Injection gas dose and pressure coordinated optimization density uniformity control method and system

Abstract

The application discloses a density uniform control method and system based on injection gas dose and pressure collaborative optimization, and relates to the technical field of injection control; the method comprises the following steps: collecting process parameters in real time; preprocessing the process parameters, and constructing a virtual cavity model based on the process parameters; carrying out defect prediction based on the virtual cavity model to obtain a defect area; generating an optimization decision scheme by comprehensively considering the defect area and corresponding process parameters; obtaining correlation coefficients between the process parameters by using a pre-constructed multivariate coupling model; calculating strategies for different defect types based on the correlation coefficients between the process parameters and the optimization decision scheme, to obtain adjustment strategies for the process parameters; and the application can effectively inhibit silver streaks and shrinkage defects, improve the surface quality, size accuracy and mechanical property stability of products, reduce the waste rate and rework cost, and simultaneously improve the injection production efficiency and product quality consistency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of injection molding control, more particularly, the present application relates to a density uniformity control method and system based on injection gas dosage and pressure collaborative optimization. BACKGROUND

[0002] In the injection molding production of complex structural parts such as automobile parts, in order to achieve the balance between lightweight and structural strength of the product, chemical or physical foaming process is often used to reduce the material density and optimize the forming performance by introducing bubbles into the molten plastic. In the prior art, the control of the injection foaming process mainly relies on preset process parameters and offline monitoring means. Its technical principle is to set the injection gas addition ratio, holding pressure, back pressure value and cooling water temperature and other parameters based on historical production experience, collect local data in the cavity through a single pressure sensor or temperature sensor, and then make simple feedback adjustment according to fixed threshold.

[0003] However, the prior art has significant limitations. Since the bubble nucleation and bubble growth process is affected by multiple factors such as material properties, melt flow state and mold temperature field, the existing monitoring means cannot capture the dynamic changes of bubble density distribution and radial growth of bubbles in real time, resulting in a lag in parameter optimization. At the same time, the control of injection gas dosage, holding pressure and cooling gradient is independent of each other, and no collaborative mechanism is formed. When the melt viscosity fluctuates or the cavity is not filled uniformly, it is easy to cause imbalance of bubble distribution. The local bubble density is too high, which will cause the melt strength to decrease and form silver lines at the stress concentration; and the bubble growth rate is not synchronized with the cooling shrinkage, which will cause shrinkage depression in the thick wall area of the product. Such defects not only reduce the mechanical properties and structural stability of the product, but also cause insufficient assembly precision of the parts, increase the waste rate and rework cost, and seriously restrict the production efficiency and product quality stability.

[0004] In view of this, the present application proposes a density uniformity control method and system based on injection gas dosage and pressure collaborative optimization to solve the above problems. SUMMARY

[0005] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purposes, the present application provides the following technical scheme: a density uniformity control method based on injection gas dosage and pressure collaborative optimization, comprising:

[0006] Real-time acquisition of process parameters;

[0007] Pretreatment of process parameters, and construction of a virtual cavity model based on process parameters; defect prediction based on the virtual cavity model to obtain a defect area; comprehensive defect area and corresponding process parameters to generate an optimization decision scheme;

[0008] Using a pre-constructed multivariate coupling model, correlation coefficients between each process parameter are obtained;

[0009] Based on the correlation coefficient between each process parameter and the optimization decision scheme, the strategy calculation is carried out for different defect types, and the adjustment strategy of each process parameter is obtained.

[0010] Further, the process parameters include cavity pressure distribution, holding pressure, mold temperature field data, cooling water temperature, melt viscosity, bubble nucleus density distribution, bubble growth rate, injection gas actual injection amount and back pressure real-time value.

[0011] Further, the method for obtaining the bubble nucleus density distribution comprises:

[0012] The original echo electric signal is collected by the ultrasonic sensor array, the amplitude, phase and propagation time of the original echo electric signal are extracted by the wavelet transform algorithm, the propagation speed is obtained by the preset ratio of the cavity wall thickness to the propagation time, the real-time density value of the melt is obtained by substituting the propagation speed into the preset correlation formula, and the bubble nucleus density of different regions of the cavity is analyzed by substituting the real-time density value of the melt into the preset corresponding relationship model, thereby forming the bubble nucleus density distribution.

[0013] Further, the method for obtaining the bubble growth rate comprises:

[0014] The amplitudes and phases of the original echo electric signals at each time are substituted into the preset bubble radius-echo characteristic parameter correlation model respectively, the bubble radius value at the corresponding time is calculated, the difference between the bubble radius values of the adjacent two times is calculated to obtain the bubble radius difference value, and the bubble growth rate in the time interval between the adjacent two times is obtained by dividing the bubble radius difference value by the time interval between the adjacent two times.

[0015] Further, the method for obtaining the defect region comprises:

[0016] The bubble nucleus density of each cavity region is obtained by counting the bubble nucleus density distribution.

[0017] The deviation degree of the bubble nucleus density of each cavity region from the average value of the bubble nucleus density of all regions is calculated to obtain the density standard deviation, and the variation coefficient is obtained by dividing the density standard deviation by the average value of the bubble nucleus density of all regions.

[0018] The numerical difference between the variation coefficient of each cavity region and the preset variation coefficient threshold value is calculated, and if the numerical difference is greater than zero, the corresponding cavity region is marked as a bubble nucleus density abnormal region.

[0019] The cooling and solidification speed of each cavity region is calculated by combining the mold temperature field data and the preset heat conduction model, and the synchronization parameter of each cavity region is obtained by difference operation between the cooling and solidification speed of each cavity region and the bubble growth rate of the corresponding region.

[0020] if the bubble nucleus density of the bubble nucleus density abnormal region is higher than the average value of the bubble nucleus density of the surrounding region by a preset proportion coefficient, and the melt viscosity of the bubble nucleus density abnormal region is lower than a preset reference viscosity value, the bubble nucleus density abnormal region is determined as a craze risk region;

[0021] if the synchronism parameter of the bubble nucleus density abnormal region exceeds a preset difference threshold value, the bubble nucleus density abnormal region is determined as a shrinkage risk region.

[0022] Further, the method for generating an optimization decision scheme comprises:

[0023] For the craze risk region, a parameter adjustment mode is selected according to the size of the corresponding region bubble nucleus density excess value, when the corresponding bubble nucleus density excess value is less than a set first difference threshold value, a way of reducing the injection gas dose is used as the optimization decision scheme; when the bubble nucleus density excess value of the corresponding region is greater than or equal to the first difference threshold value, a way of applying cavity gas back pressure is used as the optimization decision scheme; the bubble nucleus density excess value is the difference between the actual bubble nucleus density of the craze risk region and the preset ideal bubble nucleus density.

[0024] For the shrinkage risk region, a way of adjusting the holding pressure and mold temperature field data is used as the optimization decision scheme.

