A process parameter optimization control method and system for an automobile injection molding part

Abstract

The application discloses a kind of process parameter optimization control method and system of automobile injection molding piece, it is related to injection molding intelligent control technical field, obtain the three-dimensional structure of automobile injection molding piece, wall thickness, gate and cooling structure, and according to flow path, wall thickness variation and structural feature division multiple molding area;In the injection molding process, the cavity pressure of each molding area, cavity temperature and the injection pressure of injection molding machine, screw position and injection speed are collected, according to the feature value extracted according to injection time axis, form regional actual molding state fingerprint;Actual molding state fingerprint is compared with the target molding state fingerprint established based on qualified product;According to the sensitivity matrix between pre-established process parameter, molding area and quality defect, the process parameter to be adjusted is inverted;Under the constraint of process parameter range, molding pressure and single cycle variation, the minimum disturbance parameter adjustment amount is obtained and executed, to realize the dynamic optimization and closed-loop control of automobile injection molding piece process parameter.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for injection molding, specifically to a method and system for optimizing and controlling process parameters of automotive injection molded parts. Background Technology

[0002] Plastic parts such as car bumpers, dashboards, door panels, pillar trims, car armrests, and various structural supports are usually large in size, thin in wall, curved in surface, and have complex structures such as reinforcing ribs, snap-fit ​​pillars, and multiple gates. Their injection molding quality is affected by multiple process parameters, including melt temperature, mold temperature, injection speed, injection pressure, V / P switching position, holding pressure, holding time, and cooling time. Furthermore, there is a clear coupling relationship between different process parameters.

[0003] For example, existing technologies have employed Gaussian prediction models to predict the warpage of automotive front grilles based on parameters such as injection pressure, melt temperature, holding pressure, holding time, and injection time. Other technologies use neural networks to establish relationships between various injection molding process parameters and part quality. On the other hand, existing technologies are beginning to utilize cavity pressure curves to identify features such as pressure build-up points, pressure abrupt change points, V / P switching inflection points, and holding pressure plateaus to assist in setting injection molding process parameters. However, existing solutions typically use the warpage, shrinkage, weight, or dimensions of the entire part as overall quality evaluation indicators, making it difficult to further distinguish the different molding states between the gate area, thin-walled area, thick-walled area, reinforcing rib area, and far-fill area. Furthermore, process data such as cavity pressure are mostly used for anomaly monitoring or initial parameter setting, and cannot fully address the continuous correlation issues of "which area the anomaly occurs in, which process parameter causes it, and when it should be adjusted."

[0004] Furthermore, the sensitivity of process parameters varies in different regions. For example, while increasing the holding pressure can improve the shrinkage compensation effect in thick-walled areas, it may increase the risk of warping in the reinforcing rib areas. Therefore, simply adjusting parameters based on a single quality indicator can easily lead to new quality defects. Existing automotive armrest injection molding solutions have also disclosed a technical approach based on cavity topology partitioning and implementing partition compensation. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for optimizing and controlling process parameters of automotive injection molded parts, so as to solve the technical problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing and controlling process parameters of automotive injection molded parts, the method comprising the following steps: The three-dimensional structure, wall thickness, gate and cooling structure of automotive injection molded parts are obtained, and the molded parts are divided into multiple molding areas according to the material flow path, wall thickness change and structural characteristics. During the injection molding process, the cavity pressure, cavity temperature, injection pressure, screw position and injection speed of the injection molding machine in each molding area are collected synchronously. The data are processed synchronously according to the injection time axis to extract the pressure peak, pressure integral, pressure change rate, temperature change rate, filling time and cooling rate of each molding area to form the actual molding state fingerprint. The actual molding state fingerprint is compared with the target molding state fingerprint formed by historical data of qualified products and trial mold data to identify abnormal molding areas and their state deviations. Based on the pre-established sensitivity matrix between process parameters, molding area and quality defects, the process parameters to be adjusted corresponding to the state deviation are inverted, and the adjustable parameters and adjustment window are determined in combination with the current filling, speed and pressure switching, pressure holding or cooling stage. Under the constraints of process parameter variation range and molding pressure, solve for the parameter adjustment amount that satisfies the minimum quality deviation and the minimum parameter disturbance, and then perform parameter adjustment; The adjusted next mold forming state and quality inspection results are obtained. The sensitivity matrix and target forming state fingerprint are updated based on the verification results to form a continuous closed-loop process parameter optimization control.

