A method for optimizing and controlling process parameters in plastic injection molding

Abstract

This invention relates to the field of plastic injection molding process control technology, and more specifically, to a method for optimizing and controlling plastic injection molding process parameters. The method involves collecting process parameters, screw position, and cavity state data from consecutive injection molding cycles to establish a stable molding baseline and stable parameter snapshots. For each parameter action, a corresponding action carrier and parameter response fingerprint are established. The first deviation position of the current cycle is identified, and a responsibility candidate set is formed based on the fingerprint matching results of the action attainability conditions and necessary witness responses. After determining the responsible parameters, adjustments are made within the verified action range and safe process boundaries, and a correction action carrier of the same type is established. The adjustment results are confirmed through correction response fingerprints. When multiple parameters overlap, interactive residues are identified, and cross-cycle plasticizing parameters are combined with the batch and residual state of the plasticized melt to determine responsibility attribution. This method is applicable to closed-loop adjustment of process parameters in the injection molding process.

Description

Technical Field

[0001] This invention relates to the field of plastic injection molding process control technology, and more specifically, to a method for optimizing and controlling plastic injection molding process parameters. Background Technology

[0002] During injection molding, process parameters such as injection speed, injection pressure, V / P switching position, holding pressure, barrel temperature, back pressure, plasticizing screw speed, and mold temperature collectively affect melt filling, pressure transmission, holding pressure compensation, and cooling and solidification. For parts with significant wall thickness variations, long flow paths, or multiple cavities and branched flow channels, local process conditions are highly sensitive to parameter changes. Improper process parameter settings can easily lead to quality problems such as short shots, flash, shrinkage, warpage, and dimensional fluctuations. Therefore, using injection molding machine operating data, cavity pressure data, and product quality data to identify changes in molding conditions during continuous production and adjusting process parameters accordingly has become a common practice in injection molding process control.

[0003] Adjustments to existing process parameters are typically determined based on the final part quality, current cavity pressure deviation, or the correlation between process parameters and quality indicators in historical production data. For example, when the cavity end pressure is too low or the part is underfilled, one or more parameters can be selected for compensation, such as injection speed, injection pressure, V / P switching position, or holding pressure parameters. When product dimensions or shrinkage changes, mold temperature, holding pressure, or plasticizing parameters can also be adjusted in conjunction with historical process windows. This method can utilize existing process experience to narrow down the adjustment range, but the statistical correlation between process parameters and abnormal states does not necessarily mean that the current parameter action has actually been applied to the location where the abnormality occurred.

[0004] The effects and timing of different process parameters during the injection molding cycle vary significantly. The influence of segmented injection speed and pressure is related to the corresponding screw stroke and filling section; the effects of V / P switching and holding pressure parameters unfold as the cavity pressure builds up and is transmitted; changes in barrel temperature, back pressure, and plasticizing screw speed typically occur as the corresponding plasticized melt enters subsequent mold passes; and after mold temperature adjustment, there is a gradual process of thermal establishment. Therefore, when a pressure deviation first occurs at a certain cavity location, the production system may simultaneously retain multiple previous parameter actions, but these actions may not all have an effective effect on the deviation location. If the responsible parameter is selected solely based on the parameter adjustment time, historical correlation, or current deviation magnitude, parameters that have not yet affected the current abnormal area may be included in the adjustment scope.

[0005] On the other hand, the duration of action of different parameters during injection molding may overlap. When two parameters can both affect the same cavity pressure variable, the detected current state may simultaneously include the effects of both parameters; in some cases, the state response resulting from the combined change of parameters may even exceed the existing response range when each parameter changes individually. Existing methods for assigning responsibility for anomalies based on correlation, weight, or empirical rules for single parameters are insufficient to distinguish between the effects of a single parameter and the combined effects of parameters. Furthermore, executing multiple adjustment actions consecutively may cause subsequent detection results to lose a clear parameter correspondence.

[0006] The above problems are even more pronounced for plasticizing parameters that affect multiple molding cycles. Melt from the previous plasticizing cycle may remain in the screw tip, nozzle, or hot runner, and enter the next injection cycle along with melt from subsequent plasticizing cycles, resulting in no completely separate boundary between adjacent plasticizing batches. If the current molding anomaly is directly attributed to the cycle in which the plasticizing action occurs, the influence of the residue from the previous batch on the current cycle's state is easily overlooked.

[0007] Therefore, we propose a method for optimizing and controlling the process parameters of plastic injection molding to solve the above problems. Summary of the Invention

[0008] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method for optimizing and controlling process parameters of plastic injection molding to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing and controlling process parameters in plastic injection molding, comprising: Data on process parameters, screw position, and cavity status from consecutive injection molding cycles are collected. A stable molding baseline and corresponding stable parameter snapshots are established using stable molding cycles. Process parameters include at least one of the following: segmented injection speed, segmented injection pressure, V / P switching parameters, holding pressure parameters, barrel temperature, back pressure, plasticizing screw speed, and mold temperature. For process parameter actions that change relative to the stable parameter snapshots, corresponding action carrier records are established according to the process parameter type, and associated with corresponding parameter response fingerprints. The action carrier record characterizes the expected action space, pressure propagation state, plasticized melt batch affiliation, or mold thermal state formation of the corresponding process parameter action. The parameter response fingerprint includes at least the necessary witness position, response state variable, response direction, and response sequence. The cavity status data of the current cycle is compared with the stable molding baseline to determine the process position or cavity position where the state deviation first occurs, as the first deviation position. Process parameter actions that meet the action reachability condition are selected from the process parameter actions that have not yet ended in the action carrier record. For the expected action space or pressure propagation state, the action reachability condition is that the corresponding action range has covered the first deviation position. For the allocation of plasticized melt batches, the condition for action is that the current mold uses the corresponding plasticized melt batch; for the formation of mold thermal state, the condition for action is that the corresponding mold thermal state has been formed; then, the necessary witness response corresponding to the first deviation position is matched with the corresponding parameter response fingerprint to form a responsibility candidate set; when there is a process parameter action in the responsibility candidate set, its corresponding process parameter is determined as the responsibility parameter; when there are multiple process parameter actions, the response direction and response order of each parameter response fingerprint in the necessary witness are compared. If there are distinguishable differences, the responsibility parameter is determined accordingly. If they are still indistinguishable, one process parameter is selected for isolation testing, and the responsibility parameter is determined based on the matching result of the state response formed by the isolation test and the corresponding parameter response fingerprint; the responsibility parameter is adjusted within the verified action range and safety process boundary of the responsibility parameter, and a correction action carrier of the same type as the corresponding action carrier of the responsibility parameter is established; after the correction action carrier meets the corresponding witness conditions, the actual recovery response is matched with the correction response fingerprint corresponding to the responsibility parameter. If they match, the parameter adjustment is confirmed to be effective; if they do not match, the parameter adjustment is not confirmed as effective.

[0010] In a preferred embodiment, establishing a stable molding baseline includes aligning the process data during the mold filling stage according to a unified screw position sequence, and aligning the process data after V / P switching according to a time sequence relative to the V / P switching node; obtaining stable samples from the continuous product qualified and process state stable modules, using the minimum and maximum values ​​of the effective stable samples of the state variables corresponding to the same alignment position as the lower and upper boundaries of the base, respectively; using the larger of the repeated detection fluctuation of the corresponding sensor and the undisturbed background fluctuation as the expansion amount, extending outwards to the lower and upper boundaries of the base to form a stable allowable zone at the corresponding position; the stable allowable zones at each alignment position form the stable molding baseline, and the process parameter combination of the corresponding stable molding module is saved as a stable parameter snapshot.

[0011] In a preferred embodiment, the action carrier corresponding to the segmented injection speed or segmented injection pressure is the equivalent injection melt segment, which represents the discharge range corresponding to the corresponding parameter action and its expected action space; the action carrier corresponding to the V / P switching parameter or holding pressure parameter is the pressure propagation range; the action carrier corresponding to the barrel temperature, back pressure or plasticizing screw speed is the plasticizing melt batch; and the action carrier corresponding to the mold temperature is the mold thermal state range.

[0012] In a preferred embodiment, the equivalent melt injection segment is established and updated as follows: The corresponding screw stroke is determined based on the screw start and end positions during the actual execution of the parameter action; the theoretical sweep space corresponding to the screw stroke is converted into the equivalent effective discharge amount according to the current effective discharge correction coefficient; the cumulative equivalent effective discharge amount at the start of the parameter action is used as the migration starting point, and the newly added equivalent effective discharge amount after the start of the parameter action is accumulated; using the same type of parameter action in at least three effective cycles under stable operating conditions, the newly added equivalent effective discharge amount corresponding to the formation of repeatable responses at each cavity detection position is established, and the allowable migration increment for each cavity detection position is established; when the newly added equivalent effective discharge amount corresponding to the current parameter action enters the allowable migration increment range of a cavity detection position, it is determined that the expected action space corresponding to the parameter action covers the cavity detection position, and when a state response conforming to the parameter response fingerprint appears at the cavity detection position, it is confirmed that the process parameter action forms an action witness at the cavity detection position.

