Balanced runner and injection speed control method of multi-cavity mold
By collecting and analyzing melt flow parameters in real time in a multi-cavity mold and dynamically adjusting the injection speed, the problem of insufficient accuracy in real-time adaptive control of injection speed in a multi-cavity mold is solved. This achieves uniform melt filling and stability of the production process, thereby improving product quality and production efficiency.
Patent Information
- Application Number
- CN202610078899.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-01-21
- Publication Date
- 2026-05-05
AI Technical Summary
Existing multi-cavity molds lack sufficient precision in real-time adaptive control of injection speed during die casting, resulting in uneven melt filling, increased product defects, and low production efficiency. They also cannot effectively respond to real-time process changes such as melt temperature fluctuations, mold wear, or alloy composition changes.
By arranging sensors in the flow channels and cavities of a multi-cavity mold to collect melt flow parameters in real time, the injection speed is dynamically adjusted by comparing the control system with the balanced flow channel model. The flow requirements between the multi-cavity mold are optimized by combining the influence of melt viscosity and mold resistance, and the continuity and accuracy of the adjustment are ensured by closed-loop feedback control.
It achieves real-time adaptive control of balanced flow channels and injection speed in multi-cavity molds, improving product quality and production efficiency, reducing the probability of defects, and enhancing the traceability and maintainability of the control process.
Smart Images

Figure CN121972632A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of metal casting technology, specifically to a method for controlling the balanced flow channel and injection speed of a multi-cavity mold. Background Technology
[0002] In the die casting industry, multi-cavity molds are widely used in the mass production of high-precision products such as automotive parts and electronic housings. The balanced flow channel design and injection speed control method directly affect the uniformity of melt filling, product defect rate, and production efficiency. In traditional die casting processes, the injection speed is usually adjusted by preset fixed parameters to achieve a balanced melt distribution among multiple cavities. However, with the development of product complexity and material diversification, existing control methods are difficult to cope with real-time variables, such as melt temperature fluctuations, mold wear, or changes in alloy composition, leading to uneven filling, increased porosity or cold shut defects, and thus affecting product quality and yield. For example, invention patent CN102717488A discloses a mold runner structure, including a first runner plate and a second runner plate, which achieves uniform melt distribution through multi-layer runner connection. The advantage of this patent is that it optimizes the static runner geometry, reduces pressure loss at branch points, and improves the consistency of multi-cavity filling. However, this method relies on a preset runner design and cannot dynamically adjust the injection speed to respond to real-time process changes, leading to amplified differences in melt flow resistance and limited filling accuracy during high-pressure die casting. CN106378432A discloses a die casting method for aluminum alloy steering knuckles, which controls the filling speed by adjusting the injection speed and the cross-sectional area of the ingate, and sets the mold temperature within the range of 150-280℃. The advantage of this patent is the introduction of multi-stage injection speed adjustment (such as low-speed advancement and high-speed filling), effectively reducing the risk of sputtering and premature solidification, and improving product density. However, its control is still based on static parameter settings, lacking a closed-loop feedback mechanism, and cannot adapt to melt viscosity or external interference in real time, resulting in increased filling deviation between cavities and a higher product defect rate in multi-cavity mold applications. While the aforementioned existing technologies have improved the balance of flow channels and injection speed control to some extent, they generally suffer from insufficient accuracy in real-time adaptive control of injection speed. Specifically, existing preset parameter methods cannot accurately respond to instantaneous process fluctuations, resulting in uneven filling of multi-cavity molds, increased product defects, and low production efficiency. Therefore, a control method that can monitor and adaptively adjust the injection speed in real time is needed to improve the accuracy and reliability of multi-cavity die casting. Summary of the Invention
[0003] To address the shortcomings of existing technologies, this invention provides a method for balanced flow channels and injection speed control in multi-cavity molds, which solves the problem of insufficient real-time adaptive control accuracy of injection speed in traditional methods.
[0004] To achieve the goal of improving the precision and reliability of multi-cavity die casting mentioned in the background section, the present invention provides the following technical solution: A method for controlling the balanced flow channel and injection speed of a multi-cavity mold, comprising: S1: After the die-casting machine is started, the key parameters of the melt flow are collected in real time by sensors arranged in the flow channels and cavities of the multi-cavity mold to obtain the initial flow state data. S2: Transmit the parameters collected in real time to the control system and compare them with the preset balanced flow channel model. Analyze the differences between parameters to identify potential areas of uneven melt flow and provide a basis for subsequent adjustments. S3: Based on the difference analysis results, generate injection speed adjustment commands and optimize the flow requirements between multiple cavities according to the influence of melt viscosity and mold resistance; S4: The adjustment command is sent to the injection mechanism, which dynamically adjusts the speed curve of the injection plunger according to the command and transitions from the slow stage to the fast stage to adapt to the real-time flow conditions. S5: During the adjustment process, the changes in melt parameters continue to be fed back in real time through sensors. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment. S6: Repeat the feedback and adjustment process until the end of the entire die-casting cycle, and record all parameter changes to optimize the initial settings for the next cycle.
