Plastic product injection molding processing control method and system

Abstract

The invention belongs to the technical field of plastic product injection molding processing, and provides a plastic product injection molding processing control method and system.The method comprises the steps that a near infrared spectrum detector is introduced to obtain raw material data, historical process parameters are combined, a material-process parameter mapping model is constructed through an algorithm, and a material-process parameter mapping model is constructed; outputting prediction parameters such as glue injection speed, melt temperature, dwell pressure and the like, taking the prediction parameters as initial values, collecting the mold cavity pressure, the screw position and the melt temperature in the injection molding process in real time, establishing a process parameter relation model, predicting the influence of the parameters on product precision, and performing dynamic optimization; according to the method, the machining precision and efficiency are improved, dynamic changes such as raw material batch differences and small equipment disturbance are coped in time, the production stability is guaranteed, the machining quality of plastic products is ensured, the reject ratio is reduced, the method is suitable for multi-scene injection molding production, intelligent upgrading is assisted, and good industrial practicability and economic benefits are achieved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of injection molding of plastic products, and particularly relates to a plastic product injection molding control method and system. BACKGROUND

[0002] In the prior art, the injection molding precision of daily-use plastic products has many problems, which seriously affect product quality and production efficiency, and the specific manifestations are as follows:

[0003] During the injection molding process, there is a strong coupling relationship among the injection speed, holding pressure, and melt temperature, and the traditional control method is difficult to realize precise collaborative control; for example, when producing thin-walled plastic products, a slight fluctuation in the injection speed may lead to underfilling or flash defects.

[0004] The dynamic changes in the process caused by factors such as raw material performance fluctuations and equipment aging, but the existing control system lacks effective real-time monitoring and self-adaptive adjustment capabilities, and cannot respond to these changes in time, thereby affecting the stability of product precision.

[0005] Therefore, how to improve the injection molding precision of plastic products has become a technical problem to be solved in the field.

[0006] Therefore, the application provides a plastic product injection molding control method and system. SUMMARY

[0007] In order to make up for the deficiencies of the prior art and solve at least one technical problem proposed in the background art.

[0008] The technical scheme adopted by the application to solve the technical problems is:

[0009] In a first aspect, the application provides a plastic product injection molding control method, comprising:

[0010] S1: introducing a near-infrared spectrum detector to obtain detection data, combining the detection data with historical process parameters, using a random forest algorithm to construct a material process parameter mapping model, and outputting injection speed, melt temperature, and holding pressure prediction process parameters;

[0011] S2: using the injection speed, melt temperature, and holding pressure prediction process parameters output by the material process parameter mapping model as initial values, simultaneously collecting the mold cavity pressure, screw position, and melt temperature in the injection molding process to establish a process parameter relationship model, predicting the influence of different process parameters on product precision according to the process parameter relationship model, and dynamically optimizing the injection speed, melt temperature, and holding pressure according to the prediction results.

[0012] As a further improved scheme of the application, the specific way of introducing the near-infrared spectrum detector is:

[0013] At the hopper outlet of the injection molding machine, a near-infrared spectrum detector is installed, and the melt flow rate and melt density are analyzed in real time before the raw material enters the cylinder.

[0014] As a further improvement of the present application, the historical process parameters are specifically:

[0015] In the production of daily-use plastic products, the injection speed, melt temperature, holding pressure, melt flow rate and melt density of different batches of PP particle raw materials are recorded.

[0016] As a further improvement of the present application, the specific process of constructing the material process parameter mapping model is:

[0017] The historical process parameters of injection speed, melt temperature, holding pressure, melt flow rate and melt density are derived from the PLC system of the injection molding machine, the production batch corresponding to each process parameter combination is recorded, the process combination corresponding to the qualified daily-use plastic products without defects is manually screened as the optimal process parameter combination, and the process combination corresponding to the daily-use plastic products with defects is marked with defects; all historical process parameters are obtained, all historical process parameters are denoised, a random forest algorithm is selected, and a nonlinear mapping between the melt flow rate, melt density and injection speed, melt temperature and holding pressure is fitted;

[0018] A multi-output random forest regressor is used to encapsulate 100 decision trees, the maximum tree depth is set to 10, the denoised melt flow rate and melt density are input, and the injection speed, melt temperature and holding pressure are output;

[0019] The accuracy of the injection speed, melt temperature and holding pressure is verified using the training set to meet the requirements of industrial production, and the material process parameter mapping model is constructed.