[0013] In a preferred embodiment, the parameter response fingerprint is established by performing restricted parameter actions in two opposite directions relative to the reference parameter value under stable operating conditions. At least three valid moduli are obtained in each direction, and no new adjustment actions are generated for other process parameters that are not identifiable at the necessary witness positions of the process parameter to be calibrated. The repeated necessary witness positions, response state variables, response directions, response sequences, response intervals, and response amplitude ranges are written into the parameter response fingerprint. The parameter response fingerprint is also associated with a valid operating condition domain, which includes at least the material state, melt flow state, mold thermal state, and parameter action amplitude range. When the current operating condition exceeds the corresponding valid operating condition domain, the automatic responsibility access is suspended using the parameter response fingerprint.

[0014] In a preferred embodiment, before executing a new process parameter action, based on the type of the action carrier corresponding to the proposed process parameter action and the response position and response modulus information in the parameter response fingerprint, the expected effective modulus range, expected action area, and necessary witness position of the proposed process parameter action are determined, and the remaining effective modulus range, flow branch, action area, and necessary witness position of the process parameter action that has not yet been released are read; the proposed process parameter action is compared with the expected effective modulus range and the remaining effective modulus range of the process parameter action that has not yet been released, and it is determined whether there is any overlap in their flow branch, action area, and necessary witness position; if the expected effective modulus range and the remaining effective modulus range do not overlap, or if there is no overlap in their flow branch, action area, and necessary witness position, the proposed process parameter action is allowed to be executed; if there is overlap... When comparing the response direction, response time range, response amplitude range, and response order of the necessary witness positions of the corresponding parameter response fingerprints, if there is an identifiable difference between two process parameter actions in at least one of the comparison contents, the proposed process parameter action is allowed to be executed, and the corresponding necessary witness position is registered as the witness pre-occupancy position of the process parameter action. Before the corresponding witness is completed or the process parameter action is released, other process parameter actions that are not identifiable at that witness position are not executed. If none of the comparison contents can distinguish between the two process parameter actions, the proposed process parameter action is temporarily suspended, or the proposed process parameter action is executed in an isolation module that keeps other overlapping process parameters unchanged. After the proposed process parameter action is actually executed, the corresponding effective module range, action area, and necessary witness status are updated according to the actual formed action carrier and necessary witness results.

[0015] In a preferred embodiment, when establishing a plasticized melt batch, the amount of residual melt corresponding to the pad after the end of the previous injection cycle and the amount of fixed residual melt present in the screw head, nozzle, or hot runner are included in the equivalent residual amount of the previous batch. The ratio of the equivalent residual amount of the previous batch to the equivalent effective injection amount of the next cycle is used as the residual ratio of the previous batch. In consecutive cycles after the plasticizing parameters are restricted by a step, when the parameter response of the previous batch enters a stable background fluctuation range for at least three consecutive cycles, and the parameter response of the current plasticized melt batch can still be confirmed, the upper limit of the corresponding residual ratio of the previous batch is determined as the batch overlap threshold. When the residual ratio of the previous batch is greater than or equal to the batch overlap threshold, the responsibility candidate qualification for the plasticizing parameter action of the previous plasticized melt batch and the current plasticized melt batch is retained. When the residual ratio of the previous batch is lower than the batch overlap threshold, or the previous batch parameters exceed the last allowed response cycle, the responsibility candidate qualification for the plasticizing parameter action of the previous plasticized melt batch is released.

[0016] In a preferred embodiment, when multiple process parameter actions pass the responsibility access test, the allowable response range corresponding to the response fingerprint of each process parameter is converted into a directional incremental interval relative to the stable molding baseline. The lower bounds of each directional incremental interval are summed to obtain the combined lower boundary, and the upper bounds of each directional incremental interval are summed to obtain the combined upper boundary. The combined lower boundary and combined upper boundary are then extended outward using the background fluctuations at the corresponding detection positions to form a joint allowable range. If the actual state response exceeds the joint allowable range, or if a response position, response sequence, or response direction not included in any single parameter response fingerprint occurs after multiple process parameter actions overlap, the corresponding state change is recorded as an interaction residual, and a record of the parameter coupling effect associated with the overlapping process parameter actions is established. Single parameter isolation testing is prioritized for responsibility candidates with interaction residuals. After the interaction residuals of the same parameter combination are confirmed by at least three valid observations, a combination adjustment is allowed for the corresponding parameter combination, wherein the single adjustment amount of each process parameter does not exceed the valid action range and safe process boundary of the corresponding parameter that has been verified.

[0017] In a preferred embodiment, when the product quality is restored to the product qualification condition, but no recovery response matching the correction response fingerprint is formed after the correction action carrier meets the corresponding witnessing condition, this adjustment is recorded as an unconfirmed improvement, and further automatic adjustment of the responsibility parameter is frozen; when no matching response is formed after continuous observation until the last allowed witnessing position of the corresponding parameter response fingerprint or the last allowed response module, the responsibility parameter is restored to the value corresponding to the most recent stable parameter snapshot, provided that the equipment and product safety boundaries are met; the parameter recovery action is used as a rollback parameter action to establish a corresponding rollback action carrier and necessary witnessing record. Before the rollback action completes the necessary witnessing, no other process parameter adjustments that are not identifiable at the key witnessing position of the rollback parameter action are performed; when the action carrier record, witnessing record, and unconfirmed improvement record participating in this round of optimization have all completed the corresponding necessary witnessing, release, responsibility confirmation, or safe rollback, and there are no longer any pending witnessing, pending rollback, or unconfirmed states, and when the preset number of modules are continuously reached while simultaneously meeting the product qualification condition and stable molding baseline requirements, the stable parameter snapshot and stable molding baseline are updated.

[0018] The technical effects and advantages of this invention are as follows: Compared to existing methods that directly select adjustment parameters based on final product quality, single cavity state deviation, or historical correlation, this invention maps different process parameters to their actual operating states. Segmented injection parameters characterize the expected operating space through equivalent melt injection segments; V / P switching and holding pressure parameters characterize the pressure operating range through pressure propagation intervals; plasticizing parameters characterize the cross-mold operation through plasticized melt batches; and mold temperature characterizes the thermal state formation process through mold thermal state intervals. This allows for the determination of whether different parameters meet the conditions for explaining the current initial deviation, based on their respective operating mechanisms.

[0019] This invention uses both action reachability conditions and parameter response fingerprints to constrain responsibility candidates. A parameter action only enters the responsibility candidate set when the corresponding action state has met the reachability conditions for the initial deviation position, and when the necessary witnessing response state, response direction, and response sequence are consistent with the parameter fingerprint. This excludes parameters that have not yet acted on the abnormal region, or parameters that, although historically correlated with quality indicators, currently have no practical basis for action.

[0020] When multiple parameters simultaneously meet the liability conditions, the present invention further utilizes response direction, response time, response amplitude, and necessary witness sequence to distinguish the function of the parameters; when the existing witness conditions are insufficient to distinguish them, isolation probing is adopted to avoid accumulating new adjustment actions under the condition that multiple unidentifiable parameters change at the same time.

[0021] For non-single-parameter responses resulting from the overlap of multiple parameters, this invention combines the directional allowable response ranges corresponding to each single-parameter fingerprint into a joint allowable range. When the actual state exceeds this range, or when a new response position, response order, or response direction not present in the single-parameter fingerprint appears, the corresponding portion is recorded separately as an interaction residual, without being forcibly assigned to a single parameter. For parameter coupling effects that have been repeatedly confirmed, further restricted combination adjustments are implemented.

[0022] For parameters across mold cycles, such as barrel temperature, back pressure, and plasticizing screw speed, this invention incorporates the residual material in the gasket and the fixed amount of residue in the screw head, nozzle, or hot runner into the equivalent residual amount of the previous batch. It retains or releases the responsibility candidate qualification of the previous plasticizing melt batch by using the residual ratio of the previous batch and the batch overlap threshold, thus avoiding simply equating the action boundary of adjacent plasticizing batches with the mold cycle boundary.