[0005] In a preferred embodiment, after the die-casting machine is started, key parameters of the melt flow are collected in real time by sensors arranged in the flow channels and cavities of the multi-cavity mold to obtain initial flow state data, including: Before injection, thermal equilibrium is determined based on the temperature measurement point sequence, and the temperature difference threshold, fluctuation threshold and stable determination period are solidified into parameter versions. Pressure and temperature sensors are flush-mounted at the main channel inlet, branch nodes, and inlet of each cavity. High-temperature resistant sealing and limiting fixation are used, and the sensor link is calibrated and self-tested. Degradation compensation is performed for abnormal measuring points. During the injection phase, each measuring point is sampled synchronously under a unified acquisition clock and a timestamp is added. The sampling configuration is managed by version, and the start and stop of the acquisition window are determined by de-jittering based on the plunger status and the pressure threshold at the main channel inlet. After signal conditioning of the acquired signals, outlier removal, short-term jitter suppression and drift compensation are performed to generate a dataset with pressure attenuation characteristics, pressure difference characteristics and their changing trends, and with version labels.
[0006] In a preferred embodiment, transmitting the real-time acquired parameters to the control system includes: The data within the acquisition window is encapsulated into data packets and sent to the control system via a wired or wireless link. The data packets carry a unified timestamp sequence, a health status label for the measurement point, a threshold version label, and a sampling configuration version label. Perform field integrity checks and timing consistency checks on the received data packets. If the checks fail, the data packets are either discarded or retransmitted. In cases where padding is allowed, the missing intervals are padded according to the packet loss tolerance rules, and a padding marker is generated.
[0007] In a preferred embodiment, the model is compared with a preset balanced flow channel model to analyze the differences between parameters in order to identify potential areas of uneven melt flow and provide a basis for subsequent adjustments, including: Data of measuring points are filtered based on the health status labels of the measuring points. Data of abnormal measuring points are replaced by data of neighboring measuring points, or missing data are estimated based on the flow channel topology constraints. The source of replacement and estimation method are recorded. Establish a mapping relationship between the measurement points and the nodes or flow segments of the balanced flow channel model, and align the measurement point data to a unified time grid; Calculate pressure deviation, temperature deviation and their changing trends, and determine the limits based on the deviation threshold and the persistence judgment window; The system locates the flow path based on the flow channel topology and the propagation relationship of deviations between adjacent nodes, and outputs structured difference analysis data.
[0008] In a preferred embodiment, based on the difference analysis results, an injection speed adjustment command is generated, including: Read the version tag from the difference analysis results and generate an adjustment task based on the node or flow channel segment identifier. The adjustment task must have at least the fields of deviation amount, persistence index and confidence index. Mediation tasks are diverted according to confidence level rules. Low-confidence mediation tasks are subject to amplitude limiting and weight decay and are marked for review. The remaining mediation tasks are processed in a preset priority order.
[0009] In a preferred embodiment, the flow requirements between multiple cavities are optimized based on the influence of melt viscosity and mold resistance, including: The control system determines the viscosity correction factor based on the temperature deviation and the resistance correction factor based on the pressure drop deviation. The viscosity correction factor and the drag correction factor are fused according to the influence factor weights to obtain the comprehensive correction amount; Solve for the segmented velocity setpoint, segment switching condition, and velocity change slope parameter within the feasible region defined by the constraint set; Based on the target node or target flow channel segment identifier, the position segment association consistency is checked, and a structured injection speed adjustment command is generated. The command carries the command number, version identifier, target segment identifier and speed trajectory parameters, and sets the validity conditions.
[0010] In a preferred embodiment, sending an adjustment command to the injection mechanism includes: Command frames are sent via industrial field communication links or isolated input / output links. The command frames carry command number, sequence number, version field, check field, speed trajectory parameter set and switching condition flag; Pre-verification is performed based on the validity conditions of the instruction, and a handshake mechanism is used to complete the issuance confirmation; When a response timeout or verification failure occurs, the control system performs a retransmission based on the retransmission threshold configured in the process parameter table. When the injection mechanism returns a rejection code, the control system records the rejection code and triggers a re-evaluation or a switch to a safety strategy.
[0011] In a preferred embodiment, the injection mechanism dynamically adjusts the speed profile of the injection plunger according to instructions, and transitions from a slow phase to a fast phase, including: The instruction frame is checked for integrity and version consistency, and the instruction number and sequence number are checked for deduplication and order. When the check passes, the velocity trajectory parameters are written into the execution buffer, and the segmented velocity trajectory is shaped by limiting the slope and acceleration according to the constraint set. The stage switching is triggered based on displacement conditions, timing conditions, or pressure conditions, causing the speed curve to switch from a slow stage to a fast stage; Closed-loop tracking control is performed based on speed and displacement feedback, and tracking error is monitored by criteria. When the error criteria are met, the injection mechanism performs degraded control or activates the safety speed curve, and generates an execution log to be sent back to the control system.
