Injection molding process optimization method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of process control, in particular to an injection molding process optimization method and device, computer equipment and a storage medium, and the method comprises the following steps: obtaining stress signals of four tie bars of an injection molding machine, calculating a total mold locking force, and preprocessing the stress signals through temperature compensation and moving average filtering; real-time mold cavity pressure is calculated based on a mold filling stage linear model, a pressure maintaining stage quadratic function model and a cooling stage exponential attenuation model triggered by an injection molding machine control signal; generating a standard curve template with a confidence interval according to the characteristic parameters of the qualified sample, and dynamically updating the standard curve template; and comparing the current pressure curve with a standard template output quality judgment result in real time, and triggering automatic parameter adjustment to form closed-loop control. The device and the method have the effect of solving the problems of high cost, poor compatibility and difficulty in maintenance when the pressure of the die cavity is directly measured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of process control, and in particular to an injection molding process optimization method and device, computer equipment and a storage medium. BACKGROUND

[0002] In the field of injection molding quality detection and process control, mold cavity pressure is a core indicator reflecting the quality of injection molded parts, which directly affects the density, dimensional accuracy and mechanical properties of the product. With the continuous development of the injection molding industry, the requirements for the quality of injection molded parts are becoming higher and higher, and accurate monitoring of mold cavity pressure and optimization of injection molding process are particularly important. Accurate mold cavity pressure monitoring can help enterprises timely detect product quality problems, improve production efficiency and reduce production costs, which is of great significance to improving the market competitiveness of enterprises.

[0003] Currently, traditional mold cavity pressure detection mainly relies on manual sampling inspection or single threshold judgment. Manual sampling inspection requires operators to periodically extract a certain number of products for pressure detection during the production process. This method not only consumes a lot of manpower and time, but also the accuracy of the detection results is affected by the experience and subjective judgment of the operators. Single threshold judgment is to set a fixed pressure threshold, and when the mold cavity pressure exceeds or is lower than the threshold, the product is determined to be unqualified. This method ignores the pressure changes at different stages of the injection molding process. In addition, the method of directly measuring the mold cavity pressure is to embed a special pressure sensor in the mold, but this method has high cost, needs to modify the mold, and the sensor is easy to fail due to mold wear. At the same time, different mold structures differ greatly, and the installation position of the sensor needs to be customized and designed, which is difficult to be universal. Moreover, in the high-temperature and high-pressure environment of the mold, the sensor is prone to drift and needs to be calibrated frequently.

[0004] Traditional manual sampling inspection and single threshold judgment lack full-cycle dynamic monitoring, only focus on the pressure peak value, ignore the curve shape difference in the filling, pressure maintaining and cooling stages, and are prone to miss potential defects. The quality judgment is highly subjective and depends on the experience of the operators to judge whether the pressure curve is qualified, and the standard is not unified. The process adjustment is lagging, and after discovering defects, the operators need to analyze the reasons and adjust the parameters, which is low in efficiency and poor in accuracy, and is also disconnected with the control signals of the injection molding machine, cannot be associated with the injection, pressure maintaining and other process signals in real time, and is difficult to realize closed-loop optimization. The method of directly measuring the mold cavity pressure has high hardware cost, poor compatibility and difficult maintenance, which limits the economy and universality of mold cavity pressure monitoring. SUMMARY

[0005] In order to solve the problems of high cost, poor compatibility and difficult maintenance of directly measuring the mold cavity pressure, the present application provides an injection molding process optimization method and device, computer equipment and a storage medium.

[0006] The above invention purpose of the present application is achieved by the following technical solutions: An injection molding process optimization method, comprising: Acquiring stress signals of four columns of a injection molding machine and calculating total locking force, and pre-processing the stress signals through temperature compensation and sliding average filtering; Based on the linear model triggered by the filling stage control signal of the injection molding machine, the quadratic function model of the pressure maintaining stage and the exponential decay model of the cooling stage, the real-time mold cavity pressure is calculated; According to the characteristic parameters of the qualified sample, a standard curve template with a confidence interval is generated and dynamically updated; Real-time comparison of the current pressure curve with the standard template output quality judgment result, triggering parameter automatic adjustment to form a closed loop control.

[0007] By adopting the above technical solutions, the mold cavity pressure is indirectly measured by using the stress of the column, without the need to modify the mold, reducing the hardware cost, and the stress meter is easy to install and maintain; it is suitable for various injection molding machines and is not limited by the mold structure; the stress-pressure correlation model is established in stages, combined with dynamic template comparison, to improve the detection accuracy; directional adjustment suggestions are generated for different stages of abnormalities, combined with closed loop control, to improve the process adjustment efficiency and product qualification rate; the stress-pressure accurate conversion is realized through the stage-by-stage nonlinear model; the adaptive updating mechanism makes the template always match the current process state; the data-driven parameter optimization algorithm reduces the number of process adjustment iterations; it can be realized based on the existing stress meter and touch screen all-in-one machine without additional hardware modification.

[0008] In a preferred example, the present application can be further configured as follows: The pre-processing of the stress signals through temperature compensation and sliding average filtering specifically includes: Acquiring the ambient temperature by using a built-in temperature sensor; Compensating the original stress value by using a preset temperature-stress correction formula Δσ = 0.02 × με / ℃ × (T-25), and triggering automatic calibration when the compensation error exceeds a set threshold, wherein T is the ambient temperature collected by the temperature sensor.

[0009] By adopting the above technical solutions, the ambient temperature is collected by using a built-in temperature sensor, and the original stress value is compensated by using a preset temperature-stress correction formula, which can reduce the influence of ambient temperature on the stress signal of the column and improve the accuracy of the stress signal; when the compensation error exceeds a set threshold, automatic calibration is triggered, which can ensure the accuracy of temperature compensation, make the subsequent calculation of total locking force and calculation of mold cavity pressure based on stress signals more accurate, and further improve the detection accuracy and quality judgment accuracy of the injection molding process optimization method.

[0010] The application can be further configured in a preferred example as follows: the filling stage linear model, the holding stage quadratic function model and the cooling stage exponential decay model based on the injection molding machine control signal trigger to calculate the real-time mold cavity pressure, specifically including: The filling stage linear model adopts a linear relationship P1=a1xF_total+b1, wherein a1 and b1 are calibration coefficients; The holding stage quadratic function model adopts a quadratic function P2=a2xF_total2+b2xF_total+c2, wherein a2, b2 and c2 are calibration coefficients; The cooling stage exponential decay model adopts an exponential decay formula P3=a3xe(b3xt)+c3xF_total, wherein a3, b3 and c3 are calibration coefficients.