[0025] Further, the method for obtaining the adjustment strategy of each process parameter comprises:

[0026] The prediction time domain is set as the sum of the remaining time of the current injection cycle and the next complete injection cycle;

[0027] For the craze risk scenario, if the optimization decision scheme is to use the way of reducing the injection gas dose, the reduction value of the injection gas dose and the final adjustment value of the back pressure are calculated based on the bubble nucleus density change prediction result in the prediction time domain obtained through the multivariate coupling model and the correlation coefficient between each process parameter; if the optimization decision scheme is to use the way of applying cavity gas back pressure, the pressure peak value of the gas back pressure is determined based on the melt filling progress prediction result in the prediction time domain obtained through the multivariate coupling model and the correlation coefficient between each process parameter.

[0028] For the shrinkage risk scenario, based on the synchronism parameter prediction result in the prediction time domain obtained through the multivariate coupling model, the holding pressure action time, the extension amount of the mold high temperature maintenance time in the mold temperature field data, and the adjustment gradient of the cooling water temperature are calculated.

[0029] Further, the method for obtaining the holding pressure action time, the extension amount of the mold high temperature maintenance time in the mold temperature field data, and the adjustment gradient of the cooling water temperature comprises:

[0030] The correlation coefficient between the synchronism parameter and the holding pressure action time adjustment amount is denoted as the first correlation coefficient.

[0031] The synchronism parameter is multiplied by the first correlation coefficient to obtain a shortening range of the holding pressure action time;

[0032] The correlation coefficient of the temperature and the holding pressure is recorded as a second correlation coefficient;

[0033] The shortening range of the holding pressure action time is multiplied by the second correlation coefficient to obtain an initial extension of the mold high-temperature maintenance time;

[0034] Based on the mold temperature field dynamic prediction data in the prediction time domain, a cooling speed deviation coefficient is obtained;

[0035] The mold high-temperature maintenance time extension is the initial extension of the mold high-temperature maintenance time multiplied by the cooling speed deviation coefficient;

[0036] According to the determined mold high-temperature maintenance time extension and the cooling solidification speed of the thick wall area in the prediction time domain, a cooling waterway temperature drop gradient is calculated.

[0037] Further, the method for obtaining the cooling speed deviation coefficient comprises:

[0038] Based on the mold temperature field dynamic prediction data in the prediction time domain, the cooling solidification speed of the thick wall area at different time nodes in the prediction time domain is calculated through a preset heat conduction model, the cooling solidification speed corresponding to the highest shrinkage risk time node of the thick wall area is selected as a calculation reference, and the highest shrinkage risk time node is the time node corresponding to the maximum value of the synchronism parameter;

[0039] The cooling solidification speed corresponding to the highest shrinkage risk time node of the thick wall area is divided by the ideal cooling speed obtained in advance to obtain the cooling speed deviation coefficient.

[0040] Further, the method for obtaining the cooling waterway temperature drop gradient comprises:

[0041] The correlation coefficient of the cooling waterway temperature adjustment gradient and the mold high-temperature maintenance time is recorded as a third correlation coefficient;

[0042] The correlation coefficient of the cooling waterway temperature adjustment gradient and the cooling solidification speed is recorded as a fourth correlation coefficient;

[0043] The mold high-temperature maintenance time extension is multiplied by the third correlation coefficient to obtain a first gradient adjustment component;

[0044] The cooling solidification speed is multiplied by the fourth correlation coefficient to obtain a second gradient adjustment component;

[0045] The first gradient adjustment component and the second gradient adjustment component are added to obtain a cooling waterway temperature drop gradient initial value;

[0046] The pre-stored cooling water path temperature drop gradient constraint range in the multi-variable coupling model is called, if the cooling water path temperature drop gradient initial value is in the cooling water path temperature drop gradient constraint range, the cooling water path temperature drop gradient initial value is the final cooling water path temperature drop gradient, if the cooling water path temperature drop gradient initial value exceeds the cooling water path temperature drop gradient constraint range, the boundary value of the cooling water path temperature drop gradient constraint range is taken as the final cooling water path temperature drop gradient.

[0047] The density uniformity control system based on injection molding gas dose and pressure collaborative optimization comprises:

[0048] The data acquisition module is used for acquiring process parameters in real time.

[0049] The scheme generation module is used for pre-processing the process parameters, and constructing a virtual cavity model based on the process parameters; defect prediction is carried out based on the virtual cavity model to obtain a defect area; and an optimization decision scheme is generated by comprehensively considering the defect area and corresponding process parameters.

[0050] The correlation coefficient module is used for obtaining the correlation coefficients between the process parameters by using the pre-constructed multi-variable coupling model.

[0051] The strategy generation module is used for calculating strategies for different defect types based on the correlation coefficients between the process parameters and the optimization decision scheme, to obtain adjustment strategies for the process parameters.

[0052] Compared with the prior art, the technical effects and advantages of the density uniformity control method and system based on injection molding gas dose and pressure collaborative optimization are as follows:

[0053] The cavity pressure distribution, holding pressure, mold temperature field data, cooling water path temperature, melt viscosity, bubble nucleation density distribution, bubble growth rate, injection gas actual injection amount and back pressure real-time value and other process parameters are acquired in real time, the virtual cavity model is constructed based on the pre-processed process parameters, defect prediction is carried out through the virtual cavity model to obtain a defect area, and an optimization decision scheme is generated by comprehensively considering the defect area and corresponding process parameter deviation degree; the correlation coefficients between the process parameters are obtained by using the pre-constructed multi-variable coupling model, and strategies are calculated for different defect types such as silver line risk scenarios and shrinkage risk scenarios based on the correlation coefficients and the optimization decision scheme, to obtain adjustment strategies for the process parameters such as injection gas dose reduction value, back pressure final adjustment value, holding pressure action time shortening range, mold high temperature maintenance time extension amount and cooling water path temperature drop gradient, and finally the adjustment strategies are coded into control instruction sets conforming to a communication protocol and are sent to each execution unit to complete control.

[0054] The present application solves the problem that the existing injection molding foaming control relies on preset process parameters and offline detection, cannot capture the bubble nucleus density distribution and bubble growth dynamic change in real time, and the injection gas dose, holding pressure back pressure and cooling gradient regulation are independent and have no synergy, which easily causes defects such as crazing and shrinkage, leads to the problems of decreased product mechanical properties, insufficient assembly precision, high waste rate and low production efficiency. The advantages are that the key data of molding can be obtained in real time, the working condition is intuitively presented through the virtual cavity model and the defects are accurately predicted, the process parameters are synergistically regulated by means of the multivariate coupling model, and the limitations of single parameter adjustment are avoided; the beneficial effects are that the defects of crazing and shrinkage are effectively inhibited, the product surface quality, size precision and mechanical property stability are improved, the waste rate and rework cost are reduced, and the injection molding production efficiency and product quality consistency are improved. BRIEF DESCRIPTION OF DRAWINGS

[0055] Figure 1 It is a density uniform control system based on injection gas dose and pressure synergistic optimization of the embodiment of the present application.

[0056] Figure 2 It is a density uniform control method flow chart based on injection gas dose and pressure synergistic optimization of the embodiment of the present application.

[0057] Figure 3 It is a method flow chart for obtaining a defect area.