[0020] As a further improvement of the present application, the specific process of real-time acquisition of the cavity pressure, screw position and melt temperature during the injection molding process is:

[0021] The process of real-time acquisition of the cavity pressure is that two piezoelectric pressure sensors are arranged at the gate and the end of the cavity, the piezoelectric pressure sensor is installed flush with the inner wall of the cavity to avoid interference of the melt flow, and the cavity pressure is acquired in real time;

[0022] The process of real-time acquisition of the screw position is that the grating displacement sensor is directly mounted at the end of the injection screw, and the shield cover is wrapped, and the screw position is acquired in real time;

[0023] The process of real-time acquisition of the melt temperature is that the infrared temperature measurement sensor is embedded in the nozzle, which is 5-10 mm away from the gate, and the melt temperature is acquired in real time.

[0024] As a further improvement of the present application, the process parameter relationship model specifically comprises:

[0025] The initial parameters output by the material process parameter mapping model are accessed, including the injection speed, melt temperature and holding pressure, and the injection speed, melt temperature and holding pressure are processed, abnormal values are removed, the injection speed, melt temperature and holding pressure after removing abnormal values are respectively subjected to conditional constraints, the safety ranges of the injection speed, melt temperature and holding pressure are respectively obtained, and the safety ranges are used for constraints.

[0026] As a further improvement of the present application, the safety range is:

[0027] The safety range refers to the corresponding value range of different process parameters in the injection molding process of daily plastic products. If a process exceeds its corresponding value range, the qualified rate of the injection molding process of daily plastic products will be affected. The process parameters within the process safety range will not affect the qualified rate of the injection molding process of daily plastic products.

[0028] As a further improvement of the present application, the process parameter relationship model specifically further comprises:

[0029] The timestamps of the injection speed, melt temperature and holding pressure are synchronized through the injection start signal of the injection molding machine PLC, so as to ensure the accuracy of coupling analysis, and a second-order low-pass filter is used to filter out the interference of hydraulic pumps and servo motors;

[0030] The relationship between the injection speed and the mold cavity pressure is established based on the Hagen-Poiseuille law: Wherein, η(T(t)) is the melt viscosity at t time, which has an exponential relationship with temperature, Q(v(t)) is the volume flow rate at t time, which is proportional to the injection speed, L is the melt flow path length, R is the equivalent radius of the mold cavity runner, and π is the circular constant;

[0031] The training data includes X batch production data, covering the injection speed, melt temperature and holding pressure, the accuracy rate is optimized by using a loss function, and the process parameter relationship model is obtained by iterative training through the Adam algorithm.

[0032] The LSTM neural network is used to learn the time sequence dependence, the input is the injection speed, melt temperature, holding pressure, mold cavity pressure and screw position in the previous N seconds, and the output is the process parameter prediction value in the future N / 2 seconds, including the injection speed, melt temperature and holding pressure.

[0033] As a further improvement of the present application, the specific process of dynamically optimizing the injection speed, melt temperature and holding pressure is:

[0034] The future N / 2 seconds of the injection speed, the melt temperature, and the holding pressure output based on the process parameter relationship model are combined with product defects and size deviations of the daily plastic product to construct a multi-objective optimization model, specifically: min J = alpha * |Delta L| + beta * P + gamma * |Delta u| 缺陷 + gamma * |Delta u|, wherein Delta L is the size deviation of the daily plastic product, P 缺陷 is the flash probability of the daily plastic product, Delta u is a process parameter variation, including: an injection speed adjustment amount Delta v, a holding pressure adjustment amount Delta P, and a melt temperature adjustment amount Delta T, alpha, beta, and gamma are weights, and alpha + beta + gamma = 1.

[0035] According to the process safety range, the optimal solution of the multi-objective optimization model is calculated in real time, and the injection speed, the melt temperature, and the holding pressure are dynamically optimized according to the process parameter variation corresponding to the optimal solution.