[0012] In a preferred embodiment, during the adjustment process, changes in melt parameters continue to be fed back in real time via sensors. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment, including: The observation window is determined by the execution start and end time identifiers, the stage switch trigger record, and the instruction number; Within the observation window, data is collected synchronously according to the sampling configuration and standardized, and then written to the threshold version label, sampling configuration version label, and measurement point health status label. Based on the target equilibrium state, convergence determination, persistence determination, and execution consistency determination are performed respectively to obtain the corresponding determination results; Generate a residual input set and a verification flag, and encapsulate the judgment result into a confirmation report and write it into the feedback buffer.
[0013] In a preferred embodiment, the feedback and adjustment process is repeated until the end of the entire die-casting cycle, and the system records all parameter changes to optimize the initial settings for the next cycle, including: The control rounds are determined based on the confirmation report and residual set, and the data collection, difference analysis, solution and command issuance are repeated within each round; At the end of the die-casting cycle, the control system generates a cycle-level file and records the model version, threshold version, and sampling configuration version. The control system updates model parameters based on multi-period periodic archives and generates corresponding threshold version identifiers.
[0014] Compared with the prior art, the present invention provides a method for balanced flow channels and injection speed control in multi-cavity molds, which has the following beneficial effects: 1. This invention establishes a unified and reproducible acquisition and judgment caliber before and after injection. Through synchronous sampling and data standardization of pressure and temperature at multiple locations, and combined with a balanced flow channel model, it completes the mapping from measurement points to model nodes or flow channel segments and analyzes deviation trends, providing candidate regions of imbalance and their processing priorities. Based on this, deviations are categorized into viscosity correction and resistance correction factors according to their mechanisms and weighted and fused. Under the constraints of equipment capacity boundaries and forming risks, the segmented speed trajectory and switching conditions are solved, generating structured adjustment instructions carrying version information and validity conditions. The instruction issuance and execution sides employ integrity verification and handshake confirmation mechanisms, combined with deduplication sequence control, trajectory secondary shaping, and closed-loop tracking, and return execution logs for consistency verification. The control side completes convergence judgment, persistence judgment, and execution consistency judgment within the observation window, forming residuals and performing selective iteration. Simultaneously, based on periodic archives, cross-period version updates and parameter optimization are carried out, ultimately improving the balance stability of multi-cavity filling and reducing the probability of defects, ensuring batch production quality and cycle time consistency, thereby solving the problem of insufficient real-time injection speed adaptive control accuracy in traditional methods.
[0015] 2. This invention integrates mechanisms such as sensor health status identification, self-test degradation, data completion marking, command validity conditions, communication handshake and retransmission control, execution-side trajectory shaping, and anomaly code feedback throughout the entire process of data acquisition, solution, distribution, execution, and confirmation. Furthermore, it uses periodic archives to uniformly archive and associate model versions, threshold versions, and sampling configuration versions, enabling traceability and consistency verification of the basis, process records, and status information for each adjustment round. When situations such as measurement point drift, link jitter, unavailable execution resources, or abnormal equipment status occur, the system performs tiered handling based on verification markings and safety policies, and incorporates abnormal events and residual entries into subsequent iterations in a structured form, thereby enhancing the traceability, maintainability, and operational stability of the control process under complex operating conditions. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the process for a balanced flow channel and injection speed control method for a multi-cavity mold according to the present invention. Detailed Implementation
[0017] 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.
[0018] Example 1: Figure 1 A method for controlling the balanced flow channel and injection speed in a multi-cavity mold is presented, including: S1: After the die-casting machine is started, the key parameters of the melt flow are collected in real time by sensors arranged in the flow channels and cavities of the multi-cavity mold to obtain the initial flow state data. S2: Transmit the parameters collected in real time to the control system and compare them with the preset balanced flow channel model. Analyze the differences between parameters to identify potential areas of uneven melt flow and provide a basis for subsequent adjustments. S3: Based on the difference analysis results, generate injection speed adjustment commands and optimize the flow requirements between multiple cavities according to the influence of melt viscosity and mold resistance; S4: The adjustment command is sent to the injection mechanism, which dynamically adjusts the speed curve of the injection plunger according to the command and transitions from the slow stage to the fast stage to adapt to the real-time flow conditions. S5: During the adjustment process, the changes in melt parameters continue to be fed back in real time through sensors. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment. S6: Repeat the feedback and adjustment process until the end of the entire die-casting cycle, and record all parameter changes to optimize the initial settings for the next cycle.