[0023] After adjusting the responsibility parameters, this invention establishes a correction action carrier consistent with the original parameter action carrier type, and verifies the recovery process using the correction response fingerprint. If the product has been restored to qualified but lacks corresponding process witnesses, this status is retained as unconfirmed improvement, and the stable parameter snapshot is not directly updated; if necessary, the responsibility relationship is reconfirmed through a parameter recovery process with a rollback action carrier and necessary witnesses. The stable parameter snapshot and stable forming baseline are updated only after the relevant action records, witness records, interaction residues, and unconfirmed statuses have been processed, which can avoid temporary compensation or improvements of unconfirmed origin from entering the subsequent stable baseline. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Reference Figure 1 This implementation method is used for the optimization and control of process parameters in plastic injection molding. Injection speed, injection pressure, V / P switching, holding pressure, plasticizing parameters, and mold temperature all have different effects and response times during the injection cycle, and deviations in the same cavity state may be simultaneously affected by the actions of multiple parameters. During production, the current action status of different parameters is recorded, and the actual process response is used to determine whether a particular parameter has the responsibility to explain the initial deviation.

[0027] This implementation method can be executed by the injection molding machine's in-house controller, a stand-alone industrial controller, or an industrial computer that communicates with the injection molding machine's controller. The control equipment reads the screw position, actual injection speed, injection pressure, V / P switching position, V / P switching pressure, holding pressure, temperature of each section of the barrel, back pressure, plasticizing screw speed, mold temperature, material pad position, and cooling status, and also reads the status data output by the cavity pressure sensor or cavity temperature sensor inside the mold.

[0028] The process parameters processed in this embodiment include one or more of the following: segmented injection speed, segmented injection pressure, V / P switching parameters, holding pressure parameters, barrel temperature, back pressure, plasticizing screw speed, and mold temperature. Other injection molding parameters for which corresponding action carriers and parameter response fingerprints have not been established are not directly included in the automatic responsibility admission process of this embodiment.

[0029] Product quality results are linked to the batch number. Quality items that can be inspected online can be generated by online weighing equipment, visual inspection equipment, or dimensional inspection equipment; inspection items that cannot be completed online are inspected by subsequent inspection stations, and then written back to the corresponding production record through the batch number.

[0030] Product qualification criteria are pre-associated with the current product model or production task to determine whether the product quality of the corresponding mold batch meets the requirements for establishing a stable molding baseline, confirming parameter adjustment results, and updating stable parameter snapshots. Product qualification criteria can be determined by the current product's technical requirements, inspection specifications, or quality limits confirmed during the trial molding stage, and must include at least one of the following: product quality, critical dimensions, appearance defects, and molding integrity.

[0031] For quality items using online testing, the actual test results of the current module are compared with the corresponding quality limits. For quality items using subsequent inspection stations, the comparison is completed after the test results are written back through the module identifier. When all quality items required for testing in the current module meet the corresponding quality limits, the module is recorded as a qualified module. When any necessary quality item exceeds the corresponding quality limit, the module is recorded as a non-qualified module. When no valid test results have been obtained for a necessary quality item, the product qualification status of the current module remains pending confirmation and is not prematurely included as a qualified module in the establishment of stable molding baselines or the update of stable parameter snapshots.

[0032] Product qualification criteria are pre-associated with the current product model or production task to determine whether the product quality of the corresponding mold batch meets the requirements for establishing a stable molding baseline, confirming parameter adjustment results, and updating stable parameter snapshots. Product qualification criteria can be determined by the current product's technical requirements, inspection specifications, or quality limits confirmed during the trial molding stage, and must include at least one of the following: product quality, critical dimensions, appearance defects, and molding integrity.

[0033] For quality items using online testing, the actual test results of the current module are compared with the corresponding quality limits. For quality items using subsequent inspection stations, the comparison is completed after the test results are written back through the module identifier. When all quality items required for testing in the current module meet the corresponding quality limits, the module is recorded as a qualified module. When any necessary quality item exceeds the corresponding quality limit, the module is recorded as a non-qualified module. When no valid test results have been obtained for a necessary quality item, the product qualification status of the current module remains pending confirmation and is not prematurely included as a qualified module in the establishment of stable molding baselines or the update of stable parameter snapshots.

[0034] An injection molding cycle is divided into the following stages according to the actual execution sequence: plasticizing stage, initial mold filling stage, main mold filling stage, final mold filling stage, V / P switching stage, holding pressure and shrinkage compensation stage, and cooling stage. The screw position is used as the main process position axis during the mold filling stage, and the time axis relative to the switching node is used for comparison after the V / P switching.

[0035] This embodiment uses a 40mm nominal diameter injection molding machine screw for illustration. At the end of metering, the screw position is approximately 68mm, and the stable material pad position after injection is 4.8–5.2mm. The normal screw travel is approximately 63mm. For different equipment specifications, corresponding conversion relationships are established using the actual screw diameter, metering position, injection end position, and stable production data.

[0036] The mold has three cavity pressure detection positions along the main melt flow path: the gate side, the middle area, and the end area. Multi-branch or multi-cavity molds store the flow branch to which each detection position belongs, and do not directly treat state changes in unconnected branches as the same action propagation path.

[0037] Each injection molding cycle is assigned a unique cycle identifier, which is then linked to process parameter data, screw position data, cavity status data, plasticized melt batch, parameter action records, and product quality results for that cycle. The melt formed during the plasticizing stage and actually used in subsequent cycles is associated with the corresponding injection molding cycle using the plasticized melt batch identifier.

[0038] During the mold filling stage, screw position, injection pressure, and cavity pressure can be collected at a frequency of 1kHz. When comparing different mold runs, a uniform position sequence is established with a screw position interval of 0.1mm. If the original sampling point does not fall on the uniform position, the target position state value is calculated based on the position distance using adjacent actual sampling data before and after the target position. Data after V / P switching is aligned according to a uniform time interval, starting from the switching moment.

[0039] After the equipment completes preheating, mold temperature stabilization, and routine process debugging, a stable molding baseline is established from consecutive mold runs where the product qualification status has been confirmed and the process status is stable. This embodiment uses 20 consecutive effective stable mold runs as the baseline sample for the first time. In this embodiment, an effective mold run refers to a mold run where the target process parameter action is actually executed according to the current action record, the corresponding action carrier reaches the corresponding detection position or meets the corresponding action state, necessary detection data is generated, and there are no abnormal acquisitions from the relevant sensors. For parameter response fingerprint calibration, it is also required that there are no other newly added parameter actions that are indistinguishable from the target parameter at the necessary witness position. In addition to meeting the above conditions, the effective stable mold runs used for stabilizing the molding baseline also require that the product quality meets the product qualification conditions and that there are no abnormal records in the process status that have not yet been closed.

[0040] After obtaining stable samples at each uniform alignment position, the abnormal sampling is first checked to see if it belongs to sampling glitch or sensor malfunction. If a state variable has only a single abnormal sampling point, or the number of consecutive abnormal sampling points is less than the number of consecutive sampling points required for the current first deviation confirmation, and the state values ​​of adjacent positions before and after the abnormal sampling re-enter the current stable change range, and at the same time, the relevant screw motion state, injection pressure, or other related cavity state variables in the same mold do not show a continuous change synchronized with the abnormality, the corresponding abnormal sampling is recorded as a sampling glitch and is not used for the formation of the current position stable boundary.

[0041] If an abnormal state continuously reaches the number of consecutive sampling points required for the first deviation confirmation, or if other related state variables show continuous changes in the same direction during the abnormal period, it is not excluded as a sampling glitch. The corresponding data is retained, and it is checked whether the current module still meets the stable sample condition. After confirming that the data belongs to sampling glitch or sensor acquisition anomaly, the minimum state value in the remaining valid stable samples is used as the basic lower boundary, and the maximum state value is used as the basic upper boundary.

[0042] The fluctuations detected by repeated sensor readings and the fluctuations in the undisturbed background are obtained from stable production data, and the larger of the two is taken as the expansion amount. This expansion amount is applied to the lower boundary of the base data towards lower values, and to the upper boundary of the base data towards higher values, forming a stable allowable band at the current position. The expanded allowable band does not exceed the product qualification process boundary already confirmed during the trial molding stage.

[0043] The stable allowable bands formed at each uniform location collectively constitute the stable forming baseline. The combination of process parameters corresponding to the stable mold stages that establish this baseline is saved as a snapshot of the stable parameters.

[0044] When a certain position does not obtain enough effective stable samples, the position remains in a baseline state to be supplemented, and the allowable range of adjacent positions is not used to directly replace it. The stable working condition in this embodiment refers to the current product meeting the product qualification conditions, the key process state being in the corresponding stable allowable zone, the material state and mold thermal state being in the current effective range, and the production state without any pending rollback, unexplained interaction residues, or key sensor abnormalities.