[0058] Figure 4 It is a method flow chart for obtaining an adjustment strategy of each process parameter. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present application will be described in detail, clearly and completely below with reference to the drawings in the embodiments of the present application. It should be particularly noted that the specific embodiments described below are only used to better illustrate and describe the technical solutions of the present application, and are intended to enable those skilled in the art to better understand and implement the present application, and should not be understood as limiting the protection scope of the present application. Those skilled in the art can modify, adjust or equivalently replace the present application according to the content disclosed in the present application without departing from the spirit and essence of the present application, and these should be regarded as the protection scope of the present application.

[0060] Embodiment 1:

[0061] Please refer to Figure 1 The present embodiment discloses a density uniform control system based on injection gas dose and pressure synergistic optimization, which comprises a data acquisition module, a scheme generation module, a correlation coefficient module and a strategy generation module, each module is connected through wired and / or wireless connection to realize data transmission.

[0062] The data acquisition module is used for acquiring process parameters in real time.

[0063] The process parameters include cavity pressure distribution, holding pressure, mold temperature field data, cooling water temperature, melt viscosity, bubble nucleation density distribution, bubble growth rate, injection gas actual injection amount, and back pressure real-time value.

[0064] The injection gas dosage is monitored by a sensor / mass flow meter installed on the injection molding machine screw barrel or supercritical fluid injection system to monitor the amount of chemical or physical injection gas added, and the actual weight or volume of injection gas added in each injection cycle is obtained. The holding pressure is provided by the holding pressure set value and real-time measured value provided by the injection molding machine control system. The cavity pressure curve during holding can be obtained by a pressure sensor or cavity sensor, and the peak value and duration information are collected. The back pressure real-time value is the back pressure set and actual value during the pre-plasticizing stage of the screw. The back pressure applied during screw plasticizing is recorded by a pressure sensor at the end of the injection molding machine barrel, which represents the level of back pressure experienced by the melt during metering mixing. Temperature sensors are installed at the inlet and outlet of the mold cooling circuit to monitor the cooling medium temperature in real time. If necessary, each water circuit is equipped with an adjustable valve to dynamically control the local cooling flow and temperature, and the actual cooling temperature change of each area is recorded. Cavity gas back pressure: a controllable gas pressure gauge is introduced into the mold cavity to raise the internal gas pressure before injection. The cavity gas pressure value is monitored by a pressure sensor, and the inflation / pressure relief time is controlled by a solenoid valve to suppress premature foaming during melt filling and ensure the densification of the surface layer. The cavity pressure distribution is obtained by arranging multiple embedded high-precision micro pressure sensors in the mold cavity to synchronously collect the pressure changes at each point during filling and holding. The pressure sensor provides the spatial distribution and pressure gradient information of the cavity pressure, which is used to determine the position of the melt flow front and the pressure condition of each region. The mold temperature field data is obtained by arranging multiple thermocouples or optical fiber temperature sensors on the mold cavity wall to measure the temperature change of each region in real time, and the temperature field distribution in the cavity is obtained. Non-contact monitoring can also be combined with a mold surface infrared thermal imager to capture the temperature gradient and ensure a comprehensive understanding of the mold temperature dynamics. The melt viscosity is estimated by monitoring the relationship between the screw advancing speed and the injection pressure during injection, and the real-time viscosity change is calculated; or the melt state is detected by an ultrasonic sensor installed at the barrel / nozzle to indirectly infer the real-time viscosity of the melt. The change in melt viscosity can be calculated from the characteristics of the pressure-time curve, such as the increase in viscosity leading to an increase in filling pressure. The melt viscosity is used to reflect the fluctuation of material properties for dynamic adjustment of the control strategy.

[0065] First, according to the wall thickness distribution of the injection foaming molding machine, the complexity of the product structure, and the equipment specifications and product structure of the automobile parts, complete the deployment of various sensors and the input of initial process parameters. The specific method includes: using flush mounting structure to embed high-precision micro pressure sensors at key positions such as the feed port of the mold cavity, the middle and end of the cavity, etc., to ensure that the sensor probe is flush with the cavity wall to avoid disturbing the melt flow; fixing temperature sensors at the inlet and outlet of each independent water circuit of the mold cooling circuit; arranging multi-point thermocouples and optical fiber temperature sensors in different areas such as thick wall area and thin wall area of the mold cavity wall to form a three-dimensional monitoring network of the mold temperature field; installing mass flow meters in the injection gas injection port of the injection molding machine screw barrel or the pipeline of the supercritical fluid injection system for real-time measurement of the injection amount of injection gas; fixing ultrasonic sensor arrays at the transition area of the thick wall in the cavity, the corner, the end of the melt flow, and the areas where the bubble density is abnormal or the bubble growth rate is unbalanced frequently in the history of production. These areas are prone to abnormal bubble evolution due to changes in melt flow state and temperature gradient differences during injection foaming, which are the core positions affecting product density uniformity and defect generation. The probe of the ultrasonic sensor is tightly attached to the outer wall of the mold through a coupling agent to ensure that the ultrasonic wave can effectively penetrate the mold steel and act on the melt in the cavity; torque sensors and displacement sensors are installed in the pre-plasticizing section and injection section of the injection molding machine screw, respectively, for subsequent estimation of melt viscosity.

[0066] After completing the sensor deployment, the preset injection gas dose, holding pressure peak value, back pressure reference value, and cooling water target temperature are input, and the data acquisition timing and frequency are set. For example, the data acquisition frequency needs to be synchronized with the injection cycle to ensure that data capture of no less than 50 ms / time can be achieved in each stage of melt filling, holding, and cooling. After starting the injection molding machine, the pressure sensor continuously captures the pressure changes at each position of the cavity and generates a pressure curve, the temperature sensor synchronously records the dynamic data of the cooling water temperature and the mold wall temperature, the mass flow meter real-time feedbacks the instantaneous injection amount and cumulative injection amount of the injection gas, the ultrasonic sensor array emits ultrasonic waves according to the preset timing and receives the echo signals reflected by the melt, and the screw torque sensor and displacement sensor record the torque changes in the pre-plasticizing stage and the screw advancing speed in the injection stage.