[0019] S1: After the die-casting machine is started, sensors arranged in the flow channels and cavities of the multi-cavity mold are used to collect key parameters of the melt flow in real time to obtain initial flow state data. The specific implementation is as follows: In the multi-cavity die-casting production of aluminum alloy steering knuckles, after the die-casting machine is started and enters the operating state where injection control can be executed, the mold clamping and locking are completed first, so that the fixed half mold and the moving half mold are stably closed. Then, the mold temperature control unit is started to preheat the main flow channel, the branch flow channel and each cavity area to establish a repeatable initial working condition. In order to avoid the thermal state difference caused by relying solely on experience, the control system performs a consistency judgment on the mold thermal balance: First, temperature measuring points are set up in the area near the main flow channel inlet, the area near each branch flow node, the area near each cavity inlet and the area near the cavity end. The temperature sequence of each measuring point is collected within the judgment window, and the maximum temperature difference and fluctuation amplitude are calculated. When the maximum temperature difference does not exceed the temperature difference threshold, the fluctuation amplitude does not exceed the fluctuation threshold, and the above conditions are continuously met during the stable judgment period, the mold is judged to have reached the initial thermal balance state, so that the subsequent collected data mainly reflects the difference between melt flow and cavity feeding, rather than the disturbance introduced by local temperature difference. The temperature difference threshold, fluctuation threshold, and stability determination period are calibrated and fixed in the process parameter table during the trial molding stage. Temperature sequences are collected during multiple stabilization cycles, and the temperature difference threshold and fluctuation threshold are determined based on the statistical distribution of the maximum temperature difference and the fluctuation amplitude of the sliding window within the cycle. The stability determination period is determined by combining the mold thermal inertia and production cycle time. On this basis, a safety margin is introduced to form a traceable parameter version. The corresponding version is called during production to ensure the consistency and reproducibility of the determination. In addition to the thermal balance-related thresholds, the self-check thresholds and trigger thresholds involved in this method also follow a unified calibration rule: the noise boundary is determined based on the trial molding data or baseline data from the non-injection stage, and a safety margin is set on the noise boundary before being fixed as a parameter version. The sampling configuration is determined by the evaluation results of the characteristic change time scale and sensor response characteristics during the trial molding stage and is managed in association with the parameter version. After initial thermal equilibrium determination, a sensor monitoring network is constructed and real-time data acquisition begins. This network characterizes injection driving force, flow attenuation, and cavity feed differences. The network includes at least a pressure monitoring link and a temperature monitoring link. Pressure measurement points are located at the main channel inlet, each flow node, and each cavity inlet to acquire initial injection pressure, pressure attenuation characteristics at flow nodes, and pressure differences at each cavity inlet. Temperature measurement points are located corresponding to or adjacent to the pressure measurement points to acquire melt temperature changes at the same flow location, thus interpreting and correcting the thermal effects of pressure differences. The sensors employ an embedded mounting structure, with the sensor sensitive surface flush with the inner wall of the flow channel or gate, or installed according to a preset flush tolerance. High-temperature resistant sealing and limiting structures ensure stable measurement point positions and complete sealing, thereby reducing disturbance to melt flow and meeting the reliability requirements of long-term cyclic production. Before injection, the control system performs calibration and health self-check on the sensor link and generates health status labels for the measuring points; it performs zero-point calibration and range consistency check on the pressure sensor, and equivalent compensation and response consistency check on the temperature sensor; at the same time, the control system sets self-check criteria based on the baseline noise range and physical change boundaries to identify abnormal conditions such as zero-point drift, signal saturation, link disconnection, and short-term abnormal jumps; when a measuring point does not meet the self-check criteria, the control system marks the measuring point as faulty and performs degradation processing, which includes shielding the measuring point from participating in the difference judgment, using the estimated value of adjacent or similar measuring points for compensation, and writing the degradation status into the data label, thereby avoiding single-point abnormal data from entering the subsequent control decision process, while ensuring the traceability of the acquisition and processing process; After injection begins, the data acquisition unit synchronously samples each measuring point under a unified acquisition clock and adds a unified timestamp to the sampled data, ensuring that the data from different measuring points are on the same time reference. This allows for direct calculation of the differences between measuring points and the attenuation characteristics along the flow channel. The sampling resolution is determined by evaluation during the trial molding stage. The evaluation includes the time scale of changes in filling characteristics and the response characteristics of the sensors, and is set based on the ability to distinguish the sequential relationship between the flow splitting node and the cavity inlet pressure changes. If insufficient sampling resolution is detected during operation, resulting in aliasing of key features, the control system switches to a higher sampling resolution configuration and records the corresponding configuration version to ensure consistency and traceability in subsequent analyses. To ensure that the initial flow state data strictly corresponds to the early stage of filling, the control system truncates the acquisition window according to the injection sequence and employs a debouncing rule in the window start-stop determination to suppress misjudgments caused