[0036] In a second aspect, the present application provides a plastic product injection molding processing control system, comprising:

[0037] The raw material sensing and adaptive module introduces a near-infrared spectrum detector, acquires detection data, combines the detection data with historical process parameters, uses a random forest algorithm to construct a material process parameter mapping model, and outputs injection speed, melt temperature, and holding pressure prediction process parameters.

[0038] The process dynamic control module uses the injection speed, melt temperature, and holding pressure prediction process parameters output by the material process parameter mapping model as initial values, simultaneously collects cavity pressure, screw position, and melt temperature in the injection molding process to establish a process parameter relationship model, predicts the influence of different process parameters on product accuracy according to the process parameter relationship model, and dynamically optimizes the injection speed, melt temperature, and holding pressure according to the prediction results.

[0039] The beneficial effects of the present application are as follows:

[0040] 1. Through the cooperation of the raw material sensing and adaptive module and the process dynamic control module, accurate control of the injection molding process parameters is realized, the multi-parameter strong coupling problem is effectively solved, the size tolerance of the daily plastic product is stably controlled in a high-precision range, the defect occurrence rate of flash, short injection, and shrinkage is significantly reduced, the performance fluctuation of the raw material is detected in real time with the help of the near-infrared spectrum, the adaptive initial parameters are quickly output combined with the material process parameter mapping model, and the production stability is ensured by collecting data such as cavity pressure in real time and dynamically optimizing during the process to timely respond to dynamic changes such as raw material batch differences and equipment minor disturbances.

[0041] 2. Reduce the number of test molds and manual parameter adjustment time, shorten the production cycle through the prediction and optimization of the process parameter relationship model, reduce the defective product rate, improve the overall production efficiency and product qualification rate, integrate multi-source data and algorithm model, form a closed-loop system from raw material perception to process control, reduce the dependence on manual experience, provide support for intelligent and automated production of injection molding processing, and adapt to the production needs of daily-use plastic products in multiple scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0042] The application will be further described below with reference to the accompanying drawings.

[0043] Figure 1 is a step flow chart of a plastic product injection molding processing control method of the application;

[0044] Figure 2 is a system module diagram of a plastic product injection molding processing control system of the application. DETAILED DESCRIPTION

[0045] In order to make the technical means, creative features, purposes and effects realized by the application easy to understand, the application will be further described below with reference to the specific embodiments.

[0046] Example 1

[0047] As shown in Figure 1 , the plastic product injection molding processing control method according to the embodiments of the application comprises:

[0048] S1: Introduce a near-infrared spectrum detector to obtain detection data, combine the detection data with historical process parameters, use a random forest algorithm to build a material process parameter mapping model, and output the predicted process parameters of the injection speed, melt temperature and holding pressure;

[0049] Since the fluctuation of raw material performance is an important factor affecting the processing precision of daily-use plastic products, there are differences in key performance parameters such as melt flow rate and density of different batches of raw materials, which will lead to improper parameter matching in the processing process, and thus affect the product precision.

[0050] Therefore, first, a raw material dynamic adaptation system is established.

[0051] In S1, the first specific way of introducing a near-infrared spectrum detector is:

[0052] At the outlet of the hopper of the injection molding machine, a near-infrared spectrum detector is installed, and before the raw material enters the barrel, the melt flow rate (MFR) and melt density key performance parameters are analyzed in real time.

[0053] In S1, the second specific, the detection data includes: melt flow rate (MFR), melt density;

[0054] In some embodiments, during the production of daily plastic products, the melt flow rate (MFR) and density fluctuations of different batches of PP particles cause process instability; at the outlet of the injection molding machine hopper, a near-infrared spectrum detector is installed, the detection frequency is set to 1 time / second, the material flow is scanned in real time, the characteristic absorption peak of PP material in the near-infrared wave band (1200-2500nm) is used to fit the quantitative model of MFR and density, and the detection data is obtained.