[0011] By adopting the above technical solution, the stress signals of the four columns of the injection molding machine are acquired and the total locking force is calculated, after temperature compensation and sliding average filtering pretreatment, the injection molding machine control signal triggers the models of different stages (filling stage linear model, holding stage quadratic function model and cooling stage exponential decay model) to calculate the real-time mold cavity pressure, the stress-pressure accurate conversion of the stage-by-stage nonlinear model is realized, the calculation error is ≤5%, the detection accuracy is improved, the standard curve template with a confidence interval is generated and dynamically updated according to the characteristic parameters of qualified samples, the current pressure curve is compared with the standard template in real time, the quality judgment result is output, the parameter automatic adjustment is triggered to form a closed loop control, and the process adjustment efficiency and product qualification rate can be improved.

[0012] The application can be further configured in a preferred example as follows: the standard curve template with a confidence interval is generated and dynamically updated according to the characteristic parameters of qualified samples, specifically including: The timing update is performed every N mold times, and when the deviation degree of the new sample data from the current template exceeds 5%, a new template is generated by a weighted mean method and the historical version is retained, wherein N is an integer greater than or equal to 1000; The quality judgment adopts a stage-by-stage weighted scoring, and when the total matching degree is lower than 0.8 or the score of any stage is lower than 0.7, the unqualified part is judged.

[0013] By adopting the above technical solution, the indirect measurement of the mold cavity pressure by the column stress can reduce the hardware cost, is easy to install and maintain, and has strong universality; the stress-pressure correlation model is established in stages, and the dynamic template comparison can improve the detection accuracy; the standard curve template is updated at a fixed time, a new template is generated by a weighted mean method when the deviation degree of the new sample data from the current template exceeds 5%, and the historical version is retained, the adaptive updating mechanism makes the template always match the current process state, and the judgment accuracy is improved; the quality judgment mode of the stage-by-stage weighted scoring can more accurately judge the quality of the injection molded part.

[0014] The application can be further configured in a preferred example that the real-time comparison of the current pressure curve and the standard template output quality determination result triggers automatic parameter adjustment to form a closed-loop control, specifically including: When the same abnormality is determined twice in succession, automatic parameter adjustment is triggered, and if the abnormality still exists after adjustment, the adjustment amplitude is increased to 1.5 times of the original value, and when the abnormality occurs five times in succession, an audible and light alarm is triggered and production is suspended; The data tracing system stores the original stress curve, the calculated pressure curve and the adjustment record of each mold cycle, uses a time series database and associates the Gelin column number, the injection molding machine number, the mold number and the operator ID.

[0015] By using the above technical solution, when the same abnormality is determined twice in succession, automatic parameter adjustment is triggered, and if the abnormality still exists after adjustment, the adjustment amplitude is increased, and when the abnormality occurs five times in succession, an audible and light alarm is triggered and production is suspended, the intelligent closed-loop control of the injection molding process can be realized, and the process adjustment efficiency and product qualification rate are improved; the data tracing system is used to store the related curves and adjustment records of each mold cycle, a time series database is used and multiple numbers and IDs are associated, which facilitates data storage, query and tracing, and improves the convenience and accuracy of injection molding production management.

[0016] The above-mentioned second application object of the application is realized by the following technical solution: An injection molding process optimization device, the injection molding process optimization device comprising: A data preprocessing module for acquiring stress signals of four Gelin columns of an injection molding machine and calculating total locking force, and pre-processing the stress signals through temperature compensation and sliding average filtering; A pressure calculation module for calculating real-time mold cavity pressure based on a linear model triggered by an injection molding machine control signal in the filling stage, a quadratic function model in the pressure maintaining stage and an exponential decay model in the cooling stage; A parameter updating module for generating a standard curve template with a confidence interval according to the characteristic parameters of qualified samples and dynamically updating it; An injection molding control module for real-time comparison of the current pressure curve and the standard template output quality determination result, triggering automatic parameter adjustment to form a closed-loop control.

[0017] By adopting the technical scheme, the cavity pressure is indirectly measured by using the coring column stress, the mold does not need to be modified, the hardware cost is reduced, and the stress gauge is easy to install and maintain; the method is suitable for various injection molding machines and is not limited by the mold structure; a stress-pressure correlation model is established in stages, dynamic template comparison is combined, and the detection accuracy is improved; directional adjustment suggestions are generated for different stages of abnormalities, closed-loop control is combined, and the process adjustment efficiency and product qualification rate are improved; the stress-pressure accurate conversion is realized through a staged nonlinear model; the adaptive updating mechanism makes the template always match the current process state; the data-driven parameter optimization algorithm reduces the number of process adjustment iterations; the method can be realized based on the existing stress gauge and touch screen all-in-one machine, and no additional hardware modification is needed.

[0018] The fourth purpose of the present application is achieved by the following technical scheme: A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned injection molding process optimization method when executing the computer program.

[0019] The fourth purpose of the present application is achieved by the following technical scheme: A computer-readable storage medium stores a computer program, and the computer program implements the steps of the above-mentioned injection molding process optimization method when executed by a processor.

[0020] In summary, the present application includes at least one of the following beneficial technical effects: 1. Significant cost advantage, indirect measurement of cavity pressure by using coring column stress, no need to modify the mold, reduced hardware cost, and stress gauge easy to install and maintain; 2. Strong universality, suitable for various injection molding machines, not limited by the mold structure, and short calibration time after replacing the mold; 3. High detection accuracy, stress-pressure correlation model established in stages, dynamic template comparison combined, and improved determination accuracy; 4. Intelligent process optimization, directional adjustment suggestions generated for different stages of abnormalities, closed-loop control combined, and improved process adjustment efficiency and product qualification rate. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a flowchart of the injection molding process optimization method in an embodiment of the present application; Figure 2 is a principle block diagram of the injection molding process optimization system in an embodiment of the present application; Figure 3 is a device schematic diagram in an embodiment of the present application. DETAILED DESCRIPTION

[0022] The application will be further described in detail below with reference to the accompanying drawings.

[0023] In an embodiment, as shown in the accompanying drawings, Figure 1 The injection molding process optimization method disclosed by the application specifically comprises the following steps: S10: Obtain the stress signals of the four columns of the injection molding machine and calculate the total locking force. The stress signals are preprocessed through temperature compensation and sliding average filtering.