[0007] Preferably, the forming area includes at least the gate influence area, the thin-walled rapid cooling area, the thick-walled shrinkage area, the reinforcing rib or column area, the weld line sensitive area, and the far-end filling area; The boundaries of each region are determined based on the distance between each region and the gate, the wall thickness gradient, the distribution of the reinforcing structure, and the confluence of the material flow. Corresponding cavity pressure acquisition points are configured for each molding region, or state mapping is performed through adjacent pressure acquisition points.

[0008] Preferably, the actual molding state fingerprint is formed in segments according to the injection molding stage, including plasticizing state characteristics, filling state characteristics, speed and pressure switching state characteristics, holding pressure state characteristics, and cooling state characteristics; wherein, the filling state characteristics include filling time, pressure build-up time, pressure peak and pressure change rate, the holding pressure state characteristics include holding pressure plateau pressure, pressure integral and pressure decay rate, and the cooling state characteristics include the regional temperature drop rate and the temperature difference between adjacent molding regions.

[0009] Preferably, the target molding state fingerprint is established based on the process parameters, cavity pressure data, cavity temperature data, and final quality inspection data corresponding to historical qualified products, and the target pressure range, target temperature range, target pressure integral range, and target cooling rate range are determined according to the molding area; when the actual molding state fingerprint exceeds the corresponding target range, the abnormality level is determined according to the absolute value of the state deviation and the duration of the deviation.

[0010] Preferably, the sensitivity matrix between process parameters, molding area and quality defects is established by historical injection molding data and controlled parameter disturbance test data. Its matrix elements represent the degree of influence of a single process parameter change on the warpage, shrinkage, dimensional deviation, incomplete filling or weld line risk of a specified molding area; and corresponding treatment is carried out according to abnormal molding areas.

[0011] Preferably, when the current molding stage is the filling stage, the process parameters to be adjusted include injection speed, injection pressure, and melt temperature; when the cavity pressure in the far filling area is detected to be lower than the corresponding target pressure range and the melt temperature is within the allowable flow temperature range, the injection speed is increased; when the cavity pressure exceeds the set pressure limit, the injection speed is reduced or the speed / pressure switching is performed in advance. When the current molding stage is the holding pressure stage, the holding pressure or holding time is adjusted according to the deviation between the pressure integral and the target pressure integral.

[0012] Preferably, the parameter adjustment amount is obtained through a constrained minimum perturbation optimization model. The deviation between the actual molding state fingerprint and the target molding state fingerprint is used as the first optimization objective, and the process parameter adjustment amount is used as the second optimization objective. The upper limit of injection pressure, the upper limit of injection speed, the mold temperature range, the holding pressure range, and the maximum change of parameters in a single cycle are set as constraints, so that the optimized parameter adjustment amount can reduce the quality deviation while limiting the sudden change of process parameters.

[0013] Preferably, after the parameter adjustment, the cavity pressure, cavity temperature, screw position and injection pressure data of the next molded automotive injection part are re-collected, and the corresponding quality inspection results are obtained through size inspection, weight inspection and appearance inspection; the difference in molding state fingerprint before and after adjustment is correlated with the quality inspection results. When the quality indicators improve and no new abnormal areas are introduced, the adjusted parameters and corresponding state data are added to the qualified sample set; otherwise, the priority of the corresponding parameter adjustment strategy is reduced and the candidate adjustment parameters are recalculated.

[0014] Preferably, the melt plasticizing time, screw torque, back pressure, melt temperature and injection pressure are further collected, and the rheological state characterization value of the current raw material batch is calculated based on the data; when the difference between the rheological state characterization value and the target rheological state exceeds the set threshold, the pressure characteristics and filling time characteristics in the target molding state fingerprint are compensated for material state to avoid abnormal misjudgment caused by changes in raw material batches.

[0015] Preferably, the system includes: The product structure analysis module is used to obtain the three-dimensional structure, wall thickness, gate and cooling structure of automotive injection molded parts and to divide the molding area. A multi-source data acquisition module is used to simultaneously acquire cavity pressure, cavity temperature, injection pressure, screw position, and injection speed. The status fingerprint generation module is used to extract the pressure peak, pressure integral, pressure change rate, temperature change rate, filling time and cooling rate of each molding area according to the injection molding stage and form the actual molding status fingerprint. The state deviation identification module is used to compare the actual forming state fingerprint with the target forming state fingerprint and determine the abnormal forming area. The parameter sensitivity analysis module is used to determine candidate adjustment parameters based on the sensitivity matrix between process parameters, molding area, and quality defects. The constraint optimization module is used to calculate the minimum disturbance parameter adjustment amount by combining the current forming stage and the parameter change range; The execution control module is used to send parameter adjustment amounts to the injection molding equipment controller; The verification update module is used to obtain the adjusted next mold forming state and quality inspection results, and update the sensitivity matrix and target forming state fingerprint according to the verification results.

