane.

5. A method for online detection of plastic film blow molding production quality according to claim 4, characterized in that: The steps of adjusting the die gap based on the rheological data and the thickness detection data are as follows: Calculate the thickness error: ， like , adjust the die gap , the adjustment formula is: ， in, Indicates the allowable tolerance, is the adjusted die clearance, is the gap adjustment factor.

6. A method for online detection of plastic film blow molding production quality according to claim 5, characterized in that: The step of adjusting the air ring cooling rate or traction rate based on the comparison result is: Use infrared detection device to obtain the surface temperature of the film bubble and its temperature distribution data, Use laser thickness measuring device to obtain film bubble thickness , measuring different positions Thickness at: ， in, Indicates that the membrane bubble is in position The thickness of is the initial bubble thickness, is the thickness deviation at this position; Calculate the bubble thickness error using the formula: ， in, is the bubble thickness error, is the target thickness; Calculate the rate of change of bubble thickness: ， in, Indicates the rate of change of film bubble thickness, is the thickness adjustment coefficient; like , increase the air ring cooling rate , the adjustment formula is: ， like , reduce the air ring cooling rate or increase the traction rate , the adjustment formula is: ， in, is the adjusted air ring cooling rate, is the adjusted traction rate, is the cooling rate adjustment factor, is the traction rate adjustment factor.

7. A method for online detection of plastic film blow molding production quality according to claim 6, characterized in that: The steps of calculating thickness distribution data based on the thickness distribution analysis model of the convolutional neural network CNN are as follows: Ultrasonic thickness gauge to measure film thickness : ， in, For location The film thickness measured at is the base film thickness, is the thickness deviation at this position; The sliding window method is used to extract local thickness data: ， in, For the A local thickness data window, and is the window size, and is the starting position of the window; A three-layer convolutional CNN is used to calculate the thickness feature map: ， ， ， in, , , is the layer-by-layer thickness feature extracted by CNN. , , is the convolution kernel of each layer of CNN, , , is the bias for each layer, is the activation function; The final thickness distribution prediction is calculated using the fully connected layer: ， in, is the film thickness distribution predicted by CNN, is the weight matrix of the fully connected layer, is the bias term of the fully connected layer.

8. A method for online detection of plastic film blow molding production quality according to claim 7, characterized in that: The steps of adjusting the traction rate or the wind ring wind pressure based on the film thickness data measured by the ultrasonic thickness measuring device and the analysis results of the CNN thickness distribution analysis model are as follows: Film thickness measured with ultrasonic thickness gauge And CNN predicts thickness , calculate the thickness error: ， in, is the thickness error, For location The film thickness measured at The film thickness distribution predicted by CNN; Calculate the local average thickness error: ， in, For Window The average thickness error at is the number of measurement points within the window; Calculate the adjusted pulling rate: like , increase the pulling rate , the adjustment formula is: ， like , reduce the traction rate, and the adjustment formula is: ， in, is the adjusted traction rate, is the traction rate adjustment factor.

9. A method for online detection of plastic film blow molding production quality according to claim 8, characterized in that: The step of adjusting the traction rate or the wind ring wind pressure also includes: Calculate the adjusted wind ring pressure: like , reduce the wind pressure of the wind ring , the adjustment formula is: ， like , increase the wind ring wind pressure, the adjustment formula is: ， in, is the adjusted wind ring wind pressure, is the wind pressure adjustment coefficient of the wind ring.

10. A method for online detection of plastic film blow molding production quality according to claim 9, characterized in that: The steps of measuring the thickness by using an optical thickness measuring device and an infrared detection device, and adjusting the extruder temperature, shear rate or pulling rate based on the measured data are as follows: The final thickness of the film is measured by an optical thickness measuring device, and the uniformity data of the film thickness is obtained; Use infrared detection device to detect the surface temperature distribution of the film to determine whether there is an abnormal temperature area; Calculate the uniformity of thickness data and make adjustments if the thickness deviation exceeds the set tolerance range; Combined with the data from the infrared detection device, the temperature distribution of the film is analyzed to determine the correlation between temperature and thickness; If the overall thickness of the film is too large or too small, adjust the extruder temperature according to the temperature data of the infrared detection device; If the film thickness varies greatly in different areas, adjust the shear rate to optimize melt flow; If there is a systematic error in the film thickness in the longitudinal direction, adjust the pulling rate.

Images