[0045] In this embodiment, a stable operating condition means that the product qualification status of the current mold batch has been confirmed, the current key process status is within the stable allowable zone corresponding to the stable molding baseline, the material batch and material moisture content have not shown any abnormal changes exceeding the current production allowable range, the mold thermal status is in a confirmed stable thermal state, and there are currently no abnormal states such as pending rollback, unexplained interactive residues, key sensor abnormalities, or other unclosed abnormal states.

[0046] The stable molding baseline also preserves the current melt flow state and mold thermal state. The melt flow state can be characterized by the actual screw speed, injection pressure, and pressure of at least one cavity within the same filling section. The mold thermal state can be characterized by the actual temperatures of the moving mold, the critical locations of the stationary mold, and the temperature changes in successive molding cycles.

[0047] When performing parameter response fingerprint calibration, equivalent melt injection segment migration increment calibration, or other restricted parameter actions under stable operating conditions, all process parameters other than the target process parameter that are not identifiable at the necessary witness position should maintain their current stable values. If any of the following conditions are not met: current product quality, critical process status, material status, mold thermal status, or critical sensor status, the corresponding calibration cycle will not be considered a valid calibration cycle under stable operating conditions.

[0048] In this embodiment, the final injection speed parameter fingerprint corresponds to a melt flow state with a final injection speed of 51–59 mm / s, an intermediate cavity pressure of 24–27 MPa, and an injection pressure of 72–78 MPa. When the temperature at critical mold locations is between 58 and 62°C, and the temperature change between five consecutive mold passes does not exceed 0.5°C, the mold can be recorded as being in a stable thermal state. The above values ​​are based on the current implementation conditions and are not intended as fixed thresholds for other equipment.

[0049] Action vector records are established for process parameter actions that change relative to the stable parameter snapshot. Different action vectors are used for different parameters, and a single spatial coverage relationship is not used to describe all parameters.

[0050] Segmented injection speed or segmented injection pressure corresponds to the equivalent injection melt segment. The equivalent injection melt segment represents the discharge range corresponding to the parameter action and its expected action space, and is not used to track the absolute position of a certain actual melt micro-particle in the cavity.

[0051] The pressure propagation range corresponding to the V / P switching parameter or the pressure holding parameter indicates the area where the parameter action can continue to affect the current cavity melt state and the corresponding pressure response stage.

[0052] Barrel temperature, back pressure, or plasticizing screw speed correspond to the plasticized melt batch. The plasticized melt batch indicates the responsibility for the action of a set of plasticizing parameters and the actual melt entering the subsequent molding cycle.

[0053] The mold temperature corresponds to the mold's thermal state range. When the mold has multiple independent temperature control zones, the relationship between each temperature control zone, temperature detection position, and corresponding cavity area is established in advance. Each temperature control zone is associated with at least one temperature detection position that can reflect the actual temperature change in that zone, and the cavity areas or flow paths that the temperature control zone can affect are saved.

[0054] After a specific mold temperature parameter is activated, only the temperature detection locations associated with the corresponding temperature control area are checked for changes in thermal state. Once the associated temperature detection location has moved away from its original stable thermal state and entered a new stable thermal state according to the adjustment direction, the corresponding mold thermal state range is marked as formed. If the initial deviation is within the cavity area associated with the temperature control area, the current mold temperature action meets the corresponding reachability condition; if the initial deviation is within other cavity areas not associated with the temperature control area, the current mold temperature action does not explain the initial deviation. The mold thermal state range represents the thermal state formed and continuously existing by the actual mold temperature change after the mold temperature adjustment.

[0055] For the mold thermal state range, first determine whether there is a correlation between the cavity area to which the first deviation position belongs and the temperature control area corresponding to the current mold temperature parameter; if there is a correlation and the corresponding temperature detection position has confirmed the formation of a new actual mold thermal state, then the current mold temperature action is considered to meet the conditions for action attainment.

[0056] When the current mold uses overall mold temperature control, the same mold temperature parameter acts on the target cavity area currently being molded, and the mold thermal state range is no longer divided into independent action spaces according to the cavity position. After the mold temperature action is executed, the overall mold thermal state is confirmed to have changed in the direction of adjustment by checking the key temperature detection positions of the moving mold and the fixed mold; once the overall mold thermal state meets the current stability conditions, it is considered that the mold temperature action has met the conditions for effectively judging the responsibility of the current target cavity area.

[0057] Therefore, the reachability condition for liability determination is determined according to the type of action carrier. For the equivalent melt injection segment and pressure propagation range, the reachability condition is that the corresponding expected action space or pressure propagation range has covered the initial deviation position. For plasticized melt batches, the reachability condition is that the current mold batch actually uses the corresponding plasticized melt batch, and the corresponding plasticizing parameters are judged based on the residue from the previous batch to determine whether they still qualify for liability. For molds using overall mold temperature control, the reachability condition for the mold thermal state range is that the actual mold thermal state caused by the corresponding mold temperature action has been formed; this reachability condition applies to the target cavity area currently controlled by the same mold temperature parameter.

[0058] The equivalent effective discharge volume corresponding to the equivalent injected melt section is first obtained by calculating the theoretical sweep space based on the effective cross-sectional area of ​​the screw and the actual screw stroke, and then converted using the effective discharge correction coefficient.

[0059] Taking a 40mm screw and a normal forward stroke of approximately 63mm as an example, the theoretical sweep space is approximately 79.2cm. 3 The actual material weight of a single mold for a stable product is 80.6g, while the current ABS material uses 1.04g / cm³. 3 When used as a unified mass-volume conversion benchmark, the corresponding equivalent molding volume is approximately 77.5 cm³. 3 This results in an effective discharge correction factor of approximately 97.9%.

[0060] The converted volume here is an equivalent calculation quantity and does not represent the absolute true volume of the high-temperature and high-pressure melt inside the barrel.

[0061] When the material pad and actual product usage remain within the original stable range during continuous stable cycles, the current effective discharge correction factor is used. Recalibration is required when the material pad, actual product usage, screw assembly, check valve, or nozzle condition undergoes continuous changes.

[0062] For example, if the final injection speed starts at the 30mm position of the screw and ends at the 15mm position, corresponding to approximately a 15mm screw stroke, this can be converted to approximately 18.5cm. 3 The equivalent effective discharge volume is obtained, and the equivalent injected melt segment is formed for this parameter action.

[0063] First, a baseline flow coordinate system is established using stable mold cycles to determine the cumulative equivalent effective discharge volume and the cavity detection position. In this embodiment, the detection positions at the gate side, middle, and end can correspond to approximately 15.5–16.5 cm, respectively. 3 41~43cm 3 and 67-69cm 3 The range of cumulative equivalent effective emissions.

[0064] For the internal equivalent injected melt segment formed later during the mold filling process, the arrival amount of the whole mold melt front is not used directly to determine the effect of parameters.

[0065] The cumulative equivalent effective discharge volume at the start of the parameter action is used as the migration starting point, and the newly added equivalent effective discharge volume continues to accumulate from this starting point. Under the aforementioned stable operating conditions, at least 3 effective cycles are obtained for the same type of parameter action. The range of newly added equivalent effective discharge volume corresponding to the formation of repeatable parameter responses at each cavity detection position is recorded, and the allowable migration increment for the corresponding cavity detection position is established respectively.

[0066] During online operation, when the newly added equivalent effective discharge volume enters the allowable migration increment range of a certain cavity detection position, it is recorded that the expected action space of the current parameter action has already covered the detection position. At this time, the status response of the detection position is checked again.

[0067] If the expected scope of action has been covered, and a response matching the parameter response fingerprint appears at the corresponding location, the current parameter action is confirmed to have formed an action witness at that detection location. If the expected scope of action has been covered but there is no corresponding response, only the space reachability record is retained, and the current parameter is not automatically identified as the source of responsibility.

[0068] The arrival of the initial melt front at the detection position in the mold cavity can be confirmed by continuous pressure establishment. In this embodiment, at a sampling frequency of 1 kHz, the peak-to-peak value of the pressure noise in the stable non-arrival state is approximately no more than 0.28 MPa. Therefore, five consecutive sampling points exceeding 0.60 MPa and continuing to rise or remain above this value can be used as a set of arrival judgment conditions. After changing the mold or sensor, the corresponding threshold is recalibrated using actual noise.

[0069] After the V / P switching action is executed, the switching position and the current pressure at each cavity detection position are recorded, and the pressure changes after the switching action are continuously observed. When the corresponding pressure change forms a continuous response along the already filled cavity area, the pressure propagation range corresponding to the V / P switching action is updated.