[0016] Compared with the prior art, the beneficial effects of the present invention are: 1. Based on the three-dimensional structure, wall thickness variation, gate location, material flow path, reinforcing ribs, and melt confluence location, this invention divides the product into multiple molding areas, including the gate influence area, thin-walled rapid cooling area, thick-walled shrinkage area, reinforcing rib and pillar area, weld line sensitive area, and far-end filling area. For each area, features such as cavity pressure, temperature, pressure integral, pressure change rate, filling time, and cooling rate are extracted to form a regionalized molding state fingerprint. This enables independent characterization of the molding state of different structural areas of automotive injection molded parts, allowing the system to locate specific areas of quality anomalies. This solves the problem of existing technologies using overall part quality indicators for process evaluation, which cannot accurately identify molding anomalies in local areas. It is particularly suitable for large thin-walled automotive parts and thick-thin-thin composite structural parts. 2. This invention correlates injection speed, melt temperature, V / P switching position, holding pressure, holding time, and cooling parameters with specific molding areas and quality defects. By establishing a three-dimensional sensitivity relationship through historical production data and controlled parameter disturbance data, it obtains the degree of influence of different process parameters on warpage, shrinkage, dimensional deviation, incomplete filling, and weld line risk in different areas. This solves the problem that traditional injection molding parameter adjustment mainly relies on the experience of engineers and the problem of "not knowing which parameter to adjust after discovering defects". It realizes the transformation from "detecting anomalies" to "analysis of the cause of anomalies", reducing process fluctuations caused by blind parameter adjustment and simultaneous adjustment of multiple parameters. 3. This invention divides a single injection molding process into stages such as plasticizing, filling, V / P switching, holding pressure, and cooling, and determines the adjustable process parameters based on the physical molding mechanism of each stage. For example, in the filling stage, the main controls are injection speed, injection pressure, and melt temperature; in the holding pressure stage, the main controls are holding pressure and holding time; and in the cooling stage, the focus is on adjusting the cooling time and cooling circuit status. This addresses the problem that while traditional methods can identify the relationship between parameters and quality, they do not further solve the problem of "when to adjust parameters," ensuring that the selection of process parameters matches the injection molding stage, improving the timeliness and effectiveness of parameter adjustments, and reducing ineffective adjustments. 4. After determining the parameters to be adjusted, this invention does not directly adopt large-scale parameter changes. Instead, it constructs an optimization objective that simultaneously considers quality deviation and parameter change, and sets constraints such as injection pressure, injection speed, mold temperature, holding pressure, and single-cycle parameter change. It then solves for the minimum parameter adjustment that can reduce molding state deviation. This solves the problem that traditional optimization methods may drastically adjust process parameters to improve a certain quality index, thereby causing new warping, flash, internal stress, or dimensional deviations in other areas. This improves the stability of process control, enabling the system to perform incremental optimization based on the original stable process, and reducing the risk of parameter oscillation. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the optimized control process of the present invention. Detailed Implementation

[0018] Please see Figure 1 One embodiment of the present invention provides: optimization and control of automotive door panel injection molding process parameters based on regional molding state fingerprints. This embodiment focuses on injection-molded automotive door panel interior trim. The automotive door panel interior trim is made of polypropylene material, with a maximum length of 1180mm, a maximum width of 820mm, and an average wall thickness of 2.5mm. The equivalent wall thickness of the local reinforcing rib area reaches 4.5–6.0mm. It features multiple snap-fit ​​posts, screw posts, and reinforcing ribs, and uses a three-point gate for melt injection. Due to the large-sized thin-walled structure of the product, along with the presence of thick-thin-thin wall transitions, reinforcing ribs, and a distal filling area, traditional fixed process parameters can easily lead to problems such as insufficient distal filling, shrinkage marks in the reinforcing rib area, localized warping, and dimensional deviations.

[0019] First, a 3D CAD model of the automotive door panel interior part was obtained, and the outer contour, wall thickness distribution, gate coordinates, reinforcing rib coordinates, snap-fit ​​column coordinates, and cooling water channel location of the part were extracted from the 3D CAD model. A 2D unfolding coordinate system was established for the part based on the gate location and melt flow direction, dividing the part into six molding areas: the first gate influence area, the second thin-walled rapid cooling area, the third thick-walled shrinkage area, the fourth reinforcing rib and snap-fit ​​column area, the fifth weld line sensitive area, and the sixth far-end filling area.