by instantaneous fluctuations. When the injection stage enters the filling state and the start conditions are met, the acquisition window is opened. The start conditions include the plunger displacement or plunger velocity entering a preset filling range, and the main channel inlet pressure showing a significant increase relative to the non-injection baseline. Meeting one or more of these conditions triggers the opening. When the injection stage switches to the holding pressure stage or the end conditions are met, the acquisition window is closed. The end conditions include stage flag switching, the plunger displacement entering a preset termination range, or the main channel inlet pressure exhibiting stable holding pressure characteristics. Meeting one or more of these conditions triggers the closing. The pressure trigger threshold is determined based on the pressure baseline noise range of the non-injection stage, taking the upper limit of the noise and adding a safety margin to balance trigger timeliness and noise resistance. The raw signals within the acquisition window are conditioned and converted from analog to digital before being input into the control system. The control system performs outlier removal, short-term jitter suppression, and drift compensation on the data according to unified rules to obtain a standardized sequence. Based on the standardized sequence, an initial flow state dataset is generated. The initial flow state dataset includes at least the instantaneous pressure, instantaneous temperature, and timestamp of each measuring point, as well as the pressure attenuation characteristics from the main channel inlet to the branch node, the pressure difference characteristics from the branch node to each cavity inlet, the pressure differences and their trends between each cavity inlet, and the temperature differences and their trends corresponding to the pressure differences, calculated from the standardized sequence. It also carries measuring point health status labels, threshold version labels, and sampling configuration version labels. Based on the data formed by thermal balance confirmation, synchronous acquisition, window truncation, and structured encapsulation, the initial flow state can be reproduced, aligned, and traced. It also serves as the input basis for comparison with a preset balanced flow channel model, identification of unbalanced locations, and triggering parameter adjustments.
[0020] S2: The parameters collected in real time are transmitted to the control system and compared with the preset balanced flow channel model. The differences between the parameters are analyzed to identify potential areas of uneven melt flow, providing a basis for subsequent adjustments. The specific implementation is as follows: Continuing with the initial flow state dataset generated earlier, the data acquisition unit sends the data packets generated within the acquisition window to the control system. These data packets include at least the instantaneous pressure, instantaneous temperature, and their unified timestamp sequence for each measuring point, as well as pressure decay characteristics, pressure difference characteristics, pressure difference trend characteristics, temperature difference characteristics, and their trend characteristics calculated based on the timestamp sequence. Furthermore, the data packets carry measuring point health status labels, threshold version labels, and sampling configuration version labels. Upon receiving the data packets, the control system merges and sorts them according to their timestamps and loads the analysis configuration corresponding to the version labels to complete subsequent alignment, comparison, and judgment under a unified time reference and parameter system. Data transmission can be achieved via wired or wireless links, and the transmission links have error control and secure access mechanisms. The control system performs integrity and timing consistency checks on the received data packets. Integrity checks are used to confirm that all fields are complete and consistent, while timing consistency checks are used to confirm that the timestamps are monotonically increasing and the sampling interval is consistent with the sampling configuration. If the check fails, the control system discards the abnormal data packets or triggers retransmission and records the reason for the abnormality. If a short-term missing data packet occurs and does not affect timing alignment, the control system determines whether to allow padding based on the packet loss tolerance rules preset in the threshold version. If padding is allowed, an interpolation strategy or a hold strategy consistent with the sampling configuration is used to complete the padding and mark the padding interval. If padding is prohibited, the missing interval is marked as unavailable and the confidence of this comparison is reduced. Before comparison, the control system filters and downgrades measurement points based on their health status labels. For measurement points marked as abnormal, the control system does not directly use their data for difference determination, but instead uses a substitution strategy to obtain equivalent input. The substitution strategy includes using redundant measurement points from neighboring measurement points in the same flow channel segment or from similar nodes, or estimating missing values based on the constraint relationships between adjacent nodes in the topology. The control system records the substitution source, estimation method, and the measurement point numbers involved in this analysis to ensure the traceability of the results. Subsequently, the control system aligns the data from each measuring point to the same time grid based on a unified timestamp. When there is a systematic link delay, it is corrected according to the pre-calibrated delay compensation parameters. After alignment, the control system performs model node mapping, establishing a correspondence between the measuring point number and the node number or flow channel segment number in the balanced flow channel model. The balanced flow channel model includes at least a multi-cavity flow channel topology, a set of key nodes, node connectivity, and the expected pressure and temperature relationships between nodes. The model parameters are determined by fusing simulation calculations with trial benchmark data. The fusing process includes identifying and updating the model resistance and heat loss parameters by minimizing errors using the trial benchmark data as the target, resulting in a model version for comparison. The control system calculates deviations layer by layer according to nodes and flow channel sections. In the pressure dimension, the control system calculates the deviation of the main flow channel inlet relative to the expected model relationship, the pressure drop deviation