[0055] In S1, the third specific process of combining the detection data with the historical process parameters to construct the material process parameter mapping model is as follows:

[0056] The historical process parameters are specifically the injection speed, melt temperature, holding pressure, melt flow rate (MFR), and melt density of different batches of PP particle raw materials during the production of daily plastic products.

[0057] The historical process parameters of injection speed, melt temperature, holding pressure, melt flow rate (MFR), and melt density are derived from the injection molding machine PLC system, the production batch corresponding to each process parameter combination is recorded, the process combination corresponding to the qualified daily plastic product without defects is manually selected as the optimal process parameter combination, and the process combination corresponding to the defective daily plastic product is labeled with defects; all historical process parameters are obtained, all historical process parameters are denoised, the random forest algorithm is selected, and the nonlinear mapping between melt flow rate (MFR) and melt density and injection speed, melt temperature, and holding pressure is fitted.

[0058] A multi-output random forest regressor is used to encapsulate 100 decision trees with a maximum tree size of 10 to balance accuracy and computing speed, and the denoised melt flow rate (MFR) and melt density are input to output the injection speed, melt temperature, and holding pressure.

[0059] The training set is used to verify the accuracy of the injection speed, melt temperature, and holding pressure to meet the requirements of industrial production, and the construction of the material process parameter mapping model is completed.

[0060] The real scene raw material, process, and quality data are collected to ensure that the model fits the actual situation, the nonlinear relationship is fitted by random forest, the executable process parameters are directly output, the incremental training and new data are used to continuously evolve the model with the changes of raw materials and equipment, and the industry pain points of raw material fluctuations and parameter mismatch leading to accuracy decline are completely solved.

[0061] This method is not only suitable for daily use of PP materials, but also can be extended to various daily plastic products (such as PE, PS, ABS), and has universality and replicability, providing core support for the precision of injection molding processing.

[0062] S2: The material process parameter mapping model outputs the shot speed, melt temperature, and holding pressure predicted process parameters as initial values, while real-time acquisition of the mold cavity pressure, screw position, and melt temperature during the injection molding process is performed to establish a process parameter relationship model. The influence of different process parameters on product accuracy is predicted according to the process parameter relationship model, and the shot speed, melt temperature, and holding pressure are dynamically optimized according to the prediction results.

[0063] After the material process parameter mapping model is established, a process parameter relationship model is introduced to solve the problem of strong multi-parameter coupling and slow dynamic response in the existing daily plastic product process parameter control method.

[0064] The process parameter relationship model takes the shot speed, melt temperature, and holding pressure provided by the material process parameter mapping model as initial values, while real-time acquisition of the mold cavity pressure, screw position, and melt temperature real-time data during the injection molding process of daily plastic products is performed.

[0065] Through algorithm analysis of these multivariate data, a process parameter relationship model is established to accurately predict the influence of different process parameter changes on product accuracy.

[0066] According to the prediction results, the shot speed, holding pressure, melt temperature, and other key processing parameters are dynamically optimized to shorten the response time of process parameter control and ensure that when the raw material fluctuates or the equipment is slightly disturbed, the parameters can be quickly adjusted to maintain processing stability.

[0067] In S2, the first specific process of real-time acquisition of the mold cavity pressure, screw position, and melt temperature is as follows:

[0068] The process of real-time acquisition of the mold cavity pressure is as follows: Two piezoelectric pressure sensors are arranged at the gate and the end of the cavity, and the piezoelectric pressure sensor is installed flush with the inner wall of the cavity to avoid interference from melt flow such as turbulence and stagnation, and real-time acquisition of the mold cavity pressure is performed.

[0069] The process of real-time acquisition of the screw position is as follows: The grating displacement sensor is directly mounted at the end of the injection screw, and a shield is wrapped around it to avoid the influence of injection molding machine vibration and electromagnetic interference on measurement accuracy, and real-time acquisition of the screw position is performed.

[0070] The process of real-time acquisition of the melt temperature is as follows: An infrared temperature measurement sensor is embedded in the nozzle, 5-10 mm away from the gate, to avoid affecting the normal ejection of the melt, and real-time acquisition of the melt temperature is performed.