[0020] Among them, the gate influence zone is used to characterize the initial pressure build-up state after the melt enters the cavity; the thin-walled rapid cooling zone is used to judge the risk of freezing of the flow front due to the small wall thickness of the melt; the thick-walled shrinkage zone is used to evaluate the pressure holding and feeding capacity; the reinforcing rib and snap-fit ​​column zone is used to evaluate local pressure concentration and uneven cooling; the weld line sensitive zone is used to evaluate the temperature and pressure state when the flow fronts of two melts merge; and the far-end filling zone is used to evaluate the final filling capacity of the melt.

[0021] Subsequently, cavity pressure sensors and cavity temperature sensors are arranged in the corresponding areas of the mold cavity. Preferably, a first pressure acquisition point, a second pressure acquisition point, and a third pressure acquisition point are respectively set near the three gates, a fourth pressure acquisition point is set in the far-end filling area, a fifth pressure acquisition point is set in the thick-walled shrinkage area, and a temperature acquisition point is set in the weld line sensitive area. The injection molding machine controller synchronously outputs the actual injection pressure, screw position, screw speed, injection speed, holding pressure, mold temperature, and V / P switching position.

[0022] Because different sensors have different sampling frequencies, a unified timestamp is added to each collected data point, and the moment the injection molding machine screw begins to advance is taken as the zero time. The pressure, temperature, screw position, and injection speed data are time-aligned. For sampling points with isolated outliers, linear interpolation of adjacent sampling points is used for correction; for continuous outlier data, the data segment is marked as invalid and is not included in the target state fingerprint calculation.

[0023] After data synchronization is completed, a single injection molding cycle is divided into the plasticizing stage, filling stage, V / P switching stage, holding pressure stage, and cooling stage.

[0024] During the filling stage, the peak pressure, pressure build-up time, pressure change rate, and filling time of each region are extracted in real time; during the holding stage, the pressure plateau value, pressure integral, and pressure decay rate are extracted; during the cooling stage, the cavity temperature drop rate, cooling end temperature, and temperature difference between different regions are extracted.

[0025] Taking the i-th forming region as an example, its forming state fingerprint is represented as follows: Q i =[P max,i ,I P,i ,KP,i ,T max,i ,K T,i ,t fill,i ,R cool,i ]; Among them, P max,i I represents the maximum cavity pressure in the i-th region. P,i K represents the pressure-time integral. P,i T represents the rate of change of pressure. max,i Indicates the highest temperature, K. T,i t represents the rate of temperature change. fill,i R represents the filling time. cool,i Indicates the cooling rate.

[0026] The fingerprints of the molding state in six regions are combined to form the actual molding state fingerprint of the entire product: Q real =[Q1,Q2,Q3,Q4,Q5,Q6]; Before establishing the actual molding state fingerprint, data from 50 continuously produced qualified products are collected, and 40 molds with stable states are selected as the target sample. The mean and standard deviation of each parameter are calculated, with the mean as the target center value, and the standard deviation above and below the mean forming the target allowable interval, thus obtaining the target molding state fingerprint Q. ∗ .

[0027] For example, for the distal filling zone, the target maximum cavity pressure is 42 MPa, with an allowable range of 38–46 MPa; the target pressure integral is... The permissible range is 115~ The target filling time is 1.82s, with an allowable range of 1.70 to 1.94s.

[0028] During the production of mold number 51, the system detected that the actual maximum cavity pressure in the distal filling zone was only 34 MPa, with a pressure integral of... The filling time reached 2.08 seconds, indicating a significant risk of insufficient filling in the distal filling area.

[0029] At this point, the system does not directly increase the injection pressure, but instead further checks the melt temperature in the distal region. The test results show that the temperature in this region is still within the target temperature range, indicating that the current anomaly is mainly due to insufficient pressure build-up, rather than premature melt cooling.

[0030] Further establish a process parameter-molding area-quality defect sensitivity matrix.

[0031] Through historical experiments, small-amplitude controlled disturbances were applied to the injection speed, melt temperature, V / P switching position, holding pressure, and holding time, and the pressure and temperature changes in six molding zones were recorded. The response coefficients of each parameter to the state indicators of each zone were calculated.