[0024] Specifically, during the injection molding production process, the column is the core component of the locking mechanism of the injection molding machine, and the stress change thereof has a strong correlation with the cavity pressure. Therefore, obtaining the stress signals of the four columns of the injection molding machine is the basis for indirectly monitoring the cavity pressure. Specifically, a pasting type strain gauge (precision ±1με) or a wireless stress sensor (Bluetooth 5.0, sampling frequency 100Hz) is installed along the axial direction of the middle section of the four columns (avoiding the stress concentration area), and the adhesion degree with the column surface is ensured to be >95%. Before installation, the surface of the column needs to be pretreated, including polishing (roughness Ra1.6), cleaning (wiping with anhydrous ethanol), and coating primer (enhancing adhesion), so as to ensure that the stress gauge can accurately collect stress signals.

[0025] The collected original stress signals will be affected by various interference factors, such as power frequency interference and transient noise. In order to improve the quality of the signals, preprocessing is needed. First, the original stress signals are subjected to 50Hz low-pass filtering (Butterworth filter, order 4), to suppress power frequency interference. Then, sliding average filtering (window size 5 sampling points) is performed to remove transient noise and improve the smoothness of the signals by ≥40%. In addition, the change of the environmental temperature will also affect the stress measurement, so temperature compensation is needed. The environmental temperature is collected by the temperature sensor (precision ±0.5℃) built in the stress gauge, and the stress value is corrected based on the preset temperature-stress correction formula (Δσ = 0.02×με / ℃×(T - 25), T is the measured temperature). The compensation coefficient is automatically calibrated every 30 minutes, to ensure that the compensation error is ≤3%.

[0026] When calculating the total locking force, the locking force of a single column is calculated first. The formula is F_i = σ_i×E×S×K, where F_i is the locking force (kN) of the i-th column, σ_i is the corrected stress value (MPa, converted from με: 1με = 1MPa / E×10 6E represents the Young's modulus of the tie rod (GPa, set via touchscreen, default 206GPa, adjustable according to the material manual, adjustment step 0.1GPa), S represents the cross-sectional area of ​​the tie rod (mm², S = π×(D / 2)², D is the diameter, obtained through parameter settings, accuracy ±0.1mm), and K is the structural correction coefficient (K = 1.0 without drilling; K = 0.9 - 0.002×d / D with drilling, d is the drilling diameter, value range 5 - 20mm). Then, the total clamping force F_total = ΣF_i (i = 1 - 4, taking the combined force of 4 rods). When the data of a certain rod deviates from the mean by >20%, it is judged as an outlier, and the mean of the other 3 rods is used to complete the calculation, with a completion error ≤5%.

[0027] Through the stress signal acquisition and preprocessing steps described above, the stress signal of the tie rod can be accurately and stably acquired, and the total clamping force can be calculated, providing a reliable data basis for subsequent mold cavity pressure calculation. This preprocessing method not only improves the signal quality but also reduces the influence of environmental factors on the measurement results, thereby improving the accuracy and reliability of the entire monitoring system.

[0028] S20: Real-time mold cavity pressure is calculated based on the linear model of the filling stage, the quadratic function model of the holding stage, and the exponential decay model of the cooling stage triggered by the injection molding machine control signal.

[0029] Specifically, calculating the real-time mold cavity pressure based on the linear model of the filling stage triggered by the injection molding machine control signal, the quadratic function model of the holding stage, and the exponential decay model of the cooling stage is one of the key aspects of this method. Since the stress change of the tie rods and the mold cavity pressure are strongly nonlinearly coupled, the error of the traditional linear model is >20%, thus requiring staged modeling to address this issue.

[0030] During the mold filling stage, 10 sets of sample experiments (adjusting parameters such as injection speed and pressure to obtain different clamping forces and corresponding mold cavity pressures) were conducted. The least squares method was used to fit the linear relationship between clamping force and mold cavity pressure: P1 = a1×F_total + b1, where a1 and b1 are calibration coefficients, and the goodness of fit R² ≥ 0.95. During the holding pressure stage, considering the nonlinear transmission of pressure, a quadratic function model was adopted: P2 = a2×F_total² + b2×F_total + c2. The coefficients a2, b2, and c2 were determined through 20 sets of sample experiments, and the goodness of fit R² ≥ 0.93. During the cooling stage, based on the exponential decay model P3 = a3×e^(b3×t) + c3×F_total (t being the time after the holding pressure ends), the coefficients a3, b3, and c3 were determined through 15 sets of sample experiments, and the goodness of fit R² ≥ 0.90.

[0031] The system automatically calls the corresponding model according to the current process stage (mold filling / pressure maintaining / cooling), triggers the model switching through the stage switching signals (such as injection start signal, pressure maintaining switching signal) sent by the injection molding machine, the switching response time is ≤10 ms, and the output cavity pressure value (precision ±5%). This way of modeling in stages fully considers the nonlinear relationship between the cavity pressure and the clamping force in different process stages, improving the accuracy of the cavity pressure calculation. Compared with the traditional linear model, the staged model can better adapt to the characteristics of different process stages, thereby reducing the calculation error and providing a more accurate basis for the optimization of the injection molding process.

[0032] S30: generating a standard curve template with a confidence interval according to the qualified sample characteristic parameters and dynamically updating.

[0033] Specifically, generating a standard curve template with a confidence interval according to the qualified sample characteristic parameters and dynamically updating is an important means to realize dynamic judgment of injection molded part quality. First, sample screening is performed, 300 pieces of qualified products (no shrinkage, flash and other defects in appearance detection + dimensional tolerance ≤0.1 mm) are selected for continuous production, the first 50 pieces (device preheating stage) and abnormal fluctuation pieces (single mold curve deviation from the mean value >8%) are removed, and 200 effective samples are retained.

[0034] Further, the characteristic parameters are extracted, including time node parameters and pressure characteristic parameters. The time node parameters include the mold filling start time t0 (the rising edge of the injection signal triggers), the pressure maintaining switching time t1 (the rising edge of the pressure maintaining signal triggers), the cooling start time t2 (the falling edge of the pressure maintaining signal triggers), and the mold opening time t3 (the rising edge of the mold opening signal triggers). The pressure characteristic parameters are different in different stages. The mold filling stage includes the rising slope k1 (MPa / s), the peak pressure Pmax1, and the peak time t_peak1. The pressure maintaining stage includes the fluctuation standard deviation σ_p (MPa), the average pressure P_avg2, and the pressure decay rate r2 (MPa / s). The cooling stage includes the decay slope k3 (MPa / s), the N-second pressure residual value P_res3, N (which can be set), and the pressure stable time t_stable3.