[0067] The collected original signals are processed in real time. The original signals refer to the initial electric signals output directly by the sensors during the data collection process without processing, including the voltage signals output by the high-precision micro pressure sensor based on the cavity pressure change, the resistance signals output by the temperature sensor based on the temperature change, the original echo electric signals received by the ultrasonic sensor array, the pulse signals output by the mass flow meter based on the injection gas flow change, the current signals output by the screw torque sensor based on the torque change, and the position signals output by the screw displacement sensor based on the displacement change. For the pressure signals output by the high-precision micro pressure sensor and the temperature signals output by the temperature sensor, a low-pass filtering algorithm is used to filter out the high-frequency noise generated by the equipment vibration, so as to obtain the cavity pressure distribution and the mold temperature field data. For the original echo electric signals received by the ultrasonic sensor array, the amplitude, phase and propagation time of the original echo electric signals are extracted through a wavelet transform algorithm, and then the bubble nucleus density distribution is analyzed in combination with the correlation formula of the sound velocity and density of the melt. The specific method includes: obtaining the propagation velocity through the preset ratio of the cavity wall thickness and the propagation time; the mold cavity wall thickness data is a fixed parameter determined in the mold design stage; then the correlation formula of the sound velocity and density of the melt is called, which is obtained through the previous experiment. In the experiment process, the sound velocity of the corresponding melt is tested under different bubble nucleus densities, and the multiple sets of sound velocity and density data obtained by testing are fitted to obtain the functional relationship between the sound velocity and the density; the calculated propagation velocity of the ultrasonic wave in the melt is substituted into the correlation formula to obtain the real-time density value of the melt; the previously established corresponding relationship model of the melt density and the bubble nucleus density is called, which is obtained through the previous experiment. In the experiment process, the density of the corresponding melt is tested under different bubble nucleus densities to establish the mapping relationship between them; the obtained real-time density value of the melt is substituted into the corresponding relationship model to analyze the bubble nucleus density in different regions of the cavity, and then the bubble nucleus density distribution is formed.Meanwhile, the bubble growth rate is calculated based on the time sequence change rate of the echo signal, and the specific method comprises the following steps: calling a parameter correlation model of bubble radius and echo characteristics which is established in advance for the material used in the current injection molding, the parameter correlation model is obtained through previous experiments, the experiment process is to observe the bubble growth process by using a high-speed camera device, and the bubble radius and the corresponding ultrasonic echo amplitude and phase data at different times are recorded synchronously, a plurality of groups of bubble radius and echo characteristic parameter data are fitted to obtain a functional relationship between bubble radius and echo characteristic parameter; based on the original echo electric signals at different times collected by the ultrasonic sensor array in real time, the amplitude and phase characteristic parameters of the echoes at each time are extracted through a wavelet transform algorithm; the amplitude and phase characteristic parameters at each time are substituted into the bubble radius and echo characteristic parameter correlation model respectively, and the bubble radius value at the corresponding time is calculated; the bubble radius difference between adjacent two times is calculated, and then the bubble radius difference is divided by the time interval between adjacent two times to obtain the bubble growth rate in the time interval between adjacent two times; for the torque data recorded by the screw torque sensor and the displacement data recorded by the displacement sensor, the power-law fluid model is combined with the injection pressure data of the injection molding machine, and the viscosity estimation formula constructed based on the correlation equation of the screw shear rate and the torque is used to derive the melt viscosity in real time.

[0068] The scheme generation module is configured to preprocess the process parameters, and construct a virtual cavity model based on the process parameters; based on the virtual cavity model, defect prediction is carried out to obtain a defect area; and an optimization decision scheme is generated by comprehensively considering the defect area and the corresponding process parameters.

[0069] The method for preprocessing the process parameters comprises the following steps: using a time sequence alignment algorithm to match the same cycle data collected by different sensors according to the time stamp, so as to eliminate the time deviation caused by the difference in response speed of the sensors; using a moving average method and other data smoothing algorithms to further filter the fluctuation noise of the pressure and temperature data, and simultaneously performing spatial interpolation processing on the bubble nucleus density distribution data obtained by ultrasonic analysis, so as to fill in the data gaps in the local area of the cavity caused by the insufficient density of the sensor arrangement, and form a complete cavity bubble nucleus density spatial distribution map; performing abnormal value detection on the melt viscosity estimation value, if there is any time when the viscosity value exceeds the upper and lower limits of the viscosity preset based on the melt index of the plastic, then the viscosity data of adjacent times are used for linear interpolation correction to ensure the effectiveness of the data.

[0070] Based on the real-time collected process parameters, a virtual cavity model consistent with the actual molding process is constructed, the dynamic processes of melt flow, bubble evolution and cooling and solidification are simulated, and the working condition state inside the cavity is intuitively presented. The specific method comprises: obtaining the three-dimensional geometric parameters of the cavity determined in the mold design stage, including the overall profile size of the cavity, the wall thickness of each region, the corner radius, the position and size of the feeding port, and establishing a three-dimensional geometric model of the cavity in the simulation platform based on the three-dimensional geometric parameters of the cavity; importing the real-time collected process parameters in the three-dimensional geometric model of the cavity, mapping the cavity pressure distribution to the pressure boundary condition of the model, mapping the mold temperature field data to the temperature boundary condition of the model, inputting the melt viscosity as the melt flow characteristic parameter, inputting the bubble nucleus density distribution as the initial state parameter of the bubble, inputting the actual injection amount of injection gas as the material source parameter of bubble generation, and inputting the real-time value of back pressure as the back pressure constraint parameter of the melt plasticizing stage; building a coupled physical field model in the simulation platform, including a melt flow model, a bubble evolution model and a cooling and solidification model, wherein the melt flow model uses continuum mechanics equations to describe the flow state of the melt in the cavity, and the equation parameters are determined by the melt viscosity estimation value and the cavity pressure distribution data; the bubble evolution model uses bubble dynamics equations to describe the nucleation, growth and merging process of the bubble, and the equation parameters are determined by the bubble nucleus density distribution data and the actual injection amount data of the injection gas; the cooling and solidification model uses heat conduction equations to describe the temperature transfer process of the cavity and the melt, and the equation parameters are determined by the mold temperature field data; the coupled physical field model is solved by numerical calculation methods such as finite element method, and the latest collected process parameters are called in real time to update the parameters of the virtual cavity model during the solving process, so as to ensure that the calculation results of the virtual cavity model are consistent with the dynamic changes of the actual molding process; the results of the melt flow trajectory, bubble size change, temperature distribution change and the like obtained by solving are presented in a visual form in the three-dimensional geometric model of the cavity, and a complete virtual cavity model is obtained.

[0071] Please refer to Figure 3 As shown, based on the constructed virtual cavity model, defect prediction is carried out to obtain a defect area, and the specific method comprises:

[0072] The standard deviation is obtained by the deviation degree of the nucleation density of different regions of the statistical cavity from the average value of the nucleation density of all regions; the coefficient of variation is obtained by dividing the standard deviation by the average value of the nucleation density of all regions; the coefficient of variation of the nucleation density of different regions of the cavity is compared with the coefficient of variation threshold value obtained based on the test of the same qualified product; the numerical difference between the current coefficient of variation and the coefficient of variation threshold value is calculated during the comparison, if the numerical difference is greater than zero, it indicates that the current bubble hole distribution uniformity does not meet the preset requirements, and the corresponding cavity region is marked as a nucleation density abnormal region; if the numerical difference is less than or equal to zero, it indicates that the current bubble hole distribution uniformity meets the preset requirements, and through the comparison, it can be determined whether the bubble hole distribution is uniform and the specific region of the abnormal distribution; at the same time, combined with the mold temperature field data, the cooling and solidification speed of each region of the cavity is calculated through the heat conduction model, and the cooling and solidification speed of each region of the cavity is subtracted from the bubble growth rate of the corresponding region to obtain the synchronization parameter of the cooling and solidification speed of each region of the cavity and the bubble growth rate of the corresponding region.