from the main flow channel inlet to the branch node, the pressure drop deviation from the branch node to each cavity inlet, and the relative pressure difference deviation and its trend deviation between the cavity inlets. In the temperature dimension, the control system calculates the temperature deviation and temperature difference trend corresponding to the above nodes, and uses the temperature deviation as an auxiliary factor in interpreting the pressure deviation to determine whether the equivalent resistance change is mainly caused by thermal or non-thermal factors. The deviation threshold is determined during the trial molding stage at full capacity. The data is obtained statistically from stable cycles under sufficient thermal equilibrium conditions. Based on this, the control system performs over-limit judgment on various deviations. To suppress false triggering caused by instantaneous noise, the control system introduces a persistent judgment mechanism. A sliding judgment window is constructed based on the sampling configuration, and the over-limit results are confirmed through continuous over-limit rules and cumulative over-limit rules. When the confirmation conditions are met, the control system marks the corresponding node or flow channel segment as an unbalanced candidate anomaly and generates a persistent index and a confidence index. The confidence index comprehensively considers the over-limit magnitude, persistent performance, measurement point degradation status, and completion mark, and is used to rank the candidate anomalies. During the differential positioning stage, the control system performs inversion positioning and risk classification of candidate anomalies based on the flow channel topology and the propagation characteristics of deviations between adjacent nodes: when the deviation shows a sudden increase from small to large between adjacent nodes, the anomaly is located at the corresponding flow channel segment or gate segment; when multiple cavity inlet deviations are in the same direction and the flow splitting nodes are synchronously abnormal, the anomaly is located at the flow splitting node and its adjacent flow splitting segment and is judged as uneven flow distribution; when a single cavity inlet deviation is significant and accompanied by synchronous abnormal temperature deviation, the anomaly is located at the cavity inlet and its adjacent flow channel segment and marked as thermal factor influence; when a single cavity inlet deviation is significant and the temperature deviation is not significant, the anomaly is judged as a local resistance anomaly and marked as non-thermal factor influence; the control system combines the deviation magnitude, persistence index, confidence index and forming sensitive area priority rules to output the risk classification results and the corresponding processing priority sequence; Finally, the control system outputs the structured difference analysis results, which include at least the cavity number, model node or flow channel segment identifier, deviation type, deviation amount, persistence index, confidence index, influence factor label, as well as the measurement point degradation status label and completion label. The results are used as the input basis for the subsequent generation of injection speed adjustment commands.
[0021] S3: Based on the difference analysis results, generate injection speed adjustment commands, and optimize the flow requirements between multiple cavities according to the influence of melt viscosity and mold resistance. Specifically, the implementation is as follows: Following the difference analysis results output previously, the control system reads the difference analysis result table and its corresponding model version, threshold version, and sampling configuration version, and generates a list of adjustment tasks based on the adjustment-related fields. These fields include at least the deviation, persistence index, confidence index, impact factor marker, and the model node or flow channel segment identifier obtained from the previous location. The control system then uses the confidence level grading rules in the threshold version to filter and allocate adjustment tasks: when the confidence level of an adjustment task is lower than the adjustable level, or when the reliability of the corresponding data is reduced due to degradation or padding, the control system assigns the task to the conservative adjustment path and initiates... A limiter is used to restrict its contribution to speed correction, and a verification mark is written for re-evaluation in subsequent cycles. When the confidence level of the adjustment task meets the adjustable level and the persistence index indicates that the deviation is a stable offset, the control system classifies the task into the regular adjustment path and processes it in the order of priority output above, so that the adjustment is given priority to the tasks related to the split nodes, key flow channels and cavity inlets that have a more significant impact on multi-cavity balance. The limiting boundary of the limiter and the conservative adjustment triggering condition are preset by the threshold version, and their values are determined by the deviation stability statistics and link missing impact assessment during the trial modeling stage to ensure that the conservative strategy has a reproducible triggering and execution basis. After the task is distributed, the control system performs a mechanistic interpretation of each adjustment task and constructs an intermediate quantity for solving the velocity trajectory. The control system uses the previous judgment results on thermal factors and non-thermal factors to map the adjustment tasks to the melt viscosity correction path and the mold resistance correction path respectively, and merges the two types of correction results according to the influence factor weight to obtain the comprehensive resistance correction amount. The melt viscosity correction factor takes the temperature deviation and temperature difference trend corresponding to the adjustment task as input. The control system calls the material rheological mapping rule in the process parameter table to convert the temperature state change into an equivalent characterization quantity of the flowability change. The rheological mapping rule is used to describe the monotonic relationship and range sensitivity of temperature change to equivalent viscosity or equivalent flow resistance, and can be modified in combination with the control mode of the filling stage to characterize the response difference of the same material in different filling stages. In order to suppress transient noise, the control system constructs a smoothing window according to the sampling period determined by the sampling configuration version and performs smoothing processing on the