[0035] Then, the template is constructed. Taking "mold filling start" as the reference point (t = 0), the time axis of all samples is unified through interpolation algorithm, and the time deviation of each stage is ≤0.1 s. The mean value μ of the characteristic parameters of 200 samples is taken, the standard deviation σ is calculated, and the upper and lower thresholds are μ±3σ, forming a template curve with a confidence interval, and the interval coverage is ≥99.7%.

[0036] In terms of dynamic updating, there are two ways: timing update and abnormal triggering. Timing update automatically starts the update process every 1000 production cycles, and the update time is ≤30 seconds, which does not affect normal production. When the average of 50 consecutive cycles deviates from the template by >5%, the update is triggered, and a prompt box is popped up to inform the operator. The update algorithm processes differently according to the deviation D of the new sample from the current template. If D≤5%, only the template threshold (±3σ_new, σ_new is the standard deviation of the new sample) is updated, and the original mean μ is retained; if 5%<D≤10%, the original template is retained as the "historical template", and a new template is generated as the "current template", supporting switching comparison, and the new template parameter mean μ_new = 0.7×μ + 0.3×μ_recent (μ_recent is the mean of the new sample); if D>10%, it prompts "process condition changes dramatically, suggest recalibration", which needs to be confirmed by the operator before rebuilding the template, and the rebuilding process automatically calls the calibration process.

[0037] S40: Real-time comparison of the current pressure curve with the standard template output quality determination result, triggering parameter automatic adjustment to form a closed-loop control.

[0038] Specifically, real-time comparison of the current pressure curve with the standard template output quality determination result, triggering parameter automatic adjustment to form a closed-loop control, is the core of intelligent optimization of injection molding process. The quality determination is performed by using a phased weighted scoring method, which automatically divides the four stages by the injection molding machine signal, and the matching weights of each stage are different (30% for filling, 40% for pressure maintaining, 20% for cooling, and 10% for opening). The weights can be adjusted in the system settings, with an adjustment range of 5%-50%.

[0039] For each sampling point P(t) of the current curve, the deviation from the template interval [P_low(t), P_high(t)] is calculated, and then the stage matching degree and the total matching degree are calculated. The full score is 1.0, with 3 decimal places. The quality determination standard is divided into qualified parts, suspicious parts and unqualified parts. Qualified parts require Stotal≥0.9 and Sj≥0.85 in each stage; suspicious parts are 0.8≤Stotal<0.9 or 0.7≤Sj<0.85 in any stage; unqualified parts are Stotal<0.8 or Sj<0.7 in any stage. For suspicious / unqualified parts, the specific abnormal points (such as "pressure exceeds the upper limit by 12% at 15s in the pressure maintaining stage") are located, and are associated with the corresponding process parameters (such as pressure maintaining pressure, speed), and the marked information is displayed in red flashing on the touch screen, with a duration of 10 seconds or until the operator confirms.

[0040] In terms of closed-loop control, when two consecutive mold times are determined as "suspicious parts" and the main cause is consistent, automatic adjustment is triggered, and the parameter adjustment algorithm is called. The parameter adjustment algorithm includes adjustment amount calculation and priority planning. Adjustment amount calculation is performed for parameters such as injection speed, holding pressure, holding time, and injection volume, and each parameter after adjustment does not exceed the maximum limit of the device. In terms of priority planning, parameters that affect product appearance (such as injection speed) are adjusted first, with an adjustment response time of ≤2 seconds. Parameters with an adjustment amount <5% can be automatically executed, and >5% require manual confirmation. The confirmation interface displays the comparison of parameters before and after adjustment and the expected effect, and simultaneously outputs the expected effect after adjustment (such as "increase holding pressure by 3 MPa, expected σ_p fluctuation from 5 MPa to 3 MPa"). The expected effect is based on historical adjustment data fitting, with an error ≤10%.

[0041] After adjustment, 3 mold times are monitored, and if they return to "qualified", the parameters are fixed, and after parameter fixation, "optimized" is marked in the system log; if it is still abnormal, upgrade the suggestion (such as increase the adjustment amount or switch the adjustment parameter), the upgrade amplitude is 1.5 times the first adjustment amount; when 5 consecutive mold times are still unqualified, the system automatically suspends production and sends an audible and visual alarm (alarm sound level 80 dB, lasting 30 seconds), prompting the operator to check the equipment and mold.

[0042] Through this real-time comparison and closed-loop control mechanism, quality problems in the injection molding process can be detected in a timely manner, and process parameters can be automatically adjusted to achieve intelligent optimization of the injection molding process. The closed-loop control method can continuously optimize the process parameters, improve the product qualification rate, and reduce the scrap rate, thereby reducing production costs and improving production efficiency. At the same time, the data traceability system stores the original stress curve, calculated pressure curve, and adjustment record of each mold time, uses a time series database and associates the Gelin column number, injection molding machine number, mold number, and operator ID, supports multi-dimensional combination queries, and facilitates the traceability and analysis of the production process, further improving the management level of the entire injection molding process.

[0043] In an embodiment, in step S10, the stress signal is preprocessed by temperature compensation and moving average filtering, specifically including: S11: Collecting the ambient temperature through the built-in temperature sensor.

[0044] Specifically, the ambient temperature is collected through the built-in temperature sensor. The built-in temperature sensor has high precision, with an accuracy of ±0.5°C, which enables it to accurately perceive the subtle changes in the ambient temperature. In terms of installation, the temperature sensor is reasonably placed inside the stress meter, which ensures that the measured temperature truly reflects the ambient temperature of the stress meter, as the stress meter is in close contact with the Gelin column, and the ambient temperature directly affects the stress of the Gelin column.

[0045] During the process of collecting the ambient temperature, the sensor will continuously collect data. An accurate temperature value will be obtained every sampling period, which will serve as an important basis for subsequent temperature compensation. Since the ambient temperature may fluctuate during the injection molding process, continuous collection can ensure timely capture of temperature changes, thereby providing real-time data support for accurate correction of stress values.

[0046] S12: Compensate the original stress value using the preset temperature-stress correction formula Δσ = 0.02 × με / ℃ × (T - 25), and trigger automatic calibration when the compensation error exceeds the set threshold, where T is the ambient temperature collected by the temperature sensor.