[0070] The pressure holding process begins with the establishment of a pressure propagation record at the start of the corresponding pressure holding stage and continues until the final effective pressure is witnessed or the gate sealing and release conditions are met.

[0071] The gate closure reference range can be calibrated by gradually extending the holding time. With other process parameters kept stable, the final holding time is extended by 0.5 seconds each time, and at least three stable cycles are completed at each time level. If, after further extending the holding time, neither the product quality nor the gate-side pressure exhibits repeatable changes exceeding the stable background fluctuation, the corresponding time range is taken as the gate closure reference range.

[0072] When establishing a plasticized melt batch, the residual melt from the previous cycle is also considered. The amount of residual melt corresponding to the material pad after the end of the previous cycle is included in the equivalent residual amount of the previous batch; if there is fixed residual melt in the screw head, nozzle, or hot runner, the corresponding fixed residual amount is also included.

[0073] The fixed retention capacity can be determined by the known effective volume, or it can be equivalently calibrated by the decay of the continuous modal response after material switching or a step change in plasticizing parameters.

[0074] The ratio of the equivalent residual amount of the previous batch to the equivalent effective injection amount of the next batch is used as the residual ratio of the previous batch.

[0075] The batch overlap threshold is calibrated under steady-state conditions using a restricted step change in plasticizing parameters. Keeping injection speed, V / P switching, holding pressure, and mold temperature constant, the target plasticizing parameters are changed starting from a single plasticized melt batch, and the parameter responses and residual proportions from the previous batch are continuously recorded in subsequent mold runs.

[0076] When the current batch response can be confirmed, and the previous batch parameter responses for at least three consecutive cycles have entered a stable background fluctuation range, the upper limit of the corresponding previous batch residual ratio is saved as the batch overlap threshold for the current device. This embodiment can obtain a batch overlap threshold of approximately 10%.

[0077] When the residual proportion of the previous batch is greater than or equal to the batch overlap threshold, the corresponding plasticizing parameter actions of both the previous plasticized melt batch and the current plasticized melt batch retain their responsibility candidate status. Once the residual proportion of the previous batch falls below the batch overlap threshold, or the previous batch parameter exceeds its last permissible response modulus, the responsibility candidate status for the corresponding plasticizing parameter action of the previous plasticized melt batch is released. The responsibility candidate status for the previous batch is also released once the residual proportion of the previous batch falls below the batch overlap threshold, or the previous batch parameter exceeds its last permissible response modulus.

[0078] After adjusting the mold temperature parameters, the actual temperatures at key temperature detection locations on both the moving and fixed molds are continuously read. When any key temperature detection location deviates from its original stable thermal state according to the adjustment direction, a new mold thermal state range is established. Once all key temperature detection locations have entered the new stable range, and after a preset number of stable confirmation cycles are reached, there is no longer any continuous drift in the same direction exceeding the allowable variation, the current mold thermal state is recorded as established. In this embodiment, five consecutive cycles are used as the number of mold thermal state stability confirmation cycles.

[0079] After establishing the action carriers for each parameter, corresponding parameter response fingerprints are further established. A parameter response fingerprint must at least store the necessary witness positions, response state variables, response direction, and response sequence. It can also store the response interval, response amplitude range, the first allowed response position, the last allowed response position, and the last allowed response module corresponding to cross-module parameters. The first allowed response position, the last allowed response position, or the last allowed response module are used to limit the expected range within which the corresponding parameter actions can form a valid response.

[0080] Parameter response fingerprints can be established from existing isolation parameter adjustment records that meet the aforementioned stable operating conditions. When historical data is insufficient, restricted parameter actions are performed in two opposite directions relative to the current baseline parameter values ​​under stable operating conditions.

[0081] At least three valid patterns must be obtained in each direction. During fingerprint calibration, no new adjustments are made to other parameters that are not identifiable at the necessary witness locations as the target parameters.

[0082] In valid modulo operations in the same direction, when the necessary witness positions, response directions, and response sequences remain consistent, the corresponding states are written into the parameter response fingerprint. The minimum and maximum valid responses at each necessary witness position, combined with stable background fluctuations, form the response amplitude range, and the valid response intervals between adjacent necessary witnesses form the response interval range.

[0083] Taking the terminal injection velocity as an example, a baseline value of 55 mm / s can be used. In the first direction, the terminal injection velocity is increased to 59 mm / s, and in the second direction, it is decreased to 51 mm / s.

[0084] In a set of calibration states in this embodiment, after increasing the final speed, the pressure increments at the gate side, middle and end cavities can be approximately 0.30–0.40 MPa, 0.75–0.90 MPa and 1.00–1.20 MPa, respectively; the response interval from the gate side to the middle position is approximately 18–24 ms, and from the middle to the end is approximately 22–30 ms.

[0085] For parameters that propagate within the same injection mold, the response interval is expressed as the time interval between necessary witness positions; for parameters that affect plasticized melt batches across molds, the response interval is expressed as the difference in mold sequence number between the target plasticized melt batch entering the mold and the corresponding necessary response occurring in the mold; for mold temperature parameters, the time interval between the start of the mold temperature action and the formation of the corresponding mold thermal state or the difference in mold sequence number can be used.

[0086] When the reverse adjustment produces a corresponding reverse response, and the same principal positional order is maintained, they can jointly form a response fingerprint of the terminal injection velocity parameter.

[0087] Barrel temperature, back pressure, plasticizing screw speed, and mold temperature are parameters with relatively long response times. The second-direction action will not be executed until the final allowable response cycle of the first-direction action is completed or the corresponding thermal state has stabilized.

[0088] The parameter response fingerprint is simultaneously associated with the effective operating condition domain. The effective operating condition domain includes the material state, melt flow state, mold thermal state, and the verified range of parameter action amplitudes.

[0089] If the current operating condition exceeds the corresponding valid operating condition domain, automatic responsibility determination will not be completed by expanding the fingerprint allowable range. If a historical adjacent valid operating condition fingerprint exists, the corresponding fingerprint will be invoked; otherwise, the system will enter a re-witnessing state.

[0090] Before a parameter action enters the current first deviation responsibility candidate set, it must pass through two conditions: action reachability and response identity.

[0091] The current process status is continuously compared with the stable molding baseline. The filling stage is judged at a unified screw position, and the V / P switching and holding pressure stages are judged at a unified relative time.

[0092] In this embodiment, five consecutive sampling points exceeding the current position's stable allowable band are used as a set of initial deviation confirmation conditions. If the number of abnormal sampling points is less than five, and the subsequent state re-enters the stable allowable band, and other related state variables do not form synchronous and persistent anomalies, an initial deviation is not established; if five consecutive sampling points exceed the stable allowable band, they are no longer treated as sampling spikes, and the initial deviation candidate is still judged.

[0093] If the same sensor repeatedly exhibits abnormalities at the same or adjacent sampling positions in multiple consecutive modes, or if it experiences continuous zero-point drift, continuous lack of effective output, or long-term inconsistency with upstream and downstream related states, it will not be treated as a single sampling glitch. Instead, the sensor will be marked as pending confirmation, and the automatic responsibility judgment that relies on the necessary witness of this sensor will be suspended.

[0094] After confirming the deviation, save the mold number identifier, process stage, screw position or cavity detection position, deviation status variable, deviation direction and deviation magnitude, and take the position that first meets the deviation confirmation conditions as the first deviation position.

[0095] For the equivalent injected melt section and pressure propagation range, check whether the corresponding expected action space or pressure propagation range covers the initial deviation position.

[0096] For plasticized melt batches, check the actual plasticized melt batches used in the current mold and the overlap status of previous batches.

[0097] For the mold thermal state range, check whether the new thermal state caused by the corresponding mold temperature action has been formed.

[0098] Parameter actions that do not meet the conditions for reaching the corresponding function will not be included in the subsequent responsibility candidate.

[0099] After the action reachability condition is met, the necessary witness responses that have occurred corresponding to the first deviation are checked. If the response state variables, response direction, and response sequence at the necessary witness location are consistent with the current parameter response fingerprint, the parameter action is added to the responsibility candidate set.

[0100] When the candidate set of responsibilities contains only one parameter action, the corresponding process parameter is determined as the responsible parameter.

[0101] When multiple responsibility candidates exist, first compare the response direction and order of each parameter in the necessary witnessing events that have already occurred. If different parameters have opposite response directions or different necessary witnessing orders, mismatched parameters can be eliminated accordingly.

[0102] When multiple candidates still exist, the response time range and response amplitude range can also be compared.