[0032] For example, for the distal filling region, the sensitivity of injection speed changes to filling pressure is 0.82, melt temperature is 0.43, V / P switching position is 0.71, and holding pressure is 0.12. Therefore, under the current abnormal filling conditions, injection speed is the primary adjustment parameter, and V / P switching position is the second primary parameter.

[0033] The system further determines the current stage of the injection molding process. If the abnormality is detected while the process is still in the filling stage, dynamic correction of the injection speed is allowed; if the process has already entered the holding pressure stage, the speed parameters for this injection stage are prohibited from being changed directly, and the correction value is used for the next injection cycle.

[0034] For mold 51, since the system detected a distal pressure deviation during the filling stage, the parameter adjustment amount was calculated as follows: ; Where E represents the deviation between the actual fingerprint and the target fingerprint, and M is the sensitivity matrix, with the first term used to reduce quality deviation and the second term used to limit the range of parameter changes. Calculations show that the injection speed for the next mold is increased by 6%, the V / P switching position is adjusted backward by 1.5 mm, and other parameters remain unchanged.

[0035] At the same time, set the upper limit of single-cycle parameter changes, such as the maximum adjustment range of injection speed not exceeding 10% and the maximum adjustment range of V / P switching position not exceeding 3mm, so as to avoid excessive adjustment at one time causing the pressure peak to exceed the allowable range of the equipment or mold.

[0036] After the 52nd mold was produced, fingerprints of the molded state were re-collected. The results showed that the maximum cavity pressure in the distal region increased from 34 MPa to 40.5 MPa, and the pressure integral increased from... Increase to The filling time was reduced from 2.08s to 1.88s, both of which are within the target allowable range.

[0037] Subsequently, a 3D dimensional inspection device was used to scan the 52nd molded part to obtain the dimensional deviation of the far-end area, the dimensional deviation of the reinforcing rib area, and the overall warpage. If the inspection results meet the preset quality requirements, the parameter adjustment is deemed effective.

[0038] The system establishes a correspondence between the abnormal state of the 51st module, the parameter adjustment amount, the state change of the 52nd module, and the quality detection results, and adds them as new training samples to the parameter sensitivity database.

[0039] This completes a full control process: anomaly region location → sensitive parameter identification → current stage judgment → minimum disturbance adjustment → next mode verification → sensitivity update.

[0040] Please see Figure 1 One embodiment of the present invention is: thick-walled region shrinkage control based on pressure holding integral: this embodiment is for controlling local thick-walled structures in the frame of an automobile dashboard.

[0041] The product has an average wall thickness of 2.8 mm, and a reinforcing structure with a thickness of approximately 5.5 mm is provided in the installation and support area. Due to the large volume of the material inside the reinforcing structure, it is prone to volume shrinkage during the cooling and curing process, and the traditional fixing and holding pressure can easily cause shrinkage marks on the surface of the reinforcing structure.

[0042] The system first determines the thick-walled shrinkage area based on the product's CAD model, and then uses the pressure sensor data corresponding to this area as an independent control object.

[0043] In normal, qualified products, the target pressure integral during the pressure holding stage in the thick-walled region is: The allowed range is When the pressure integral of a thick-walled region of a certain molded product is detected to be only [value missing] during continuous production, [the following text is incomplete and likely refers to a separate event:] Simultaneously, when 3D appearance inspection predicts a high risk of shrinkage marks in this area, the system first checks the holding pressure plateau value. If the overall holding pressure plateau is lower than the target value, the holding pressure is used as the first candidate adjustment parameter; if the plateau pressure is normal but the pressure decays too quickly, the holding time is used as the first candidate adjustment parameter. For example, if the current holding pressure is 32 MPa, the target range is 34–38 MPa, and the holding time is within the normal range, the system calculates the first adjustment scheme of increasing the holding pressure by 2 MPa. However, the system also checks the warping risk in the reinforcing rib area. Since the historical sensitivity matrix shows that the holding pressure has a high positive sensitivity to warping in this area, warping constraints are further set. Ultimately, the holding pressure is adjusted from 32 MPa to 34 MPa, instead of directly to 38 MPa. After the next batch of production, the pressure integral in the thick-walled area is increased to... The risk of shrinkage marks is reduced, and the warping in the reinforcing rib area does not exceed the allowable range. The system thus confirms the adjustment is effective. The key to this embodiment is that it does not solely target shrinkage marks for optimization, but rather correlates the integral of the holding pressure, the shrinkage in the thick-walled area, and the warping in the reinforcing rib area, thereby improving shrinkage marks while limiting other quality problems caused by increased holding pressure.