[0047] Specifically, the original stress value is compensated using the preset temperature-stress correction formula Δσ = 0.02 × με / ℃ × (T - 25). This formula is derived from a large number of experiments and data analysis, and it can accurately describe the relationship between temperature change and stress value. Where T is the ambient temperature collected by the temperature sensor, and 25℃ is a reference temperature value. When the ambient temperature T is not equal to 25℃, a compensation value Δσ is calculated according to the formula, and then the compensation value is applied to the original stress value to obtain the corrected stress value.

[0048] In practical applications, the system will automatically substitute the collected ambient temperature into the formula for calculation. For example, if the collected ambient temperature T is 30℃, then substituting the formula gives Δσ = 0.02 × με / ℃ × (30 - 25) = 0.1 × με. This compensation value will be accurately added to the original stress value to eliminate the influence of temperature on stress measurement.

[0049] When the compensation error exceeds the set threshold, automatic calibration is triggered. The set threshold is to ensure the accuracy of compensation. If the compensation error is too large, the corrected stress value may have a large deviation, thereby affecting the subsequent cavity pressure calculation and process optimization. The system will monitor the compensation error in real time and compare the compensation value obtained each time with the actual stress change.

[0050] Once it is found that the compensation error exceeds the set threshold, such as the compensation error exceeds 3%, the system will immediately trigger the automatic calibration mechanism. The automatic calibration process will recheck the accuracy of the temperature sensor to ensure that the measured ambient temperature is accurate. At the same time, the preset temperature-stress correction formula will be verified again to check whether the parameters in the formula are still applicable. If it is found that the parameters need to be adjusted, the system will optimize according to the historical data and the current actual situation to ensure that the subsequent compensation can be more accurate.

[0051] During the automatic calibration process, the system will pause the current stress data acquisition and processing, and focus on the calibration operation. After the calibration is completed, the system will resume the normal workflow, continue to acquire the ambient temperature and perform temperature compensation to ensure that the accuracy of the stress signal is always at a high level.

[0052] Optionally, in step S20, the real-time mold cavity pressure is calculated based on the injection molding machine control signal triggered mold filling stage linear model, holding stage quadratic function model and cooling stage exponential decay model, specifically including: S21: The mold filling stage linear model adopts a linear relationship P1 = a1 x F_total + b1, where a1 and b1 are calibration coefficients.

[0053] Specifically, in the mold filling stage, a linear relationship P1 = a1 x F_total + b1 is used to construct a linear model. In order to determine the calibration coefficients a1 and b1, a series of rigorous sample experiments need to be conducted. Specifically, the injection speed, pressure and other key parameters will be adjusted to obtain different clamping forces and corresponding mold cavity pressures. By carefully selecting 10 such sample experimental data, the least squares method is used for fitting. The least squares method is a very effective mathematical method, which can make the fitted linear relationship as close as possible to the actual relationship between the clamping force and the mold cavity pressure. After such a fitting process, the linear relationship obtained can well reflect the linear correlation between the clamping force and the mold cavity pressure in the mold filling stage, and the goodness of fit R² can reach ≥0.95. This means that the linear model has high reliability and accuracy in the mold filling stage, and can accurately calculate the mold cavity pressure. In the actual injection molding production process, the system will automatically trigger the call of this mold filling stage linear model according to the injection start signal and other stage switching signals sent by the injection molding machine. Once the corresponding signal is received, the system will respond quickly, with a response time ≤10 ms, and then calculate the real-time mold cavity pressure value according to the current total clamping force F_total using the calibrated coefficients a1 and b1 according to the linear relationship P1 = a1 x F_total + b1, and the accuracy of the pressure value can be controlled within ±5%.

[0054] S22: The holding stage quadratic function model adopts a quadratic function P2 = a2 x F_total² + b2 x F_total + c2, where a2, b2 and c2 are calibration coefficients.

[0055] Specifically, in the holding stage, since the pressure transmission presents a nonlinear characteristic, a quadratic function P2 = a2xF_total2 + b2xF_total + c2) is used to construct the model. Similarly, in order to determine the calibration coefficients a2, b2 and c2, more detailed sample experiments need to be carried out. This stage will carry out 20 groups of sample experiments, and through the analysis and processing of these experimental data, the most suitable quadratic function model is determined. Through a large number of experiments and accurate calculation, the goodness of fit R2 of this quadratic function model can reach ≥0.93. This shows that the model can well capture the nonlinear relationship between the mold cavity pressure and the locking force in the holding stage. In the holding stage of injection molding production, when the injection molding machine sends a holding switching signal, the system will immediately automatically call this quadratic function model of the holding stage. The system responds quickly, and within ≤10ms, according to the current total locking force F_total, combined with the already determined calibration coefficients a2, b2 and c2, the real-time mold cavity pressure is accurately calculated through the quadratic function (P2 = a2xF_total2 + b2xF_total + c2, and the accuracy of the pressure value can still be guaranteed at ±5%.

[0056] S23: Exponential decay model in cooling stage P3 = a3xe(b3xt) + c3xF_total, wherein a3, b3 and c3 are calibration coefficients.

[0057] Specifically, in the cooling stage, the mold cavity pressure will present an exponential decay trend with time, so the exponential decay formula P3 = a3xe(b3xt) + c3xF_total is used to construct the model. Here t is the time after the holding ends. In order to determine the calibration coefficients a3, b3 and c3, 15 groups of sample experiments will be carried out. Through in-depth research and analysis of these experimental data, the most suitable exponential decay model is determined. The goodness of fit R2 of this model can reach ≥0.90, which shows that it can well describe the change law of the mold cavity pressure with time and the locking force in the cooling stage. In the cooling stage of injection molding production, when the injection molding machine sends a signal indicating that the cooling starts, such as the falling edge of the holding signal, the system will quickly automatically call the exponential decay model in the cooling stage. The system responds within ≤10ms, according to the current total locking force F_total and the time t after the holding ends, and uses the already determined calibration coefficients a3, b3 and c3, to accurately calculate the real-time mold cavity pressure through the exponential decay formula P3 = a3xe(b3xt) + c3xF_total, and the pressure value accuracy is also controlled at ±5%.

[0058] In an embodiment, in step S30, the standard curve template with confidence interval is generated according to the qualified sample characteristic parameters and is dynamically updated, specifically including: S31: The timing update is performed every N mode, and when the deviation of the new sample data from the current template exceeds 5%, a new template is generated by the weighted mean method and the historical version is retained, wherein N is an integer greater than or equal to 1000.