[0071] In S2, the second specific process of establishing the process parameter relationship model is as follows:

[0072] The initial parameters output by the access material process parameter mapping model include: injection speed, melt temperature, and holding pressure, and the injection speed, melt temperature, and holding pressure are processed, abnormal values are removed, the injection speed, melt temperature, and holding pressure after removing abnormal values are respectively subjected to conditional constraints, the safe ranges of the injection speed, melt temperature, and holding pressure are respectively obtained, and the safe ranges are used for constraints;

[0073] It should be noted that the safe range refers to the corresponding value range of different process parameters in the injection molding process of daily plastic products. If a process exceeds its corresponding value range, it will affect the qualified rate of the injection molding process of daily plastic products. The process parameters within the process safety range will not affect the qualified rate of the injection molding process of daily plastic products.

[0074] For example, the safe range of the injection speed is [100, 120] mm / s, the safe range of the melt temperature is [210, 220] degrees Celsius, and the safe range of the holding pressure is [30, 40] bar. The injection speed, melt temperature, and holding pressure that deviate from the safe range will affect the qualified rate of the daily plastic products.

[0075] The timestamps of the injection speed, melt temperature, and holding pressure are synchronized through the injection start signal of the injection molding machine PLC to ensure accurate coupling analysis. A second-order low-pass filter is used to filter out the interference of hydraulic pumps and servo motors.

[0076] The relationship between the injection speed and the mold cavity pressure is established based on the Hagen-Poiseuille law: Where η(T(t)) is the melt viscosity at time t, which has an exponential relationship with temperature, Q(v(t)) is the volume flow rate at time t, which is proportional to the injection speed, L is the melt flow path length, R is the equivalent radius of the mold cavity runner, and π is the circular constant.

[0077] The training data includes X batch production data covering injection speed, melt temperature, and holding pressure. The accuracy is optimized using a loss function, and the process parameter relationship model is obtained through iterative training using the Adam algorithm.

[0078] LSTM neural network is used to learn time series dependence. The input includes the injection speed, melt temperature, holding pressure, mold cavity pressure, and screw position in the previous N seconds, and the output includes the process parameter prediction values in the future N / 2 seconds, including injection speed, melt temperature, and holding pressure.

[0079] For example, real-time data such as current mold cavity pressure 65 bar, screw position 30 mm, melt temperature 213℃, and injection speed 105 mm / s, holding pressure 33 bar are input into the process parameter relationship model to simulate the influence on the precision of daily plastic products, and output the process parameter prediction values.

[0080] In S2, the specific process of dynamically optimizing the injection speed, melt temperature and holding pressure according to the prediction result is as follows:

[0081] Based on the injection speed, melt temperature and holding pressure of the future N / 2 seconds output by the process parameter relationship model, a multi-objective optimization model is constructed in combination with the product defects and size deviation of the daily plastic product, specifically: min J = α·|ΔL| + β·P + γ·|Δu| 缺陷 + γ·|Δu|, wherein ΔL is the size deviation of the daily plastic product, P 缺陷 is the flash probability of the daily plastic product, Δu is the process parameter change amount, including: injection speed adjustment amount Δv, holding pressure adjustment amount ΔP, melt temperature adjustment amount ΔT, α, β, γ are weights, and α+β+γ=1;

[0082] According to the process safety range, the optimal solution of the multi-objective optimization model is calculated in real time, and the injection speed, melt temperature and holding pressure are dynamically optimized according to the process parameter change amount corresponding to the optimal solution.

[0083] For example, according to the process safety range and the predicted injection speed, melt temperature and holding pressure, the optimal solution of the multi-objective optimization model is obtained, the process parameter change amount corresponding to the optimal solution of the multi-objective optimization model is obtained, including: injection speed adjustment amount, holding pressure adjustment amount, melt temperature adjustment amount, and dynamic optimization is performed by using the PLC control system.

[0084] The step of the present application solves the industry problems of multi-parameter coupling out of control and dynamic interference response lag in the injection molding of daily plastic products, upgrades the injection molding control from experience-driven to data-driven, and provides a replicable intelligent technical path for the production of thin-walled and precise daily plastic products, supporting the industry transformation to high quality, high efficiency and low loss.