[0073] If the nucleation density of the nucleation density abnormal region is higher than the average value of the nucleation density of the surrounding region of the preset proportion coefficient, and the melt viscosity of the nucleation density abnormal region is lower than the reference viscosity value of the corresponding plastic material in the molten state, it indicates that the melt fluidity is too strong, and the bubbles are more likely to gather in the nucleation density abnormal region, further verifying the judgment of the decrease of the melt strength, and then predicting the nucleation density abnormal region as a craze risk region; the reference viscosity value is determined by the material manual query and the previous experimental calibration; the ideal range is the reasonable interval of the nucleation density coefficient of variation obtained based on the test of the same qualified product, and the reasonable interval of the nucleation density coefficient of variation is consistent with the coefficient of variation threshold value used for comparison in the previous period. The generation of craze defects is directly related to the degree of bubble aggregation in the melt and the strength of the melt itself. When the nucleation density coefficient of variation exceeds the ideal range and the local nucleation density is too high, it means that the bubbles in the nucleation density abnormal region are excessively aggregated, which will destroy the internal structure integrity of the melt and reduce the tensile and shear resistance of the melt; and the melt viscosity lower than the reference viscosity value will further enhance the melt fluidity, making the bubbles more likely to gather in the local region. Under the double action, the melt strength decreases significantly, when the melt is constrained by the cavity wall or the internal stress generated by the subsequent cooling shrinkage, the craze defects are easily formed in the nucleation density abnormal region, so the nucleation density abnormal region can be predicted as a craze risk region through the above conditions.

[0074] If the difference between the bubble growth rate and the cooling solidification rate of the abnormal area of the bubble nucleus density exceeds a preset difference threshold, it is determined that the bubble expansion rate cannot match the material cooling shrinkage rate, and the abnormal area of the bubble nucleus density is predicted to be a shrinkage risk area; the difference threshold is preset through a matching test of the material cooling shrinkage rate and the bubble expansion rate, for example, the difference accounts for no more than 10% of the cooling solidification rate, which can be determined based on production data statistics of similar products without shrinkage defects. The shrinkage defect is caused by the fact that the volume shrinkage in the material cooling process is not effectively compensated. Only when the bubble growth rate needs to be synchronized with the cooling solidification rate, can the volume shrinkage generated in the material cooling process be filled by the bubble expansion.

[0075] Finally, the optimization decision scheme is generated by comprehensively considering the defect area and the corresponding process parameters. The specific method includes:

[0076] For the silver wire risk area, the parameter adjustment mode is selected according to the size of the corresponding area bubble nucleus density difference value. When the corresponding bubble nucleus density difference value is less than a first difference threshold, the injection gas dose is reduced, the dose adjustment amount is calculated by the correlation curve of the bubble nucleus density difference value and the injection gas dose, the correlation curve is fitted by a plurality of sets of bubble nucleus density test data under different injection gas doses, and the bubble nucleus density of the area is ensured to return to the ideal range after the dose is reduced. When the bubble nucleus density difference value of the specific area is greater than or equal to the first difference threshold, the counter-pressure of the cavity gas is applied, the counter-pressure value is determined according to the difference between the real-time cavity pressure value and the ideal pressure value of the specific area, the ideal pressure value is the critical pressure for inhibiting the excessive nucleation of the material bubbles, and the critical pressure is obtained by a material nucleation pressure test, for example, the cavity pressure of the area needs to be maintained within the ideal pressure value ± 0.2 bar after the counter-pressure is applied.

[0077] For the shrinkage risk area, the adjustment length of the holding pressure action time, the extension amount of the corresponding area mold high temperature maintenance time derived based on the mold temperature field data, and the adjustment gradient of the cooling water temperature are calculated, and the corresponding optimization decision scheme is generated. The specific method comprises: shortening the holding pressure action time by the calculated adjustment length, so that the shortened holding pressure action time can reduce the excessive shrinkage of the melt; extending the mold high temperature maintenance time of the corresponding area by the calculated extension amount to slow down the cooling speed of the area and provide sufficient time for bubble growth; the mold high temperature maintenance time belongs to the regulation and control dimension of the mold temperature field data, which is calculated through the temperature maintenance parameters of the area in the mold temperature field data; the descending gradient of the cooling water temperature from the high temperature stage to the low temperature stage is increased by the calculated adjustment gradient to ensure that the cooling speed after adjustment can promote the synchronization of bubble expansion and material shrinkage; the adjustment of the cooling water temperature is realized by adjusting the cooling water related regulation and control parameters in the mold temperature field data. The above-mentioned shortening of the holding pressure action time, the extension of the mold high temperature maintenance time, and the increase of the descending gradient of the cooling water temperature are implemented in coordination to form a complete optimization decision scheme for the shrinkage risk area. Through the multi-parameter linkage adjustment, the synchronization parameter of the bubble growth rate and the cooling and solidification speed is returned to a reasonable range, so as to inhibit the generation of shrinkage defects. Among them, the adjustment length of the holding pressure action time is calculated according to the size of the synchronization parameter, the synchronization parameter is the difference between the bubble growth rate and the cooling and solidification speed of the shrinkage risk area, and the holding pressure action time is shortened by a preset first time length unit every time the synchronization parameter reaches a preset first proportion threshold. The first proportion threshold and the first time length unit are obtained through the shrinkage defect correlation test of the holding pressure parameters and the shrinkage defects of the same type qualified product; the extension amount of the mold high temperature maintenance time of the corresponding area is determined based on the difference between the cooling and solidification speed of the corresponding area and the ideal cooling speed, the ideal cooling speed is the cooling speed matching the bubble growth and material shrinkage, and the ideal cooling speed is obtained through the cooling curve test of the same type qualified product. The mold high temperature maintenance time is extended by a preset second time length unit every time the cooling and solidification speed exceeds the ideal cooling speed by a preset temperature rate threshold, and the temperature rate threshold and the second time length unit are obtained through the cooling parameter and the shrinkage defect correlation test of the same type qualified product; the adjustment gradient of the cooling water temperature is set according to the deviation degree of the synchronization parameter, and the descending gradient of the cooling water temperature from the high temperature stage to the low temperature stage is increased by a preset temperature gradient unit every time the synchronization parameter deviation reaches a preset second proportion threshold, and the second proportion threshold and the temperature gradient unit are obtained through the cooling gradient and the shrinkage defect correlation test of the same type qualified product.

[0078] The correlation coefficient module is configured to obtain correlation coefficients between the process parameters by using the pre-constructed multivariate coupling model.