[0059] Specifically, in the standard curve template generation stage, sample screening is first performed. 300 pieces of qualified products in continuous production are selected, which need to be detected for appearance without shrinkage, flash and other defects, and the size tolerance is ≤0.1 mm. This strict screening standard ensures that the samples used to generate the template have high quality and representativeness. Then, the first 50 products are removed because the products produced during the preheating stage of the equipment may have unstable factors. At the same time, abnormal fluctuation pieces, i.e. products with a deviation of >8% from the mean value of single mode curve, are removed, and finally 200 effective samples are retained.

[0060] Then feature parameter extraction is performed. In terms of time node parameters, the filling start time t0 is triggered by the rising edge of the injection signal, the holding switch time t1 is triggered by the rising edge of the holding signal, the cooling start time t2 is triggered by the falling edge of the holding signal, and the mold opening time t3 is triggered by the rising edge of the mold opening signal. These accurate time node parameters provide an important time reference for subsequent template construction. In terms of pressure feature parameters, in the filling stage, the rising slope k1, peak pressure Pmax1 and peak time t_peak1 are extracted; in the holding stage, the fluctuation standard deviation σ_p, average pressure P_avg2 and pressure decay rate r2 are extracted; in the cooling stage, the decay slope k3, N-second pressure residual value P_res3,N (which can be set) and pressure stabilization time t_stable3 are extracted. These pressure feature parameters comprehensively reflect the pressure changes in different stages of the injection molding process.

[0061] Then template construction is performed. Taking "filling start" as the reference point (t = 0), the time axis of all samples is unified by interpolation algorithm, so that the time deviation of each stage is ≤0.1 s. This can ensure the consistency of different samples on the time axis, facilitating subsequent parameter comparison and analysis. Then the mean value μ of the feature parameters of the 200 samples is taken, the standard deviation σ is calculated, and the upper and lower thresholds are μ±3σ, forming a template curve with a confidence interval, and the interval coverage is ≥99.7%. This template curve with a confidence interval can more accurately reflect the pressure change range of qualified products, improving the accuracy of quality determination.

[0062] In terms of standard curve template dynamic updating, the timing update every N mode (N is an integer greater than or equal to 1000) automatically starts the update process, and the update time is less than or equal to 30 seconds, which does not affect normal production. This timing update mechanism can ensure that the template is always synchronized with the actual production situation and adapt to small changes in the production process. When the average of the continuous 50 modes is greater than 5% deviation from the template, the update is forcibly triggered, and a prompt box is popped up to inform the operator. This abnormal triggering mechanism can timely detect abnormal situations in the production process and ensure the effectiveness of the template.

[0063] In terms of update algorithm, the deviation D of the last 200 qualified pieces from the current template is calculated (D=(∑|x_i - μ| / n) / μ, x_i is the new sample parameter, n is the number of parameters, n = 9). If D≤5%, only update the template threshold (±3σ_new, σ_new is the new sample standard deviation), and keep the original mean μ. This way can quickly adjust the threshold range of the template when the deviation is small, without changing the mean value, to adapt to small fluctuations in the production process. If 5% < D≤10%, keep the original template as "history template" and generate a new template as "current template" to support switching comparison, and the new template parameter mean μ_new = 0.7×μ + 0.3×μ_recent (μ_recent is the new sample mean). In this way, when the deviation is large, the history template is kept as a reference, and a new template is generated to adapt to the new production situation. If D>10%, prompt "process condition changes dramatically, suggest recalibration", and manual confirmation is required to rebuild the template, and the rebuilding process automatically calls the calibration process. This processing method can cope with major changes in process conditions and ensure the accuracy and reliability of the template.

[0064] S32: Quality determination adopts phased weighted scoring, and when the total matching degree is less than 0.8 or the score of any stage is less than 0.7, it is determined as unqualified piece.

[0065] Specifically, in terms of quality determination, phased weighted scoring is adopted. Four stages are automatically divided by the injection molding machine signal, and the matching weights of each stage are different (30% for mold filling, 40% for pressure holding, 20% for cooling, and 10% for mold opening). The weight can be adjusted in the system settings, and the adjustment range is 5%-50%. This phased weighted scoring method can more comprehensively consider the importance of different stages in the injection molding process and improve the accuracy of quality determination. For each sampling point P(t) of the current curve, the deviation from the template interval [P_low(t), P_high(t)] is calculated, and then the stage matching degree and the total matching degree are calculated. When the total matching degree is less than 0.8 or the score of any stage is less than 0.7, it is determined as unqualified piece. This clear determination standard can quickly and accurately identify unqualified products and provide a basis for subsequent parameter adjustment and process optimization.

[0066] In an embodiment, in step S40, the current pressure curve is compared with the standard template output quality determination result in real time, triggering automatic parameter adjustment to form a closed-loop control, specifically including: S41: When the same abnormality is determined twice in a row, automatic parameter adjustment is triggered. If the abnormality still exists after adjustment, the adjustment amplitude is increased to 1.5 times the original value. When the abnormality occurs five times in a row, an audible and light alarm is triggered and production is suspended.

[0067] Specifically, in the abnormality determination and parameter adjustment link, the system continuously monitors the mold cavity pressure curve during the injection molding production process. When two consecutive mold times are determined as "suspicious parts" and the main reason is consistent, the automatic adjustment mechanism is triggered. Once triggered, the system will immediately call the pre-set parameter adjustment algorithm. In this algorithm, the type of abnormality is first accurately classified.

[0068] If the abnormality is in the filling stage, such as the rising slope k1 < threshold value, the system will analyze the possible reasons for "inadequate injection speed" (probability 70%) or "low melt temperature" (probability 30%). The system will determine the main reason by comparing the "slope change rate after speed adjustment" in historical data. When a 1% speed adjustment results in a slope change of >0.5%, it is determined that "inadequate injection speed" is the main reason. If the peak pressure Pmax1 > threshold value, the possible reasons are "excessive injection volume" (probability 60%) or "low mold temperature" (probability 40%). When a 1% injection volume adjustment results in a peak pressure change of >1%, it is determined that "excessive injection volume" is the main reason.

[0069] For abnormality in the holding stage, if the fluctuation σ_p > threshold value, the main reason is "unstable holding pressure" (probability 80%). The system will associate the injection molding machine hydraulic system pressure fluctuation data (collection frequency 10 Hz), and confirm when the correlation between hydraulic fluctuation and mold cavity pressure fluctuation is >0.8. If the decay rate r2 > threshold value, the main reason is "insufficient holding time" (probability 75%) or "low holding pressure" (probability 25%). When the holding time is extended by 1 second, the decay rate decreases by >5%, it is determined that "insufficient holding time" is the main reason.