[0103] There is a direction difference condition when two parameters have opposite response directions at at least one necessary witness position; there is a time difference condition when the response time ranges do not overlap; there is an amplitude difference condition when the response amplitude ranges after adding background fluctuations do not intersect; there is an order difference condition when the necessary witness order is different and the current sampling conditions can distinguish the corresponding order differences.

[0104] If no identifiable differences are found, the responsibility parameter is not determined by using slight numerical differences. Instead, one of the candidate parameters is selected for isolation and trial.

[0105] During the isolation trial, other unidentifiable responsibility candidate parameters remain unchanged. After the corresponding action carrier of the trial parameter meets the necessary witnessing conditions, the generated state response is compared with the response fingerprint of that parameter. If they match, the current trial parameter is confirmed to have a responsibility role; if they do not match, its current responsibility candidate status is released, and the remaining candidate parameters are evaluated.

[0106] Before new process parameter actions are executed, a pre-emptive action carrier and a witness pre-emptive action are established. Since the process parameter actions to be executed have not yet been actually executed, their scope of action is represented by the expected state. Parameter actions that have not yet occurred are not directly recorded as having an actual effective scope.

[0107] The expected effective cycle range for the proposed process parameter actions is determined based on the corresponding action carrier type and the current effective parameter response fingerprint. For segmented injection parameters, V / P switching parameters, and holding pressure parameters that complete the main action propagation within the current injection cycle, the expected action duration range is determined by combining the expected execution position of the current parameter action, the first allowable response position of the corresponding action carrier, and the last allowable response position. The cycles containing this action duration range are then considered as the expected effective cycles. For parameters such as barrel temperature, back pressure, and plasticizing screw speed that have an effect across cycles through plasticized melt batches, the cycle expected to use the corresponding plasticized melt batch is used as the starting range, and the expected effective cycle range is determined by combining the last allowable response cycle in the parameter response fingerprint. For mold temperature parameters, the cycle that enters the thermal transition state after the mold temperature action is executed is used as the expected action starting point, and the expected effective cycle range is determined by combining the corresponding mold thermal state formation and stabilization conditions.

[0108] The expected area of ​​action and necessary witnessing locations for the proposed process parameter actions are determined based on the corresponding action carrier and the current parameter response fingerprint. The expected area of ​​action for segmented injection parameters is determined based on the expected action space corresponding to the equivalent injected melt segment; V / P switching parameters and holding pressure parameters are determined based on the expected pressure propagation range; plasticizing parameters are determined based on the actual impact of the corresponding plasticized melt batch on subsequent molds; and mold temperature parameters are determined based on the corresponding mold thermal state range.

[0109] For process parameter actions that have been executed but not yet released, read the actual action carrier that has been formed, the necessary witnesses that have been completed, and the remaining necessary witnesses, and determine the range of remaining effective modules based on the action segments or response modules that have not yet been completed.

[0110] The controller compares the expected effective modulus range of the proposed process parameter action with the remaining effective modulus range of the process parameter actions that have not yet been released, and further compares their flow branches, areas of action, and necessary witness locations. If the expected effective modulus range and the remaining effective modulus range do not overlap, or if they do not share a common flow branch, a common area of ​​action, or a common necessary witness location, no current witness conflict is formed, and the proposed process parameter action is allowed to be executed.

[0111] When there is overlap in the effective modulus range, and also overlap in the area of ​​effect and the necessary witness position, the response direction, response time range, response amplitude range, and necessary witness order of the two parameter response fingerprints are compared. If at least one of these factors can reliably distinguish the difference between the two parameter actions, the new parameter action is allowed to execute, and the corresponding necessary witness position is registered as the witness pre-occupancy position for that parameter action. Before the necessary witness is completed or the current parameter action is released, no other parameter adjustments that are not identifiable at that witness position for that parameter action are executed.

[0112] If no identifiable conditions exist, the action of the next parameter should be suspended. If it is necessary to continue to confirm the effect of the parameter, an isolation module can be established, and the action of the process parameter to be executed can be performed while keeping other unidentifiable overlapping parameters unchanged.

[0113] After the planned process parameter action is actually executed, the expected action state before execution is no longer used as the basis for final responsibility judgment. Instead, the actual execution location, actual parameter value and corresponding action carrier state of the parameter action are read, and the corresponding effective module range and action area are updated. As the necessary witnesses are completed in sequence, the remaining necessary witnesses and release status of the parameter action are updated.

[0114] When at least one reliably identifiable difference exists, a new parameter action can be executed, and the corresponding necessary witness position will be registered as the witness pre-position for that parameter action. No other parameter adjustments that are not identifiable at that witness position will be executed until the necessary witness is completed or the current parameter action is released.

[0115] If no identifiable conditions exist, postpone the action of the next parameter, or establish an isolated module for execution.

[0116] When multiple parameters pass the responsibility access test simultaneously, the combined response can be further examined.

[0117] The allowable response range corresponding to each parameter response fingerprint is first converted into a directional increment range relative to the stable forming baseline. Changes above the baseline are taken as positive increments, and changes below the baseline are taken as negative increments.

[0118] The lower bounds of the incremental intervals in each band direction participating in the current combination are summed, and the sum is used as the lower boundary of the combination; the upper bounds of the incremental intervals in each band direction are summed, and the sum is used as the upper boundary of the combination; then the lower boundary and upper boundary of the combination are extended outward according to the stable background fluctuations at the corresponding detection positions to form a joint allowable range.

[0119] For example, the allowable pressure increment at the intermediate detection position after increasing the terminal injection rate is 0.75–0.90 MPa, the allowable pressure increment after adjusting the V / P switching position is 0.30–0.50 MPa, and when the current background fluctuation is ±0.15 MPa, the combined explanatory range of the two unidirectional effects can be approximately 0.90–1.55 MPa.

[0120] When the actual response remains within the joint permissible range, the existing single-parameter responses still have a common interpretation possibility. When the actual response exceeds the joint range, or when a new response position, new response order, or new response direction appears that does not exist for any single-parameter fingerprint, the part that cannot be interpreted by the current single-parameter fingerprint will be recorded as an interaction residual.

[0121] Interactive residual associations participate in the currently overlapping parameter actions to form parameter coupling effect records.

[0122] When an interaction residual first appears, single-parameter isolation is prioritized for probing, without immediate combination adjustments. Combination adjustments are only permitted after the same parameter combination has repeatedly generated the same type of interaction residual in at least three valid observations.

[0123] The combined adjustment only changes the parameters that have already entered the corresponding parameter coupling effect record. The single action amplitude of each parameter does not exceed its own verified effective action amplitude, while satisfying the safety boundaries of the equipment and mold.

[0124] Safety process boundaries are primarily derived from the rated parameters of the injection molding machine, mold process cards, and material processing specifications. Product process boundaries established during the trial molding stage can also serve as constraints. When multiple boundaries are inconsistent, the boundary with the stricter constraint is used.

[0125] Once the responsibility parameters are determined, parameter adjustments are performed within the verified range of motion and safety process boundaries.

[0126] For example, if the stable value of the terminal injection speed is 55mm / s, and valid fingerprints have been obtained in both directions of 51mm / s and 59mm / s, then the directly supported single-step action range is ±4mm / s. If it is necessary to further increase to 63mm / s, first perform an adjustment from 55mm / s to 59mm / s, and decide whether to continue to the next step after the current adjustment has been effectively verified.

[0127] After adjusting the responsibility parameters, a corrective action carrier of the same type as the original action carrier is established.

[0128] After adjusting the final injection speed, a corrected equivalent injection melt segment is established; after switching V / P or adjusting the holding pressure parameters, a corrected pressure propagation range is established; after adjusting the plasticizing parameters, a corresponding corrected plasticized melt batch is established; after adjusting the mold temperature, a corrected mold thermal state range is established.

[0129] The corrected response fingerprint retains the original parameters, necessary witness positions, and response order, and the response direction is determined to be the direction of recovery towards the stable forming baseline.

[0130] After the modified action carrier meets the corresponding action reachability and witnessing conditions, check the actual recovery response. When the necessary witnessing location forms a recovery response according to the corresponding modified fingerprint, the validity of this parameter adjustment is confirmed.

[0131] When the upstream necessary witness has been restored, but the downstream still has deviations, the confirmed adjustment amount is retained, and the remaining downstream deviations are re-formed into the first deviation candidate, without continuing to unconditionally expand the same parameter.

[0132] If the product quality has been restored to the qualified condition, but no corresponding correction response fingerprint is formed after the correction action carrier meets the necessary witnessing conditions, an unconfirmed improvement record is established.