[0044] Please see Figure 1 One embodiment of the present invention is local warpage control based on mold thermal balance: This embodiment is used for large thin-walled injection molded parts for automobile bumpers.

[0045] The product is approximately 1500mm long and 450mm wide, featuring multiple reinforcing ribs and an arc-shaped structure. The mold is equipped with multiple independent cooling circuits. The system sets temperature acquisition points in different areas of the mold and simultaneously collects the inlet and outlet temperatures and flow rates of the cooling water. Based on continuous production data analysis, a target thermal balance for the mold is established. When the mold temperature in the left area was detected to be 68℃ and the mold temperature in the right area was detected to be 59℃ during a certain production cycle, the temperature difference between the areas reached 9℃, exceeding the allowable range. At the same time, the predicted warpage of the product on the left side reached 0.72mm, while the target value was below 0.45mm.

[0046] The system determines the following based on the sensitivity matrix: injection speed has low sensitivity to warping in this region; holding pressure has medium sensitivity; cooling flow rate has high sensitivity; and cooling time has a significant impact on the overall cycle. Therefore, instead of prioritizing modifications to injection speed or holding pressure, the system controls the corresponding cooling circuits.

[0047] The system first calculates the flow difference between the left and right cooling loops, then increases the flow rate of the left loop while keeping the flow rate of the right loop constant, gradually bringing the cooling capacities of the left and right regions closer together. After the next mold production, the left mold temperature drops to 64℃, while the right mold temperature remains at 59℃, reducing the temperature difference to 5℃. Simultaneously, the predicted warpage on the left side decreases to 0.48mm. The system then incorporates the cooling loop adjustment results into the thermal balance model. When the system maintains a stable state for several consecutive cycles, the new temperature state is updated as part of the target state fingerprint.

[0048] This method avoids directly changing the injection speed or holding pressure due to warping, which is common in traditional injection molding optimization, thereby reducing interference between process parameters.

[0049] Please see Figure 1 One embodiment of the present invention is: adaptive compensation of process parameters under batch changes of raw materials: This embodiment is used to solve the problem of molding state drift caused by batch changes of raw materials in automotive injection molded parts.

[0050] In automotive injection molding production, even if the injection molding machine settings remain constant, variations in filling pressure and filling time can occur due to differences in the melt flow characteristics of different batches of materials. Existing injection molding process optimization schemes typically optimize by setting process parameters or based on historical parameters, but changes in the state of raw materials can cause deviations from fixed target parameters.

[0051] In this embodiment, the following data are additionally collected during the injection molding process: screw torque; plasticizing time; back pressure; melt temperature; actual injection pressure; and screw position. Material rheological state characterization values ​​are then established. Among them, Tm V is the melt temperature. i For injection speed, P i For injection pressure, t p τ is the plasticizing time, and τ is the screw torque.

[0052] Reference rheological state obtained during normal production: When the new batch of materials was put into production, the system detected: If so, it can be determined that the rheological state of the material has changed.

[0053] For example, if the actual plasticizing time of a new material batch increases by 0.25 seconds and the screw torque increases by 8%, the pressure in the far-filling region decreases by approximately 5% while the melt temperature remains constant. Based on the historical sensitivity matrix, the system determines that this pressure drop is primarily due to changes in the material flow state, rather than a malfunction in the injection molding machine. In this case, instead of directly maintaining the target pressure value in the far-filling region as the original target value, the system compensates for the target state fingerprint based on the changes in material state. Simultaneously, new process parameter compensation amounts were calculated. Calculations showed that increasing the melt temperature by 5°C and the injection speed by 3%, while keeping the holding pressure and cooling time unchanged, was appropriate.

[0054] The results of the next test showed that the pressure in the far region recovered to the compensated target range, and there was no problem of excessive pressure near the gate.

[0055] After five consecutive model verifications, the system confirmed the effectiveness of the new material condition compensation model and established a mapping relationship between the material condition of this batch and the corresponding process parameters.

[0056] Please see Figure 1 The present invention provides an embodiment of multi-parameter collaborative optimization when multiple defects occur simultaneously: This embodiment is used to illustrate the parameter optimization process when multiple quality risks occur simultaneously in automotive injection molded parts.

[0057] During the production of a certain automotive door panel injection molded part, the system detected the following issues: low pressure in the distal region; insufficient pressure integration in the thick-walled region; excessively rapid temperature drop in the reinforcing rib region; and temperature below the target range in the weld line region. Using traditional manual parameter adjustment methods might simultaneously increase the injection speed, melt temperature, and holding pressure, potentially leading to new problems such as excessively high pressure in the gate region and increased warpage.