[0070] After determining the main reason for the abnormality, the system calculates the adjustment amount. For injection speed adjustment, a safety factor of 0.8 is multiplied, and the adjusted speed does not exceed 90% of the maximum speed of the device. The holding pressure adjustment does not exceed 90% of the maximum holding pressure of the device. The holding time adjustment does not exceed the pre-set maximum cooling time. The injection volume adjustment step is 0.1 cm³. At the same time, the system will prioritize the parameter adjustment. Parameters that affect the appearance of the product (such as injection speed) are adjusted first, and the adjustment response time is ≤2 seconds. Parameters with an adjustment amount <5% can be automatically executed, and parameters with an adjustment amount >5% require manual confirmation. The confirmation interface displays the parameter comparison before and after adjustment and the expected effect. The expected effect is based on historical adjustment data fitting and the error is ≤10%.

[0071] After adjustment, the system will continue to monitor 3 times. If the product returns to the "qualified" state, the system will solidify the adjusted parameters, and after parameter solidification, it will be marked as "optimized" in the system log. If there is still an abnormality after adjustment, the system will upgrade the suggestion, such as increasing the adjustment amount or switching the adjustment parameter, with an upgrade of 1.5 times the first adjustment amount. When the continuous 5 times of judgment results are abnormal, the system will trigger an audible and light alarm (alarm sound level 80dB, lasting 30 seconds), and pause production, prompting the operator to check the equipment and mold.

[0072] S42: The data tracing system stores the original stress curve, the calculated pressure curve, and the adjustment record of each mold. A time series database is used, and the column number, injection molding machine number, mold number, and operator ID are associated.

[0073] Specifically, in terms of data tracing, the data tracing system plays a crucial role. The system stores the original stress curve, the calculated pressure curve, and the adjustment record of each mold. A time series database (such as InfluxDB) is used for data storage, which can efficiently handle time series data, ensuring data storage and query efficiency. At the same time, the system associates column numbers, injection molding machine numbers, mold numbers, and operator IDs, etc. Through these associated information, operators can easily trace and analyze production data. For example, when there is a product quality problem, the data tracing system can be queried to quickly locate which injection molding machine, which mold, and which operator had problems in the production process during which time period. The system supports multi-dimensional combined queries such as product number, production time, mold number, etc., with a query response time ≤3 seconds, and the query results are displayed in the form of a list + chart, providing intuitive and comprehensive data information for operators to facilitate in-depth analysis and optimization of the production process. This closed-loop control and data tracing mechanism effectively improves the stability of the injection molding process and product quality, reduces production costs, and improves production efficiency.

[0074] It should be understood that the size of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0075] In an embodiment, an injection molding process optimization device is provided, which corresponds to the injection molding process optimization method in the above embodiments. As shown in the figure, the injection molding process optimization device includes a data preprocessing module, a pressure calculation module, a parameter updating module, and an injection molding control module. The functions of each module are described in detail as follows: Figure 2 The data preprocessing module is used to acquire the stress signals of the four columns of the injection molding machine and calculate the total locking force, and the stress signals are preprocessed through temperature compensation and sliding average filtering. The data preprocessing module is used to acquire the stress signals of the four columns of the injection molding machine and calculate the total locking force, and the stress signals are preprocessed through temperature compensation and sliding average filtering. a pressure calculation module, configured to calculate real-time mold cavity pressure based on a linear model of a filling stage triggered by an injection molding machine control signal, a quadratic function model of a pressure maintaining stage, and an exponential decay model of a cooling stage; a parameter updating module, configured to generate a standard curve template with a confidence interval according to qualified sample characteristic parameters and to dynamically update the standard curve template; an injection control module, configured to compare a current pressure curve with a quality determination result output by the standard curve template in real time, and to trigger automatic parameter adjustment to form a closed-loop control.

[0076] Optionally, the data preprocessing module comprises: a temperature acquisition submodule, configured to acquire an ambient temperature through a built-in temperature sensor; a calibration submodule, configured to compensate an original stress value by using a preset temperature-stress correction formula Δσ = 0.02 × με / ℃ × (T-25), and to trigger automatic calibration when a compensation error exceeds a set threshold, wherein T is the ambient temperature acquired by the temperature sensor.

[0077] Optionally, the pressure calculation module comprises: a filling stage calculation submodule, configured to use a linear relationship formula P1 = a1 × F_total + b1 for the linear model of the filling stage, wherein a1 and b1 are calibration coefficients; a pressure maintaining stage calculation submodule, configured to use a quadratic function formula P2 = a2 × F_total2 + b2 × F_total + c2 for the quadratic function model of the pressure maintaining stage, wherein a2, b2, and c2 are calibration coefficients; a cooling stage calculation submodule, configured to use an exponential decay formula P3 = a3 × e^(b3 × t) + c3 × F_total for the exponential decay model of the cooling stage, wherein a3, b3, and c3 are calibration coefficients.

[0078] Optionally, the parameter updating module comprises: a cyclic updating submodule, configured to update every N moldings, and to generate a new template by using a weighted mean method and to retain a historical version when a deviation degree between new sample data and a current template exceeds 5%, wherein N is an integer greater than or equal to 1000; a weighted scoring submodule, configured to use a phased weighted scoring for quality determination, and to determine a non-qualified part when a total matching degree is lower than 0.8 or a score of any stage is lower than 0.7.

[0079] Optionally, the injection control module comprises: a parameter adjustment submodule, configured to trigger automatic parameter adjustment when the same abnormality is determined for two consecutive times, to adjust the parameter to 1.5 times of an original value if the abnormality still exists after adjustment, and to trigger an audible-light alarm and to suspend production when the abnormality is determined for five consecutive times; A data traceability sub-module is configured to store original stress curves, calculated pressure curves and adjustment records of each mold cycle in a data traceability system, and uses a time series database and associates the column number, injection molding machine number, mold number and operator ID.

[0080] The specific limitations of the injection molding process optimization device can refer to the limitations of the injection molding process optimization method described above, which will not be repeated here. Each module in the above injection molding process optimization device can be realized by software, hardware and their combination. The above modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor calls and executes the operations corresponding to each module.

[0081] In one embodiment, a computer device, which can be a server, is provided, and its internal structure diagram can be as shown in Figure 3 The computer device includes a processor, a memory, a network interface and a database connected by a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with external terminals through network connection. The computer program is executed by the processor to implement an injection molding process optimization method.