[0085] Embodiment 2

[0086] As Figure 2 shown, based on embodiment 1, the present application provides a plastic product injection molding processing control system, which comprises:

[0087] Raw material sensing and adapting module: introduce near-infrared spectrum detector, obtain detection data, combine detection data with historical process parameters, use random forest algorithm to construct material process parameter mapping model, and output injection speed, melt temperature and holding pressure prediction process parameters.

[0088] The specific way of introducing the near-infrared spectrum detector is as follows:

[0089] At the outlet of the injection molding machine hopper, install a near-infrared spectrum detector, and analyze the melt flow rate and melt density in real time before the raw material enters the barrel.

[0090] The historical process parameters are specifically:

[0091] In the production of daily-use plastic products, the injection speed, melt temperature, holding pressure, melt flow rate and melt density of different batches of PP particle raw materials are obtained.

[0092] The specific process of the build-up material process parameter mapping model is:

[0093] The historical process parameters of injection speed, melt temperature, holding pressure, melt flow rate and melt density are derived from the PLC system of the injection molding machine, the production batch corresponding to each process parameter combination is recorded, the process combination corresponding to the qualified daily-use plastic product without defects is manually screened as the optimal process parameter combination, and the process combination corresponding to the daily-use plastic product with defects is marked with defects; all historical process parameters are obtained, all historical process parameters are denoised, a random forest algorithm is selected, and a nonlinear mapping between melt flow rate, melt density and injection speed, melt temperature and holding pressure is fitted;

[0094] A multi-output random forest regressor is used to encapsulate 100 decision trees, the maximum tree depth is set to 10, the denoised melt flow rate and melt density are input, and the injection speed, melt temperature and holding pressure are output;

[0095] The accuracy of the injection speed, melt temperature and holding pressure is verified using the training set to meet the requirements of industrial production, and the material process parameter mapping model is constructed.

[0096] Process dynamic control module: the injection speed, melt temperature and holding pressure predicted by the material process parameter mapping model are used as initial values, and the mold cavity pressure, screw position and melt temperature in the injection molding process are collected in real time to establish a process parameter relationship model, the influence of different process parameters on product accuracy is predicted according to the process parameter relationship model, and the injection speed, melt temperature and holding pressure are dynamically optimized according to the prediction results.

[0097] The specific process of collecting the mold cavity pressure, screw position and melt temperature in real time during the injection molding process is:

[0098] The process of real-time acquisition of mold cavity pressure is as follows: two piezoelectric pressure sensors are arranged at the gate and the end of the cavity, mold inserts are integrated, the piezoelectric pressure sensor is installed flush with the inner wall of the cavity to avoid interference of melt flow, and the mold cavity pressure is collected in real time;

[0099] The process of real-time acquisition of screw position is as follows: the grating displacement sensor is directly mounted at the end of the injection screw, and the shield cover is wrapped, and the screw position is collected in real time;

[0100] The process of real-time acquisition of the melt temperature is that an infrared temperature sensor is embedded in the inside of the nozzle, 5-10 mm away from the gate, and the melt temperature is acquired in real time.

[0101] The process parameter relationship model specifically comprises:

[0102] The initial parameters output by the access material process parameter mapping model include: injection speed, melt temperature and holding pressure, and the injection speed, melt temperature and holding pressure are processed, abnormal values are eliminated, the injection speed, melt temperature and holding pressure after elimination of abnormal values are respectively subjected to conditional constraints, the safety ranges of the injection speed, melt temperature and holding pressure are respectively obtained, and the safety ranges are used for constraints.

[0103] The safety range is:

[0104] The safety range refers to the corresponding value range of different process parameters in the injection molding process of daily plastic products. If a process exceeds its corresponding value range, the qualified rate of the injection molding process of daily plastic products will be affected. The process parameters within the process safety range will not affect the qualified rate of the injection molding process of daily plastic products.