[0058] This invention first converts each abnormal state into a deviation vector: The corresponding sensitivity matrix is ​​then represented as: Each row corresponds to a quality anomaly, and each column corresponds to a candidate process parameter.

[0059] By solving for MΔU≈−E, the combined adjustment amount of multiple process parameters is obtained. Simultaneously, a penalty for parameter changes is added: J=∣∣MΔU+E∣∣ 2 +λ∣∣ΔU∣∣ 2 Furthermore, process boundary constraints were added. The final system achieved the following: melt temperature increased by 3°C; injection speed increased by 4%; holding pressure increased by 1.5 MPa; and cooling time increased by 0.3 s. This was achieved instead of directly and significantly increasing a single parameter. After verification in the next mold, the distal pressure, thick-wall pressure integral, and weld line temperature all fell within the target range, while the warpage in the reinforcing rib area remained within the allowable range. This demonstrates synergistic parameter optimization under multiple defect conditions.

[0060] Please see Figure 1 The present invention provides an embodiment of closed-loop learning control in continuous production: This embodiment further illustrates the operation mode of the system in the continuous production process.

[0061] The system establishes an independent data record for each molded part: D k =(U k Q k ,Y k Among them: U k The actual process parameters for the k-th module; Q k The fingerprint of the actual formed state of the k-th mold; Y k This represents the final quality inspection result for the k-th mold. For example, after continuously producing 100 molds, the system obtains 100 sets of corresponding data.

[0062] For qualified products, their formed-state fingerprints are extracted to form a qualified sample set: D good ={D1,D2,...,D n Products exhibiting warping, shrinkage marks, or dimensional deviations are treated as abnormal samples. The system statistically analyzes the variation patterns of process parameters under different anomaly types and updates the sensitivity matrix accordingly.

[0063] For example, the initial sensitivity of "holding pressure to shrinkage marks in thick-walled areas" was 0.61 based on trial molding data. After updating the data from 1000 consecutive production runs, the actual sensitivity was adjusted to 0.68. The system then writes the new sensitivity into the control model, ensuring that subsequent occurrences of the same type of anomaly will preferentially select the holding pressure for minor adjustments. To prevent abnormal products from contaminating the target model, products are only allowed to enter the target sample library if they pass the size, appearance, and weight inspections.

[0064] Target state update adopts: Where ρ is the update coefficient.

[0065] When mold repair, raw material batch change, or cooling system maintenance is detected, the update coefficient is reduced to prevent sudden changes in equipment status from causing the target state to drift rapidly.

[0066] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for optimizing and controlling process parameters of automotive injection molded parts, characterized in that: The method includes the following steps: The three-dimensional structure, wall thickness, gate and cooling structure of automotive injection molded parts are obtained, and the molded parts are divided into multiple molding areas according to the material flow path, wall thickness change and structural characteristics. During the injection molding process, the cavity pressure, cavity temperature, injection pressure, screw position and injection speed of the injection molding machine in each molding area are collected synchronously. The data are processed synchronously according to the injection time axis to extract the pressure peak, pressure integral, pressure change rate, temperature change rate, filling time and cooling rate of each molding area to form the actual molding state fingerprint. The actual molding state fingerprint is compared with the target molding state fingerprint formed by historical data of qualified products and trial mold data to identify abnormal molding areas and their state deviations. Based on the pre-established sensitivity matrix between process parameters, molding area and quality defects, the process parameters to be adjusted corresponding to the state deviation are inverted, and the adjustable parameters and adjustment window are determined in combination with the current filling, speed and pressure switching, pressure holding or cooling stage. Under the constraints of process parameter variation range and molding pressure, solve for the parameter adjustment amount that satisfies the minimum quality deviation and the minimum parameter disturbance, and then perform parameter adjustment; The adjusted next mold forming state and quality inspection results are obtained. The sensitivity matrix and target forming state fingerprint are updated based on the verification results to form a continuous closed-loop process parameter optimization control.

2. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: The forming area includes at least the gate influence area, the thin-walled rapid cooling area, the thick-walled shrinkage area, the reinforcing rib or column area, the weld line sensitive area, and the far-end filling area. The boundaries of each region are determined based on the distance between each region and the gate, the wall thickness gradient, the distribution of the reinforcing structure, and the confluence of the material flow. Corresponding cavity pressure acquisition points are configured for each molding region, or state mapping is performed through adjacent pressure acquisition points.

3. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: The actual molding state fingerprint is formed in segments according to the injection molding stage, including plasticizing state characteristics, filling state characteristics, speed and pressure switching state characteristics, holding pressure state characteristics, and cooling state characteristics; among them, the filling state characteristics include filling time, pressure build-up time, pressure peak and pressure change rate; the holding pressure state characteristics include holding pressure plateau pressure, pressure integral and pressure decay rate; and the cooling state characteristics include the regional temperature drop rate and the temperature difference between adjacent molding regions.

4. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: The target molding state fingerprint is established by the process parameters, cavity pressure data, cavity temperature data and final quality inspection data corresponding to historical qualified products, and the target pressure range, target temperature range, target pressure integral range and target cooling rate range are determined according to the molding area. When the actual formed fingerprint exceeds the corresponding target range, the anomaly level is determined according to the absolute value of the state deviation and the duration of the deviation.

5. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: The sensitivity matrix between process parameters, molding area and quality defects is established by historical injection molding data and controlled parameter disturbance test data. Its matrix elements represent the degree of influence of a single process parameter change on the warpage, shrinkage, dimensional deviation, incomplete filling or weld line risk of a specified molding area; and the abnormal molding area is handled according to its state.

6. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: When the current molding stage is the filling stage, the process parameters to be adjusted include injection speed, injection pressure and melt temperature; when the cavity pressure in the far filling area is detected to be lower than the corresponding target pressure range and the melt temperature is within the allowable flow temperature range, the injection speed is increased; when the cavity pressure exceeds the set pressure limit, the injection speed is reduced or the speed / pressure switch is performed in advance. When the current molding stage is the holding pressure stage, the holding pressure or holding time is adjusted according to the deviation between the pressure integral and the target pressure integral.

7. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: The parameter adjustment amount is obtained through a constrained minimum perturbation optimization model. The deviation between the actual molding state fingerprint and the target molding state fingerprint is used as the first optimization objective, and the process parameter adjustment amount is used as the second optimization objective. The upper limit of injection pressure, the upper limit of injection speed, the mold temperature range, the holding pressure range, and the maximum change of parameters in a single cycle are set as constraints so that the optimized parameter adjustment amount can reduce the quality deviation while limiting the sudden changes of process parameters.

8. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: After parameter adjustments, the cavity pressure, cavity temperature, screw position, and injection pressure data are re-collected for the next molded automotive injection part. Corresponding quality inspection results are obtained through dimensional inspection, weight inspection, and appearance inspection. The difference in molding state fingerprint before and after adjustment is correlated with the quality inspection results. When the quality indicators improve and no new abnormal areas are introduced, the adjusted parameters and corresponding state data are added to the qualified sample set. Otherwise, the priority of the corresponding parameter adjustment strategy is reduced and the candidate adjustment parameters are recalculated.

9. The method for optimizing and controlling process parameters of automotive injection molded parts according to claim 1, characterized in that: Further data such as melt plasticizing time, screw torque, back pressure, melt temperature, and injection pressure are collected, and the rheological state characterization value of the current raw material batch is calculated based on the data. When the difference between the rheological state characterization value and the target rheological state exceeds the set threshold, material state compensation is performed on the pressure characteristics and filling time characteristics in the target molding state fingerprint to avoid abnormal misjudgments caused by changes in raw material batches.

10. A process parameter optimization and control system for automotive injection molded parts, used to execute the process parameter optimization and control method for automotive injection molded parts according to any one of claims 1-9, characterized in that: The system includes: The product structure analysis module is used to obtain the three-dimensional structure, wall thickness, gate and cooling structure of automotive injection molded parts and to divide the molding area. A multi-source data acquisition module is used to simultaneously acquire cavity pressure, cavity temperature, injection pressure, screw position, and injection speed. The status fingerprint generation module is used to extract the pressure peak, pressure integral, pressure change rate, temperature change rate, filling time and cooling rate of each molding area according to the injection molding stage and form the actual molding status fingerprint. The state deviation identification module is used to compare the actual forming state fingerprint with the target forming state fingerprint and determine the abnormal forming area. The parameter sensitivity analysis module is used to determine candidate adjustment parameters based on the sensitivity matrix between process parameters, molding area, and quality defects. The constraint optimization module is used to calculate the minimum disturbance parameter adjustment amount by combining the current forming stage and the parameter change range; The execution control module is used to send parameter adjustment amounts to the injection molding equipment controller; The verification update module is used to obtain the adjusted next mold forming state and quality inspection results, and update the sensitivity matrix and target forming state fingerprint according to the verification results.

Images