temperature deviation sequence. When the persistence index indicates that the deviation is unstable or the verification flag has been triggered, the control system limits the gain of the viscosity correction factor and increases the smoothing intensity accordingly to avoid high-frequency fluctuations in the speed command between adjacent cycles. The mold resistance correction factor takes the pressure drop deviation, cavity inlet pressure difference deviation, and positioning results mentioned above as inputs for the adjustment task. The control system maps the deviation to the target flow channel segment or gate segment corresponding to the positioning based on the node connectivity of the balanced flow channel model. According to the mapping rule between the expected pressure drop relationship of the model and the measured deviation, the deviation amplitude is converted into a representation of the equivalent resistance change. When the adjustment task is in a conservative adjustment path, the control system applies a weight attenuation to the resistance correction factor and enables a limiter to limit its contribution to the overall resistance correction. The above-mentioned smoothing window parameters, gain limit parameters, weight attenuation parameters, and limiter boundaries are all provided by the process parameter table and associated with the version configuration. After obtaining the comprehensive resistance correction, the control system enters the solution stage of the injection speed adjustment command. First, the speed trajectory parameters are determined by a solution method that unifies objectives and constraints to avoid local over-adjustment or the introduction of new imbalances caused by relying solely on fixed speed increase and decrease rules. The control system aims to reduce the pressure difference deviation at the cavity inlet, reduce the pressure drop deviation at the flow splitting node, and reduce the persistence of deviations. At the same time, it calls the constraint set in the process parameter table to construct the feasible region. The constraint set includes at least equipment capacity boundary constraints and forming quality risk constraints. The equipment capacity boundary constraints include speed boundaries, acceleration boundaries, and speed change slope boundaries. The forming quality risk constraints include at least constraints to prevent flash, air entrapment, scouring, cold shut, and insufficient filling. The control system selects the main adjustment stage based on the target position segment identifier of the adjustment task and maps the comprehensive resistance correction amount to the speed trajectory parameter correction amount. The speed trajectory parameter includes at least the segmented speed setpoint, segment switching conditions, and speed change slope parameters. The segment switching conditions can be defined by displacement triggering, time triggering, or pressure state triggering, and their triggering method and priority are configured by the process parameter table. To ensure the feasibility of the solution, the control system uses a combination of rule mapping and iterative correction to generate the final correction: first, the correction direction and action stage are determined based on the deviation type and influencing factors; then, the speed setpoint and slope parameters are limited and saturated within the feasible region; subsequently, the local correction amounts of each adjustment task are superimposed according to priority. When superposition causes constraint conflicts or contradictions in correction directions, high-priority corrections are retained based on priority adjudication, while low-priority corrections are reduced or masked. The adjudication results are recorded for traceability and verification. For tasks in conservative adjustment paths, the control system only outputs correction amounts with limited amplitude and ensures that forming quality risk constraints are met. To match the speed command with the potential imbalance region, the control system executes a position segment association strategy based on the positioning results mentioned above, performing consistency checks and re-constraints on the solved speed trajectory. If the positioning result points to a flow branch node and its adjacent flow branch segments, the global speed trajectory is corrected using the pressure drop deviation of the flow branch node as the main control index, making the feed distribution relationship of multiple cavities tend to be balanced. If the positioning result points to a single cavity inlet or its adjacent flow channel segment, the pressure difference deviation at the cavity inlet is used as the main control index to correct the speed establishment rate and switching conditions of the corresponding stage, making the pressure difference between cavities converge. Furthermore, the control system selects a correction strategy based on the influencing factor label: when thermal factors dominate, the slope of speed change is limited and the correction amplitude is constrained by the viscosity correction factor; when non-thermal factors dominate, the equivalent resistance difference is improved by driving force distribution and the temperature-related indicators are kept within an acceptable range. The position segment association strategy is executed according to the strategy parameters in the process parameter table to ensure reproducible implementation. After completing the solution, the control system generates an injection speed adjustment command and writes it into the command output buffer for subsequent steps. The injection speed adjustment command adopts a structured data format, including at least the command number, applicable model version and process parameter table version, applicable sampling configuration version, target node or target flow channel segment identifier, speed trajectory parameter set, switching condition set, constraint set reference identifier, confidence level marker, amplitude limiting and weight attenuation marker, verification marker, and priority decision record field. The control system sets validity conditions for the command, which use the operating condition judgment results and measurement point availability conditions mentioned above. If the validity conditions are not met in the next cycle, the control system does not use the command and re-triggers the difference analysis and solution process to ensure that the speed strategy is consistent with the current operating condition. Based on the task diversion, factor construction, constraint solving, position segment association, and structured encapsulation of the difference analysis results, the control system generates an executable and traceable injection speed adjustment command and uses it as input for subsequent command sending and execution.