[0133] During the period of unconfirmed improvement, further automatic adjustments to the current responsibility parameters will be frozen, and other optimization actions that are not identifiable by the current necessary witness will be suspended. Continuous monitoring will continue until the current parameters reach the last allowed witness position or the last allowed response module.

[0134] If a matching recovery response cannot be formed after the observation period, this product improvement will not be recorded as a confirmed adjustment effect of the current responsible parameter. When equipment and product safety permits, the parameter will be restored to the value corresponding to the most recent stable parameter snapshot.

[0135] The parameter recovery action is treated as a new rollback parameter action, and a rollback action carrier of the same type and necessary witness records are established. No other parameter actions that are not identifiable at the key witness location related to this rollback action are executed until the rollback witness is completed.

[0136] If the product remains in acceptable condition after a rollback, the priority liability eligibility for the previously assigned liability parameters is released. If the original deviation reappears after a rollback, and the rollback response matches the original parameter response fingerprint, this process is considered a new liability witness.

[0137] If the conditions for safe rollback cannot be met, parameters will not be forcibly restored. Instead, the corresponding responsibility record will remain in an open state and will be transferred to the safety policy or manual confirmation.

[0138] When no parameters meet the liability access criteria for the first deviation, an unattributed deviation record is established. At this time, the material batch, material moisture content, cooling water condition, equipment thermal condition, material pad condition, check valve condition, and sensor condition are checked.

[0139] Material batches can be obtained from the production management system, equipment formula records, or manual batch records. The moisture content of the material can be read from online moisture content testing data, or offline testing results of the corresponding material batch can be used.

[0140] The cooling water status can read the inlet water temperature, return water temperature, and flow rate of the mold temperature controller or cooling circuit. The equipment thermal status can read the barrel temperature and the temperature data of the existing hydraulic or servo system.

[0141] When there is no dedicated anti-reverse detection device, the anti-reverse status can be determined by combining the position of the pad in the continuous die, the actual material usage changes of the product under the same screw stroke, and the pressure build-up status in the initial injection stage.

[0142] If any inspection item lacks valid data, the item will not be directly recorded as normal, but will remain in the state of "unconfirmed cause".

[0143] Stable parameter snapshots and stable molding baselines are not updated immediately after a mold pass.

[0144] The number of stable state verification cycles is preset. This number is determined based on the product production cycle time, quality inspection cycle, and fluctuations between normal cycles. In this embodiment, the number of stable state verification cycles is set to 10 cycles.

[0145] If the preset number of consecutive molding cycles simultaneously meet the product qualification conditions and stable molding baseline requirements, and all records involved in this round of optimization have been completed, then updating the stable parameter snapshot and stable molding baseline is allowed.

[0146] The record ends when the following conditions are met: the equivalent injection melt segment has completed the final necessary witnessing or clear release; the pressure propagation range has completed the final effective witnessing or meets the release conditions; the plasticized melt batch has exceeded the allowable response mold number and there is no responsibility for the unreleased previous batch; the mold thermal state has stabilized or completed safe handling; witnessing pre-occupancy has been completed or cancelled; interactive residues have been explained through isolation testing or no longer recur; and unconfirmed improvements have completed responsibility confirmation or safe rollback.

[0147] If there are unwitting, unrollback, unexplained interaction remnants, or other unresolved responsibility states, a stable state update will not be performed.

[0148] If the preset number of cycles are achieved consecutively, simultaneously meeting the product qualification conditions and stable molding baseline requirements, and all necessary witnessing, release, responsibility confirmation, or safety rollback procedures have been completed for the action carrier records, witnessing records, and unconfirmed improvement records involved in this round of optimization, then updating the stable parameter snapshot and stable molding baseline is permitted.

[0149] When this round of optimization involves parameter coupling records or interaction remnants, it is also required that the interaction remnants have been explained through isolation probing, formed confirmed parameter coupling relationships, or have been confirmed by subsequent observations to no longer recur. If unexplained interaction remnants still exist, stable parameter snapshots and stable baseline updates will not be performed.

[0150] The parameter response fingerprint is not updated immediately after a single valid adjustment. The corresponding parameter response fingerprint is adjusted only when the same parameter has obtained at least three confirmed valid actions within the same or adjacent valid operating condition domains, and the necessary witnessing order, response direction, and main response amplitude relationship remain consistent.

[0151] Taking the low end filling pressure as an example, during the current stable production process, the gate side detection position is still within the stable allowable range. The middle detection position first forms a continuous low pressure at about 34mm screw position, and continues to spread to the end detection position.

[0152] The intermediate detection position is recorded as the initial deviation position.

[0153] The expected action space of the equivalent injected melt segment corresponding to the current terminal injection speed has covered the intermediate detection area. At the same time, the necessary witness that meets the response fingerprint of the current terminal injection speed has already appeared before the intermediate detection position. Therefore, the terminal injection speed satisfies the two conditions of action reachability and response identity.

[0154] The pressure holding phase has not yet started, and the corresponding pressure holding propagation range does not cover the intermediate detection area. Therefore, the pressure holding parameters are not included in the responsibility candidate.

[0155] The current batch of plasticized melt used in the mold has no new plasticizing parameters, the residual ratio of the previous batch is already below the batch overlap threshold, and there have been no new changes in the mold's thermal state.

[0156] The current candidate set of liability only includes the terminal injection rate, therefore the terminal injection rate is determined as the liability parameter.

[0157] The current stable value is 55 mm / s. Existing parameter response fingerprints indicate that increasing the final injection speed can restore the intermediate and final pressures to the stable molding baseline. When safe process boundaries allow, the final injection speed will be adjusted to 59 mm / s for the next complete mold run.

[0158] The adjustment action forms a new corrected equivalent injection melt segment. Once the expected effective space covers the gate side, middle, and end detection areas, and the actual pressure recovers to the stable allowable zone according to the corresponding corrected response sequence, the effectiveness of this round of adjustments is confirmed.

[0159] If the intermediate detection position is restored but the end still has a deviation, maintain the confirmed adjustment amount of 59mm / s and re-enter the responsibility screening for the remaining abnormality at the end.

[0160] When a product returns to compliance but no corresponding corrective response fingerprint is generated, it enters the unconfirmed improvement process and is not directly included in the new stable molding baseline.

[0161] After the preset number of times the stable state confirmation mode is reached, and all action carrier records, witness records, interaction residual records and unconfirmed improvement records have completed the necessary witnessing, release, responsibility confirmation or safe rollback, a new stable parameter snapshot is saved and the stable forming baseline is updated.

[0162] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method of optimizing control of parameters of a plastic injection molding process, characterized in that, include: Collect process parameter data, screw position data and cavity state data in continuous injection molding cycles, and establish a stable molding baseline and corresponding stable parameter snapshots using stable molding cycles. The process parameters include at least one of the following: segmented injection speed, segmented injection pressure, V / P switching parameters, holding pressure parameters, barrel temperature, back pressure, plasticizing screw speed and mold temperature. For process parameter actions that change relative to the stable parameter snapshot, establish corresponding action carrier records according to the process parameter type and associate them with corresponding parameter response fingerprints. The action carrier record characterizes the expected action space, pressure propagation state, plasticized melt batch affiliation, or mold thermal state formation of the corresponding process parameter action. The parameter response fingerprint includes at least the necessary witness location, response state variable, response direction, and response sequence. Compare the cavity state data of the current mold batch with the stable molding baseline to determine the process location or cavity location where the state deviation first occurs, and use this location as the first deviation location. From the process parameter actions that have not yet ended recorded by the action carrier, process parameter actions that meet the action reachability conditions are selected. Specifically, for the expected action space or pressure propagation state, the action reachability condition is that the corresponding action range has covered the first deviation position; for the plasticized melt batch affiliation, the action reachability condition is that the current mold uses the corresponding plasticized melt batch; for the mold thermal state formation status, the action reachability condition is that the corresponding mold thermal state has been formed. Then, the necessary witness responses that have occurred corresponding to the first deviation position are matched with the corresponding parameter response fingerprints to form a set of responsibility candidates. When there is a process parameter action in the candidate set of responsibility, its corresponding process parameter is determined as the responsibility parameter; when there are multiple process parameter actions, the response direction and response order of each parameter response fingerprint in the necessary witnessing have been compared. If there are distinguishable differences, the responsibility parameter is determined accordingly. If they are still indistinguishable, one of the process parameters is selected for isolation trial, and the responsibility parameter is determined based on the matching result of the state response formed by the isolation trial and the corresponding parameter response fingerprint. Adjust the responsibility parameters within the verified range of action and safety process boundaries, and establish a corrective action carrier of the same type as the action carrier corresponding to the responsibility parameters; After the correction carrier meets the corresponding witnessing conditions, the actual recovery response is matched with the correction response fingerprint corresponding to the responsibility parameter. If a match is found, the parameter adjustment is confirmed to be valid. If no match is found, the parameter adjustment is not confirmed as a valid adjustment.