[0079] The method for constructing the multivariable coupling model comprises: the multivariable coupling model comprises a multivariable coupling matrix, and the multivariable coupling matrix is a core component for quantifying the correlation between process parameters in the multivariable coupling model. The multivariable coupling model quantifies the correlation between the injection gas amount, the holding pressure, the back pressure value and the cooling water temperature based on historical data and real-time working conditions, and the specific process is as follows: first, the mutual influence factors between the process parameters are determined, the influence factors including the direct action strength and the indirect action path of the change of one process parameter on other process parameters; then, a plurality of orthogonal tests are designed, the injection foam molding is carried out under different combinations of the injection gas amount, the holding pressure, the back pressure value and the cooling water temperature, the product quality data and the process parameter dynamic response data corresponding to each test are collected, and the experimental data are obtained; based on the experimental data, the correlation coefficients between the process parameters are calculated by using the multiple linear regression or the partial least squares regression algorithm, the correlation coefficient representing the value of the change of other process parameters corresponding to the unit change of one process parameter; all the correlation coefficients are arranged according to the corresponding relationship of the injection gas amount, the holding pressure, the back pressure value and the cooling water temperature, and the multivariable coupling matrix is formed. Through the multivariable coupling matrix, the correlation range of the adjustment of other process parameters when one process parameter is adjusted is determined, and it is ensured that the adjustment of each parameter forms a synergistic effect rather than mutual interference. For example, the adjustment range of the back pressure corresponding to the decrease of 1% of the injection gas amount is determined when the melt fluidity is maintained; the gradient value of the cooling water temperature that needs to be adjusted synchronously when the peak value of the holding pressure changes by 0.5 bar is determined, and it is ensured that the adjustment of each process parameter does not interfere with each other but forms a synergistic effect.

[0080] The prediction time domain of the MPC is set as the sum of the remaining time of the current injection molding cycle and the next complete injection molding cycle. Since the defect in the injection foaming process is generated from the deviation of the process parameters to the final appearance on the product, it needs to go through the melt flow, bubble evolution and cooling stage of the remaining current cycle, and may continue to the molding process of the next cycle. The setting of the prediction time domain of the MPC can capture the dynamic change trend of the injection gas amount, holding pressure, back pressure value, cooling water temperature and other process parameters in the historical molding data and real-time working condition data within this time range, and the defect development path that these changes may cause, provide time dimension data support covering the complete process of defect generation for subsequent specific strategy calculation for defect scenarios, avoid missing the key regulation nodes due to insufficient prediction time range, and further cause strategy calculation to miss key regulation nodes; The control time domain of the MPC is set according to the response speed of each process parameter adjustment. Due to the differences in hardware structure and working principle, the execution units corresponding to different process parameters have different adjustment response speeds, such as injection gas metering device, injection molding machine hydraulic / servo system, mold temperature controller, etc. The setting of the control time domain of the MPC needs to be adapted to the response speed of each process parameter to ensure that the control instructions generated for each parameter can be issued to the execution unit within the effective time corresponding to its response speed and complete the adjustment action, avoiding the lag or advance of instruction issuance due to the mismatch between the control time domain and the response speed of the process parameter, so that the process parameter adjustment cannot timely intervene in the defect development, and further affect the regulation effect. The setting of the prediction time domain and the control time domain of the MPC is an important prerequisite for subsequent strategy calculation based on the multivariate coupling model for silver line risk scenarios and shrinkage risk scenarios. The determination of the adjustment amount and adjustment time of each process parameter in the subsequent strategy calculation needs to be based on the process parameter change prediction results within the prediction time domain to judge the development node of the defect in the time dimension, and the control time domain as the time constraint of instruction execution to ensure that the generated adjustment strategy can act on the execution unit in time before the defect worsens, realizing the accurate inhibition of the defect.

[0081] The strategy generation module calculates the adjustment strategy of each process parameter based on the correlation coefficient between each process parameter and the optimized decision scheme for different defect types.

[0082] Please refer to Figure 4 As shown, the method for calculating the adjustment strategy of each process parameter based on the correlation coefficient between each process parameter and the optimized decision scheme for different defect types includes:

[0083] For the silver risk scenario, if the optimization decision scheme is to reduce the injection gas dose, first determine the peak node of the bubble nucleus density deviation value in the time dimension in the silver risk area based on the prediction result of the bubble nucleus density change in the prediction time domain, extract the bubble nucleus density deviation value of the silver risk area corresponding to the peak node from the optimization decision scheme, and the bubble nucleus density deviation value is the difference between the actual bubble nucleus density and the ideal bubble nucleus density in the silver risk area; the ideal bubble nucleus density is obtained by testing the bubble nucleus density of the same type of qualified products, specifically collecting the bubble nucleus density data of the cavity area of multiple same type of qualified products, and calculating the average value of the bubble nucleus density data to obtain the ideal bubble nucleus density; multiply the bubble nucleus density deviation value by the correlation coefficient of the bubble nucleus density deviation value and the injection gas dose adjustment amount in the multivariate coupling matrix to obtain the reduction value of the injection gas dose, while ensuring that the reduction value of the injection gas dose corresponds to the adjustment action that can be completed within the control time domain to take effect before the bubble nucleus density deviation value reaches the peak. Because the injection gas dose is coupled with the back pressure value, reducing the injection gas dose will change the gas content of the melt, thereby affecting the viscosity and flow characteristics of the melt. If only the injection gas dose is adjusted without adjusting the back pressure synchronously, it may lead to the situation that the melt is over-sheared when the back pressure is too high in the plasticizing stage or the melt is insufficient in flowability when the back pressure is too low, the former will exacerbate the silver risk by damaging the internal structure of the melt, and the latter may lead to uneven distribution of bubble nuclei, so the back pressure adjustment value needs to be calculated synchronously based on the coupling relationship between the two. Multiply the reduction value of the injection gas dose by the correlation between the injection gas dose and the back pressure value in the multivariate coupling matrix to obtain the initial adjustment value of the back pressure; combine the prediction result of the melt viscosity change in the prediction time domain to calculate the value that the back pressure needs to be reduced synchronously, specifically: calculate the ratio of the predicted melt viscosity to the reference viscosity value of the corresponding plastic material in the molten state to obtain the melt viscosity deviation coefficient; if the predicted melt viscosity increases due to the reduction of the injection gas dose, i.e. the melt viscosity deviation coefficient is greater than 1, then the final adjustment value of the back pressure is the initial adjustment value of the back pressure multiplied by (2-melt viscosity deviation coefficient), wherein the coefficient (2-melt viscosity deviation coefficient) is determined through the preliminary melt viscosity and back pressure matching experiment to ensure that the back pressure adjustment amplitude is moderately reduced with the increase of viscosity; if the predicted melt viscosity decreases due to the reduction of the injection gas dose, i.e. the melt viscosity deviation coefficient is less than 1, then the final adjustment value of the back pressure is the initial adjustment value of the back pressure multiplied by (1 / melt viscosity deviation coefficient) to ensure that the back pressure adjustment amplitude is moderately increased with the decrease of viscosity; the value that the back pressure needs to be reduced synchronously is finally determined through the above calculation, and the execution time of the back pressure adjustment action is constrained within the control time domain to avoid excessive shearing of the melt due to excessive back pressure; the value that the back pressure needs to be reduced synchronously is finally determined through the above calculation, and the execution time of the back pressure adjustment action is constrained within the control time domain, so that the injection gas dose and the back pressure are adjusted coordinately to reduce the excessive nucleation of bubble nuclei to inhibit silver and ensure the stability of melt flow and plasticization.