[0082] In one embodiment, a computer device is provided, including a memory, a processor and a computer program stored on the memory and executable on the processor, and the processor implements the following steps when executing the computer program: Stress signals of four columns of the injection molding machine are acquired, and the total locking force is calculated, and the stress signals are preprocessed by temperature compensation and sliding average filtering; The real-time mold cavity pressure is calculated based on a linear model triggered by the filling stage of the injection molding machine control signal, a quadratic function model of the pressure maintaining stage and an exponential decay model of the cooling stage; A standard curve template with a confidence interval is generated according to the characteristic parameters of qualified samples and is dynamically updated; The quality judgment result of the current pressure curve and the standard template is compared in real time, and the parameter automatic adjustment is triggered to form a closed loop control.

[0083] In one embodiment, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the following steps: Stress signals of four columns of the injection molding machine are acquired, and the total locking force is calculated, and the stress signals are preprocessed by temperature compensation and sliding average filtering; The real-time mold cavity pressure is calculated based on a linear model triggered by an injection molding machine control signal in a filling stage, a quadratic function model in a holding stage and an exponential decay model in a cooling stage; A standard curve template with a confidence interval is generated according to the qualified sample characteristic parameters and is dynamically updated; The current pressure curve is compared with the standard template in real time to output a quality determination result, and automatic parameter adjustment is triggered to form a closed-loop control.

[0084] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, the processes of the above-mentioned embodiments can be included. In each embodiment provided in the present application, any reference to a memory, storage, database or other medium can include a non-volatile and / or volatile memory. The non-volatile memory can include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM) or a flash memory. The volatile memory can include a random access memory (RAM) or an external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0085] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is exemplified. In actual applications, the above-mentioned functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above.

[0086] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for optimizing an injection molding process, characterized in that, The injection molding process optimization method includes: The stress signals of the four gates of the injection molding machine are obtained and the total clamping force is calculated. The stress signals are preprocessed by temperature compensation and moving average filtering. Real-time mold cavity pressure is calculated based on the linear model of the mold filling stage, the quadratic function model of the pressure holding stage, and the exponential decay model of the cooling stage triggered by the injection molding machine control signal. Generate a standard curve template with confidence intervals based on the characteristic parameters of qualified samples and update it dynamically. The current pressure curve is compared with the standard template output quality judgment result in real time, and the parameters are automatically adjusted to form a closed-loop control.

2. The injection molding process optimization method according to claim 1, characterized in that, The preprocessing of the stress signal through temperature compensation and moving average filtering specifically includes: Ambient temperature is collected via a built-in temperature sensor; The original stress value is compensated using a preset temperature-stress correction formula Δσ=0.02×με / ℃×(T-25), and automatic calibration is triggered when the compensation error exceeds a set threshold, where T is the ambient temperature collected by the temperature sensor.

3. The injection molding process optimization method according to claim 1, characterized in that, The calculation of real-time mold cavity pressure based on the linear model of the mold filling stage, the quadratic function model of the holding stage, and the exponential decay model of the cooling stage, triggered by the injection molding machine control signal, specifically includes: The linear model for the filling stage adopts the linear relationship P1=a1×F_total+b1, where a1 and b1 are calibration coefficients; The quadratic function model for the pressure holding stage adopts the quadratic function P2=a2×F_total²+b2×F_total+c2, where a2, b2 and c2 are calibration coefficients; The exponential decay model for the cooling stage adopts the exponential decay formula P3=a3×e^(b3×t)+c3×F_total, where a3, b3 and c3 are calibration coefficients.

4. The injection molding process optimization method according to claim 1, characterized in that, The process of generating a standard curve template with confidence intervals based on the characteristic parameters of qualified samples and dynamically updating it specifically includes: The template is updated periodically every N template cycles. When the deviation of the new sample data from the current template exceeds 5%, a new template is generated using the weighted average method, and the historical version is retained. Here, N is an integer greater than or equal to 1000. The quality assessment adopts a phased weighted scoring method. When the total matching degree is lower than 0.8 or the score of any phase is lower than 0.7, the part is judged as unqualified.

5. The injection molding process optimization method according to claim 1, characterized in that, The real-time comparison of the current pressure curve with the standard template outputs the quality judgment result, triggering automatic parameter adjustment to form a closed-loop control, specifically including: If the same abnormality is detected twice in a row, automatic parameter adjustment will be triggered. If the abnormality still exists after adjustment, the adjustment range will be increased to 1.5 times the original value. If the abnormality is detected five times in a row, an audible and visual alarm will be triggered and production will be suspended. The data traceability system stores the original stress curve, calculated pressure curve, and adjustment records for each mold. It uses a time-series database and associates it with the gate column number, injection molding machine number, mold number, and operator ID.

6. An injection molding process optimization device, characterized in that, The injection molding process optimization device includes: The data preprocessing module is used to acquire the stress signals of the four gates of the injection molding machine and calculate the total clamping force. The stress signals are preprocessed by temperature compensation and moving average filtering. The pressure calculation module is used to calculate the real-time mold cavity pressure based on the linear model of the mold filling stage, the quadratic function model of the holding pressure stage, and the exponential decay model of the cooling stage triggered by the injection molding machine control signal. The parameter update module is used to generate a standard curve template with confidence intervals based on the characteristic parameters of qualified samples and to update it dynamically. The injection molding control module is used to compare the current pressure curve with the quality judgment result output by the standard template in real time, and trigger the automatic adjustment of parameters to form a closed-loop control.

7. The injection molding process optimization device according to claim 6, characterized in that, The data preprocessing module includes: The temperature acquisition submodule is used to acquire ambient temperature through a built-in temperature sensor. The calibration submodule is used to compensate the original stress value using a preset temperature-stress correction formula Δσ=0.02×με / ℃×(T-25), and to trigger automatic calibration when the compensation error exceeds a set threshold, where T is the ambient temperature collected by the temperature sensor.

8. The injection molding process optimization device according to claim 6, characterized in that, The pressure calculation module includes: The die-filling stage calculation submodule is used for the linear model of the die-filling stage, which adopts the linear relationship P1=a1×F_total+b1, where a1 and b1 are calibration coefficients; The pressure holding stage calculation submodule is used for the pressure holding stage quadratic function model, which adopts the quadratic function P2=a2×F_total²+b2×F_total+c2, where a2, b2 and c2 are calibration coefficients. The cooling stage calculation submodule is used for the exponential decay model of the cooling stage, which adopts the exponential decay formula P3=a3×e^(b3×t)+c3×F_total, where a3, b3 and c3 are calibration coefficients.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the injection molding process optimization method as described in any one of claims 1 to 5.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the injection molding process optimization method as described in any one of claims 1 to 5.

Images