2. The method of claim 1, wherein: Establishing a stable molding baseline includes: The process data during the mold filling stage is aligned according to a unified screw position sequence, and the process data after V / P switching is aligned according to the time sequence relative to the V / P switching node. Stable samples are obtained from the continuous product qualified and process state stable modules. The minimum and maximum values ​​of the effective stable samples of the state variables corresponding to the same alignment position are used as the basic lower boundary and the basic upper boundary, respectively. The larger of the repeated detection fluctuation of the corresponding sensor and the undisturbed background fluctuation is used as the expansion amount to expand the basic lower boundary and the basic upper boundary outward to form the stable allowable band at the corresponding position. The stable forming baseline is formed by the stable allowable bands at each alignment position, and the process parameter combination of the corresponding stable forming die is saved as a stable parameter snapshot.

3. The method of claim 1, wherein: The action carrier corresponding to the segmented injection speed or segmented injection pressure is the equivalent injection melt segment. The equivalent injection melt segment characterizes the discharge range corresponding to the action of the corresponding parameters and its expected action space. The action medium corresponding to the V / P switching parameters or pressure holding parameters is the pressure propagation range; The action carrier corresponding to the barrel temperature, back pressure, or plasticizing screw speed is the batch of plasticized melt. The temperature of the mold corresponds to the thermal state range of the mold.

4. The method of claim 3, wherein: The equivalent injected melt segment is established and updated in the following manner: Based on the screw start position and screw end position during the actual execution of the parameter action, the corresponding screw stroke is determined, and the theoretical sweep space corresponding to the screw stroke is converted into the equivalent effective discharge amount according to the current effective discharge correction coefficient; The cumulative equivalent effective discharge volume at the start of the parameter action is used as the migration starting point, and the cumulative new equivalent effective discharge volume after the start of the parameter action is accumulated. By utilizing the same type of parameter action in at least 3 effective cycles under stable operating conditions, the corresponding new equivalent effective discharge amount is generated when a repeatable response is formed at each cavity detection position, and the allowable migration increment at each cavity detection position is established. When the newly added equivalent effective discharge volume corresponding to the current parameter action enters the allowable migration increment range of a cavity detection position, it is determined that the expected action space corresponding to the parameter action covers the cavity detection position. When a state response that matches the parameter response fingerprint appears at the cavity detection position, it is confirmed that the process parameter action forms an action witness at the cavity detection position.

5. The method of claim 1, wherein: The parameter response fingerprint is established by performing constrained parameter actions in two opposite directions relative to the baseline parameter value under stable operating conditions. At least three effective moduli are obtained in each direction, and no new adjustment actions are generated for other process parameters that are not identifiable from the process parameters to be calibrated at the necessary witness positions in the effective moduli. Write the recurring necessary witness locations, response state variables, response direction, response order, response interval, and response range into the parameter response fingerprint; The parameter response fingerprint is also associated with the effective operating condition domain, which includes at least the material state, melt flow state, mold thermal state, and parameter action range. When the current operating condition exceeds the corresponding effective operating condition domain, the automatic responsibility access is suspended using the parameter response fingerprint.

6. The method for optimizing and controlling process parameters in plastic injection molding according to claim 5, characterized in that: Before executing a new process parameter action, based on the type of action carrier corresponding to the process parameter action to be executed and the response position and response modulus information in the parameter response fingerprint, determine the expected effective modulus range, expected action area and necessary witness position of the process parameter action to be executed, and read the remaining effective modulus range, flow branch, action area and necessary witness position of the process parameter action that has not yet been released. Compare the expected effective range and remaining effective range of the process parameter actions to be executed with those of the process parameter actions that have not yet been released, and determine whether there is any overlap in their flow branches, areas of action, and necessary witness locations. If the expected effective module range does not overlap with the remaining effective module range, or if their flow branches, action areas, and necessary witness positions do not overlap, the execution of the proposed process parameter action is permitted. When there is overlap, compare the response direction, response time range, response amplitude range, and response order of the corresponding parameter response fingerprints. When two process parameter actions have identifiable differences in at least one comparison content, the proposed process parameter action is allowed to be executed, and the corresponding necessary witness position is registered as the witness pre-occupancy position of the process parameter action. Before the corresponding witness is completed or the process parameter action is released, other process parameter actions that are not identifiable at that witness position are not executed. If none of the comparison criteria can distinguish between the actions of the two process parameters, the action of the process parameter to be executed shall be temporarily suspended, or the action of the process parameter to be executed shall be executed in an isolation module while keeping other overlapping process parameters unchanged. After the planned process parameters are actually executed, the corresponding effective module range, action area, and necessary witness status are updated based on the actual action carrier formed and the necessary witness results.

7. The method for optimizing and controlling process parameters in plastic injection molding according to claim 3, characterized in that: When establishing a plasticized melt batch, the amount of residual melt corresponding to the pad after the end of the previous injection cycle, as well as the amount of fixed residual melt in the screw head, nozzle or hot runner, are included in the equivalent residual amount of the previous batch. The ratio of the equivalent residual amount of the previous batch to the equivalent effective injection amount of the next batch is used as the residual ratio of the previous batch. In consecutive modulus cycles after a limited step in plasticizing parameters, when the parameter responses of the previous batch in at least three consecutive cycles enter a stable background fluctuation range, and the parameter responses of the current plasticized melt batch can still be confirmed, the upper limit of the corresponding residual proportion of the previous batch is determined as the batch overlap threshold. When the residual ratio of the previous batch is greater than or equal to the batch overlap threshold, the candidate qualification for the action of the corresponding plasticizing parameter of the previous plasticized melt batch and the current plasticized melt batch is retained; when the residual ratio of the previous batch is lower than the batch overlap threshold, or the parameters of the previous batch exceed the last allowed response mode, the candidate qualification for the action of the corresponding plasticizing parameter of the previous plasticized melt batch is released.

8. The method for optimizing and controlling process parameters in plastic injection molding according to claim 5, characterized in that: When there are multiple process parameter actions that pass the responsibility access, the allowable response range corresponding to the response fingerprint of each process parameter is converted into a band direction increment interval relative to the stable forming baseline. The lower bound of each band direction increment interval is summed to obtain the combined lower boundary, and the upper bound of each band direction increment interval is summed to obtain the combined upper boundary. The combined lower boundary and combined upper boundary are then extended outward using the background fluctuations at the corresponding detection positions to form a joint allowable range. When the actual state response exceeds the joint allowable range, or when the response position, response sequence or response direction is not included in any single parameter response fingerprint after multiple process parameter actions overlap, the corresponding state change is recorded as an interaction residual, and a record of the parameter coupling effect associated with the overlapping process parameter actions is established. For candidates with interactive residuals, single-parameter isolation testing should be prioritized; after at least three valid observations confirm the interactive residuals of the same parameter combination, combined adjustments to the corresponding parameter combination are allowed, wherein the single adjustment amount of each process parameter shall not exceed the valid action range and safe process boundary of the corresponding parameter that has been verified.

9. The method for optimizing and controlling process parameters in plastic injection molding according to claim 1, characterized in that: If the product quality is restored to the product qualification condition, but the corrective action carrier fails to form a recovery response that matches the corrective response fingerprint after meeting the corresponding witnessing conditions, this adjustment will be recorded as an unconfirmed improvement, and further automatic adjustment of the responsibility parameters will be frozen. If no matching response is formed after continuous observation until the last allowed witness position of the corresponding parameter response fingerprint or the last allowed response module, the responsibility parameter will be restored to the value corresponding to the most recent stable parameter snapshot, provided that the safety boundaries of the device and product are met. The parameter recovery action establishes a corresponding rollback action carrier and necessary witness records as a rollback parameter action. Before the rollback action completes the necessary witnessing, no other process parameter adjustments that are not identifiable at the key witnessing position of the rollback parameter action are executed. All action carrier records, witness records, and unconfirmed improvement records involved in this round of optimization have completed the necessary witnessing, release, responsibility confirmation, or safe rollback, and there are no longer any pending witnessing, rollback, or unconfirmed states. Furthermore, when the preset number of cycles are reached and the product qualification conditions and stable molding baseline requirements are met simultaneously, the stable parameter snapshot and stable molding baseline are updated.

Images