[0084] If the optimized decision scheme is to adopt the way of applying cavity gas back pressure, first, based on the melt filling progress prediction results in the prediction time domain, the key filling stage of bubble over-nucleation in the silver risk area is determined, then the melt filling progress of the current cycle is calculated through the cavity pressure sensor and displacement sensor data, which is the ratio of the current melt filling volume to the total cavity volume; the correlation constraint condition of the pressure parameter in the multivariate coupling matrix and the melt filling progress and bubble nucleation density is called, which limits the reasonable range of cavity gas back pressure under different melt filling progress; according to the bubble nucleation density control target of the silver risk area, the pressure peak value of the gas back pressure is determined within the range limited by the correlation constraint condition; at the same time, based on the melt flow velocity prediction in the prediction time domain, the melt filling progress reaches the preset threshold value as the gas back pressure application time, which is determined based on the correlation between melt flow and bubble nucleation in the multivariate coupling model, and the application time matches the control time domain, ensuring that the back pressure application action can be completed within the key filling stage, effectively inhibiting the excessive foaming of the silver risk area; finally, through the pressure closed-loop model, the actual value of the cavity gas back pressure is collected in real time, compared with the pressure peak value, and the back pressure control valve opening is dynamically adjusted to ensure that the back pressure application process is smooth and impact-free.

[0085] The correlation coefficient between the synchronism parameter and the adjustment amount of the holding pressure action time in the multivariate coupling matrix is denoted as a first correlation coefficient for the shrinkage risk area; the shrinkage risk area is multiplied by the first correlation coefficient to obtain the shortening range of the holding pressure action time, and the adjusted action time of the shortened holding pressure action time is constrained in the control time domain to ensure that the action takes effect before the synchronism parameter deteriorates and to reduce the excessive shrinkage of the melt. The correlation coefficient between the temperature and the holding pressure in the multivariate coupling matrix is denoted as a second correlation coefficient; the shortening range of the holding pressure action time is multiplied by the second correlation coefficient to obtain an initial extension of the mold high-temperature maintenance time; based on the dynamic prediction data of the mold temperature field in the prediction time domain, the cooling and solidification speed of the thick wall area at different time nodes in the prediction time domain is calculated through a preset heat conduction model, the cooling and solidification speed corresponding to the highest shrinkage risk time node of the thick wall area is selected as a calculation reference, and the highest shrinkage risk time node is the time node corresponding to the maximum synchronism parameter; the ideal cooling speed is obtained by calling the cooling curve test of the same qualified product in advance, and the ideal cooling speed is the cooling and solidification speed of the same qualified product at the same time node and the same wall thickness area corresponding to the calculation reference; the cooling and solidification speed corresponding to the highest shrinkage risk time node of the thick wall area is divided by the corresponding ideal cooling speed to obtain a cooling speed deviation coefficient; the mold high-temperature maintenance time extension is the initial extension of the mold high-temperature maintenance time multiplied by the cooling speed deviation coefficient; the extension of the mold high-temperature maintenance time corresponding to the area is finally determined through the above calculation, and the execution time of the extension action matches the control time domain to ensure timely slowing down the cooling speed. The correlation coefficient between the cooling water temperature adjustment gradient and the mold high-temperature maintenance time in the multivariate coupling matrix is denoted as a third correlation coefficient; the mold high-temperature maintenance time extension is multiplied by the third correlation coefficient to obtain a first gradient adjustment component; the correlation coefficient between the cooling water temperature adjustment gradient and the cooling and solidification speed in the multivariate coupling matrix is denoted as a fourth correlation coefficient; the cooling and solidification speed is multiplied by the fourth correlation coefficient to obtain a second gradient adjustment component; the first gradient adjustment component and the second gradient adjustment component are added to obtain an initial value of the cooling water temperature descending gradient; finally, the pre-stored cooling water temperature descending gradient constraint range in the multivariate coupling model is called, the cooling water temperature descending gradient constraint range is determined based on the maximum adjustment capacity of the injection molding equipment cooling system and the gradient effective value in the production of the same qualified product, if the initial value of the descending gradient is within the cooling water temperature descending gradient constraint range, the initial value of the descending gradient is the final cooling water temperature descending gradient; if the initial value of the descending gradient exceeds the cooling water temperature descending gradient constraint range, the boundary value of the cooling water temperature descending gradient constraint range is taken as the final cooling water temperature descending gradient, so that the calculated descending gradient not only meets the parameter correlation law, but also meets the equipment operation limitation in actual production, thereby ensuring that the cooling speed adjustment can promote the synchronization of bubble expansion and material shrinkage.

[0086] After the adjustment strategy calculation is completed, the adjustment strategy of each process parameter is encoded into a control instruction set that conforms to the communication protocol of the injection molding machine and each execution unit, and the execution time of the control instruction set strictly matches the control time domain, ensuring accurate correspondence with the defect development nodes in the prediction time domain. The control instruction set is then issued to each execution unit for corresponding control, realizing early inhibition and correction of defect development, and reducing waste and rework costs.

[0087] Embodiment 2:

[0088] Please refer to Figure 2 The embodiment provides a density uniformity control method based on injection gas dosage and pressure collaborative optimization, which comprises the following steps:

[0089] Real-time acquisition of process parameters;

[0090] Pretreatment of the process parameters, and construction of a virtual cavity model based on the process parameters; defect prediction based on the virtual cavity model to obtain a defect area; and generation of an optimization decision scheme by integrating the defect area and the corresponding process parameters;

[0091] Using a pre-constructed multivariate coupling model, correlation coefficients between each process parameter are obtained;

[0092] Based on the correlation coefficients between each process parameter and the optimization decision scheme, strategy calculation is performed for different defect types to obtain the adjustment strategy of each process parameter.