[0105] The process parameter relationship model specifically further comprises:

[0106] The timestamps of the injection speed, melt temperature and holding pressure are synchronized through the injection start signal of the injection molding machine PLC to ensure the accuracy of the coupling analysis, and a second-order low-pass filter is used to filter out the interference of the hydraulic pump and the servo motor;

[0107] The relationship between the injection speed and the mold cavity pressure is established based on the Hagen-Poiseuille law: Wherein, η(T(t)) is the melt viscosity at t time, which has an exponential relationship with temperature, Q(v(t)) is the volume flow rate at t time, which is proportional to the injection speed, L is the melt flow path length, R is the equivalent radius of the mold cavity runner, and π is the circular constant;

[0108] The training data includes: X batch production data, covering injection speed, melt temperature, holding pressure, using loss function to optimize accuracy, and using Adam algorithm for iterative training to obtain the process parameter relationship model;

[0109] LSTM neural network is used to learn time series dependence, the input is the injection speed, melt temperature, holding pressure, mold cavity pressure and screw position in the previous N seconds, and the output is the process parameter prediction value in the future N / 2 seconds, including: injection speed, melt temperature and holding pressure.

[0110] The specific process of dynamically optimizing the injection speed, melt temperature and holding pressure is:

[0111] The future N / 2 seconds of the injection speed, the melt temperature and the holding pressure output based on the process parameter relationship model are combined with the product defect and the size deviation of the daily plastic product to construct a multi-objective optimization model, specifically: min J = alpha * |Delta L| + beta * P + gamma * |Delta u| 缺陷 + gamma * |Delta u|, wherein, Delta L is the size deviation of the daily plastic product, P 缺陷 is the flash probability of the daily plastic product, Delta u is the process parameter variation, including: the injection speed adjustment amount Delta v, the holding pressure adjustment amount Delta P and the melt temperature adjustment amount Delta T, alpha, beta and gamma are weights, and alpha + beta + gamma = 1;

[0112] According to the process safety range, the optimal solution of the multi-objective optimization model is calculated in real time, and the injection speed, the melt temperature and the holding pressure are dynamically optimized according to the process parameter variation corresponding to the optimal solution.

[0113] The basic principles, main features and advantages of the present application are shown and described. It should be understood by those skilled in the art that the present application is not limited by the above examples, and the above examples and descriptions in the specification are only to illustrate the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for controlling the injection molding process of plastic products, characterized in that: include: S1: Introduce a near-infrared spectrometer to acquire detection data, combine the detection data with historical process parameters, use the random forest algorithm to construct a material process parameter mapping model, and output injection speed, melt temperature, and holding pressure to predict process parameters. S2: The injection speed, melt temperature, and holding pressure predicted by the material process parameter mapping model are used as initial values. At the same time, the mold cavity pressure, screw position, and melt temperature during the injection process are collected in real time to establish a process parameter relationship model. Based on the process parameter relationship model, the impact of different process parameters on product accuracy is predicted. Based on the prediction results, the injection speed, melt temperature, and holding pressure are dynamically optimized.

2. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The specific method for introducing the near-infrared spectrometer is as follows: A near-infrared spectrometer is installed at the outlet of the injection molding machine hopper to analyze the melt flow rate and melt density in real time before the raw materials enter the barrel.

3. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The specific historical process parameters are as follows: In the production of daily-use plastic products, the injection speed, melt temperature, holding pressure, melt flow rate, and melt density are determined for different batches of PP particle raw materials.

4. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The specific process for constructing the material process parameter mapping model is as follows: Historical process parameters such as injection speed, melt temperature, holding pressure, melt flow rate, and melt density are exported from the injection molding machine PLC system. The production batches corresponding to each combination of process parameters are recorded. The process combinations corresponding to qualified, defect-free daily-use plastic products are manually selected as the optimal process parameter combinations. Defective process combinations corresponding to defective daily-use plastic products are marked with defects. All historical process parameters are obtained and denoised. A random forest algorithm is selected to fit the nonlinear mapping between melt flow rate, melt density, injection speed, melt temperature, and holding pressure. A multi-output random forest regressor is used, which encapsulates 100 decision trees with a maximum tree depth of 10. The inputs are the denoised melt flow rate and melt density, and the outputs are the injection speed, melt temperature, and holding pressure. Using the training set, we verified the accuracy of injection speed, melt temperature, and holding pressure to meet industrial production requirements, thus completing the construction of a material process parameter mapping model.

5. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The specific process of real-time acquisition of mold cavity pressure, screw position, and melt temperature during the injection molding process is as follows: The process of real-time acquisition of cavity pressure involves placing two piezoelectric pressure sensors at the gate and at the end of the cavity, integrating them with mold inserts. The piezoelectric pressure sensors are installed flush with the inner wall of the cavity to avoid interference from melt flow and to acquire cavity pressure in real time. The real-time acquisition process of the screw position is as follows: the screw position is directly mounted on the end of the injection screw, using a grating ruler displacement sensor, and wrapped with a shielding cover to acquire the screw position in real time. The process of real-time acquisition of melt temperature involves using an infrared temperature sensor embedded inside the nozzle, 5-10 mm from the gate, to acquire the melt temperature in real time.

6. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The process parameter relationship model specifically includes: The initial parameters output by the material process parameter mapping model are input, including injection speed, melt temperature, and holding pressure. The injection speed, melt temperature, and holding pressure are processed to remove outliers. Conditional constraints are applied to the injection speed, melt temperature, and holding pressure after removing outliers to obtain the safe ranges for injection speed, melt temperature, and holding pressure, and these safe ranges are used for constraint.

7. The method for controlling the injection molding process of plastic products according to claim 6, characterized in that: The safety range is as follows: The safe range refers to the range of values ​​for different process parameters during the injection molding process of daily-use plastic products. If a process parameter exceeds its corresponding value range, it will affect the pass rate of the injection molding process of daily-use plastic products. Process parameters within the safe range will not affect the pass rate of the injection molding process of daily-use plastic products.

8. The method for controlling the injection molding process of plastic products according to claim 6, characterized in that: The process parameter relationship model specifically includes: The injection start signal from the injection molding machine PLC is used to synchronize the timestamps of injection speed, melt temperature, and holding pressure to ensure accurate coupling analysis. A second-order low-pass filter is used to filter out interference from the hydraulic pump and servo motor. The relationship between injection speed and mold cavity pressure was established based on the Hagen-Poiseuille law: Where η(T(t)) is the melt viscosity at time t, Q(v(t)) is the volumetric flow rate at time t, L is the melt flow path length, R is the equivalent radius of the mold cavity flow channel, and π is pi. The training data includes X batches of production data, covering injection speed, melt temperature, and holding pressure. The accuracy is optimized using a loss function, and the process parameter relationship model is obtained through iterative training using the Adam algorithm.

9. The method for controlling the injection molding process of plastic products according to claim 1, characterized in that: The specific process for dynamically optimizing the injection speed, melt temperature, and holding pressure is as follows: Based on the injection speed, melt temperature, and holding pressure output from the process parameter relationship model for the next N / 2 seconds, a multi-objective optimization model is constructed by combining product defects and dimensional deviations of daily-use plastic products. Specifically: min J=α·|ΔL|+β·P 缺陷 +γ·|Δu|, where ΔL is the dimensional deviation of daily-use plastic products, P 缺陷 Let denot be the flash probability of daily plastic products, Δu be the change in process parameters, and α, β, and γ be the weights, with α+β+γ=1; Based on the process safety range, the optimal solution of the multi-objective optimization model is calculated in real time. Based on the changes in process parameters corresponding to the optimal solution, the injection speed, melt temperature, and holding pressure are dynamically optimized.

10. A control system for injection molding of plastic products, characterized in that, include: Raw material sensing and adaptation module: Introducing a near-infrared spectroscopy detector to acquire detection data, combining the detection data with historical process parameters, using a random forest algorithm to construct a material process parameter mapping model, and outputting injection speed, melt temperature, and holding pressure to predict process parameters; The process dynamic control module uses the injection speed, melt temperature, and holding pressure predicted by the material process parameter mapping model as initial values. At the same time, it collects the mold cavity pressure, screw position, and melt temperature in real time during the injection molding process to establish a process parameter relationship model. Based on the process parameter relationship model, it predicts the impact of different process parameters on product accuracy and dynamically optimizes the injection speed, melt temperature, and holding pressure based on the prediction results.

Images