[0093] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited thereto. Any skilled person in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0094] Finally: the above is only a preferred embodiment of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A density uniformity control method based on the synergistic optimization of injection gas dosage and pressure, characterized in that, include: Process parameters are acquired in real time; Process parameters include cavity pressure distribution, holding pressure, mold temperature field data, cooling water temperature, melt viscosity, bubble nucleus density distribution, bubble growth rate, actual injection volume of injection gas, and real-time back pressure value. The process parameters are preprocessed, and a virtual cavity model is constructed based on the process parameters; Defect prediction is performed based on a virtual cavity model to obtain the defect region; By combining the defective areas and their corresponding process parameters, an optimized decision-making scheme is generated; Methods for obtaining defect areas include: The density of bubble nuclei in each cavity region was obtained by statistically analyzing the bubble nuclei density distribution. The deviation of the bubble nucleus density in each cavity region from the average bubble nucleus density of all regions is calculated to obtain the density standard deviation; the density standard deviation is divided by the average bubble nucleus density of all regions to obtain the coefficient of variation. Calculate the difference between the coefficient of variation of each cavity region and the preset coefficient of variation threshold; if the difference is greater than zero, mark the corresponding cavity region as a region with abnormal nucleus density. By combining the mold temperature field data, the cooling and solidification rate of each cavity region is calculated using a preset heat conduction model; the difference between the cooling and solidification rate of each cavity region and the bubble growth rate of the corresponding region is calculated to obtain the synchronization parameters of each cavity region. If the bubble nucleus density in an abnormal bubble nucleus density area is higher than the average bubble nucleus density of the surrounding area by a preset proportional coefficient, and the melt viscosity in the abnormal bubble nucleus density area is lower than the preset reference viscosity value, then the abnormal bubble nucleus density area is determined to be a crazing risk area. If the synchronization parameter of a region with abnormal bubble nucleus density exceeds a preset difference threshold, then the region with abnormal bubble nucleus density is determined to be a region at risk of shrinkage. Using a pre-constructed multivariate coupling model, the correlation coefficients between various process parameters are obtained; Based on the correlation coefficients between various process parameters and optimization decision schemes, including: For areas with silver streaks, parameter adjustment methods are selected based on the magnitude of the bubble nucleus density deviation in the corresponding area. When the deviation is less than a set first difference threshold, reducing the injection gas dosage is used as an optimization decision. When the deviation is greater than or equal to the first difference threshold, applying cavity gas back pressure is used as an optimization decision. The bubble nucleus density deviation is the difference between the actual bubble nucleus density in the silver streaks risk area and the preset ideal bubble nucleus density. For areas with shrinkage risk, adjusting the holding pressure and mold temperature field data is used as an optimization decision. The prediction time domain is set as the sum of the remaining duration of the current injection cycle and the next complete injection cycle. Strategy calculations are performed for different defect types to obtain adjustment strategies for each process parameter. Specifically, for the silver streak risk scenario, if the optimized decision is to reduce the injection gas dosage, the reduction value of the injection gas dosage and the final adjustment value of the back pressure are calculated based on the predicted bubble nucleus density change in the prediction time domain obtained through a multivariate coupling model and the correlation coefficients between various process parameters. If the optimized decision is to apply cavity gas back pressure, the peak pressure of the gas back pressure is determined based on the predicted melt filling progress in the prediction time domain obtained through a multivariate coupling model and the correlation coefficients between various process parameters. For the shrinkage risk scenario, the holding pressure is calculated based on the predicted synchronization parameter in the prediction time domain obtained through a multivariate coupling model. The extension of the high-temperature holding time of the mold and the adjustment gradient of the cooling water temperature in the mold temperature field data are used to calculate the pressure application time. The correlation coefficient between the synchronicity parameter and the adjustment of the holding pressure application time is recorded as the first correlation coefficient. Multiplying the synchronicity parameter by the first correlation coefficient yields the reduction in the holding pressure application time. The correlation coefficient between temperature and holding pressure is recorded as the second correlation coefficient. Multiplying the reduction in the holding pressure application time by the second correlation coefficient yields the initial extension of the high-temperature holding time of the mold. Based on the dynamic prediction data of the mold temperature field in the prediction time domain, the cooling rate deviation coefficient is obtained. The extension of the high-temperature holding time of the mold is the initial extension of the high-temperature holding time of the mold multiplied by the cooling rate deviation coefficient. Based on the determined extension of the high-temperature holding time of the mold and the cooling solidification rate of the thick-walled region in the prediction time domain, the decreasing gradient of the cooling water temperature is calculated.

2. The density uniformity control method based on the synergistic optimization of injection gas dosage and pressure according to claim 1, characterized in that, Methods for obtaining the density distribution of bubble nuclei include: The original echo electrical signal is acquired by an ultrasonic sensor array, and the amplitude, phase and propagation time of the original echo electrical signal are extracted by wavelet transform algorithm. The propagation speed is obtained by using the preset ratio of cavity wall thickness to propagation time. The propagation speed is substituted into the preset correlation formula to obtain the real-time density value of the melt. The real-time density value of the melt is substituted into the preset correspondence model to analyze the bubble nucleus density in different regions of the cavity, and thus form the bubble nucleus density distribution.

3. The density uniformity control method based on the synergistic optimization of injection gas dosage and pressure according to claim 2, characterized in that, Methods for obtaining bubble growth rate include: The amplitude and phase of the original echo signal at each time point are substituted into the preset bubble radius and echo characteristic parameter correlation model to calculate the bubble radius value at the corresponding time point; the difference between the bubble radius values ​​at two adjacent time points is calculated to obtain the bubble radius difference value; the bubble radius difference value is divided by the time interval between two adjacent time points to obtain the bubble growth rate within the time interval between two adjacent time points.

4. The density uniformity control method based on the synergistic optimization of injection gas dosage and pressure according to claim 1, characterized in that, Methods for obtaining the cooling rate deviation coefficient include: Based on the dynamic prediction data of the mold temperature field in the prediction time domain, the cooling and solidification rate of the thick-walled area at different time nodes in the prediction time domain is calculated by the preset heat conduction model. The cooling and solidification rate corresponding to the time node with the highest shrinkage risk in the thick-walled area is selected as the calculation benchmark. The time node with the highest shrinkage risk is the time node corresponding to the synchronicity parameter reaching the maximum value. The cooling rate deviation coefficient is obtained by dividing the cooling solidification rate corresponding to the time point with the highest risk of shrinkage in the thick-walled region by the pre-obtained ideal cooling rate.

5. The density uniformity control method based on the synergistic optimization of injection gas dosage and pressure according to claim 1, characterized in that, Methods for obtaining the temperature gradient of the cooling water circuit include: The correlation coefficient between the cooling water temperature adjustment gradient and the high temperature maintenance time of the mold is denoted as the third correlation coefficient. The correlation coefficient between the cooling water temperature adjustment gradient and the cooling solidification rate is denoted as the fourth correlation coefficient. Multiply the extension of the high temperature holding time of the mold by the third correlation coefficient to obtain the first gradient adjustment component; Multiplying the cooling and solidification rate by the fourth correlation coefficient yields the second gradient adjustment component; Add the first gradient adjustment component to the second gradient adjustment component to obtain the initial value of the cooling water circuit temperature drop gradient; The pre-stored cooling water path temperature descent gradient constraint range in the multivariate coupled model is retrieved. If the initial value of the cooling water path temperature descent gradient is within the cooling water path temperature descent gradient constraint range, then the initial value of the descent gradient is the final cooling water path temperature descent gradient. If the initial value of the descent gradient exceeds the cooling water path temperature descent gradient constraint range, then the boundary value of the cooling water path temperature descent gradient constraint range is taken as the final cooling water path temperature descent gradient.

6. A density uniformity control system based on the coordinated optimization of injection gas dosage and pressure, used to implement the density uniformity control method based on the coordinated optimization of injection gas dosage and pressure as described in any one of claims 1-5, characterized in that, include: The data acquisition module is used to collect process parameters in real time. The scheme generation module is used to preprocess the process parameters and construct a virtual cavity model based on the process parameters; Defect prediction is performed based on a virtual cavity model to obtain the defect region; By combining the defective areas and their corresponding process parameters, an optimized decision-making scheme is generated; The correlation coefficient module is used to obtain the correlation coefficients between various process parameters using a pre-built multivariate coupling model; The strategy generation module calculates strategies for different defect types based on the correlation coefficients between various process parameters and the optimization decision scheme, and obtains the adjustment strategies for each process parameter.

Images