[0022] S4: The adjustment command is sent to the injection mechanism, which dynamically adjusts the injection plunger speed curve according to the command and transitions from a slow phase to a fast phase to adapt to real-time flow conditions. Specifically, this is implemented as follows: Following the injection speed adjustment command generated and output to the command output buffer in the previous section, the control system issues the command to the injection mechanism through an industrial field communication link or an isolated I / O link. The link must have at least error detection, retransmission control, acknowledgment, and anti-interference transmission capabilities. The control system encapsulates the command using a structured sending frame. The sending frame includes a command number, a sequence number, a version field, and a verification field. The sending frame also carries a set of speed trajectory parameters, a set of switching conditions, a constraint set reference identifier, and amplitude limiting, weight attenuation, and verification markers. The command number and sequence number are used for deduplication and sequential control, the version field is used for consistency verification, and the verification field is used for integrity verification. Before issuing an instruction, the control system performs a pre-verification based on the validity conditions attached to the adjustment instruction mentioned above to determine the executability of the instruction in the current loop. If the validity conditions are not met, the control system does not issue the instruction and triggers a re-analysis and solution process. The control system uses a handshake mechanism when issuing the instruction, and waits for the injection mechanism to return a reception confirmation and parsing status after sending the issuance frame. The response timeout parameter, the upper limit of retransmission count, and the deduplication window are configured as communication thresholds by the process parameter table. The control system triggers retransmission when a response timeout occurs, the receiver returns a verification failure, or a frame duplication conflict is detected. If a valid confirmation is not obtained after exceeding the upper limit of retransmission count, the control system determines that the communication is abnormal and enters the safety control path, while recording the communication abnormality code, instruction number, and retransmission record. For the rejection status returned by the injection mechanism, the control system classifies and processes it according to the rejection code: when the versions are inconsistent, execution is prohibited and the reason for the version mismatch is recorded; when the constraint is infeasible, the issuance is terminated and a re-solution is triggered; when the execution resource is unavailable, the system switches to a safe speed strategy and waits for the equipment to recover or for manual handling. After receiving the transmitted frame, the injection mechanism performs frame-level verification and consistency verification. Frame-level verification verifies the consistency of the verification fields and eliminates transmission errors. Consistency verification checks whether the version field is consistent with the version loaded on the machine, checks whether the instruction number and sequence number meet the deduplication and order requirements, verifies whether the constraint set reference identifier can be parsed, and determines whether the validity conditions attached to the adjustment instruction mentioned above are consistent with the current device status. If the verification fails, the injection mechanism refuses to execute and sends back a rejection code and reason. Constraint infeasibility means that after applying the limiting and shaping rules of the constraint set, it is still impossible to form an executable trajectory that simultaneously satisfies the device capability boundary and the switching condition set. Execution resource unavailable means that the execution element self-test fails or the key sensor is in an unavailable state. If the verification passes, the injection mechanism writes the speed trajectory parameter set and the switching condition set into the execution buffer and writes the limiting, weight attenuation, and verification flags into the current execution context for subsequent execution and log recording. The injection mechanism generates a target velocity curve for the injection plunger based on a set of velocity trajectory parameters, and performs secondary shaping on the target velocity curve according to a set of constraints. This secondary shaping includes at least slope limiting and acceleration limiting. The injection mechanism has a built-in trajectory generator that raises or lowers the segmented velocity setpoints according to slope boundaries and saturates and limits velocity changes according to acceleration boundaries, thereby converting discrete segmented parameters into a continuously executable velocity curve to avoid risks of melt impact, air entrapment, scouring, or flash caused by sudden velocity changes. The velocity curve includes at least a slow stage and a fast stage. The slow stage is used to establish stable feeding and assist in early venting and flow stability, while the fast stage is used to meet filling efficiency requirements and suppress the risk of early solidification. The transition from the high-speed phase to the fast-speed phase is triggered by a set of switching conditions, which includes one or a combination of conditions triggered by plunger displacement, stage timing, or pressure state. During execution, the injection mechanism calculates in real time whether the triggering conditions are met, and when they are met, the trajectory generator smoothly switches to the next stage under continuous constraints. If the adjustment command mentioned above carries a limiting flag or a weight attenuation flag, the injection mechanism performs limiting processing on the speed setpoint and slope parameters of the corresponding stage during trajectory generation and shaping, and records the limiting effect status, the limited parameters, and the limiting reason. The verification flag is used to indicate that the command is in a conservative execution path. After execution, the injection mechanism returns the verification flag and the corresponding limiting record together. During the speed curve execution, the injection mechanism obtains the actual execution state through the plunger displacement detection unit and the speed feedback unit, and performs closed-loop tracking control; the controller outputs the execution control quantity with the target speed curve as a reference, so that the plunger speed and plunger displacement change according to the target trajectory; the injection mechanism monitors the tracking error and abnormal state online, and the tracking error threshold and error persistence judgment window are configured by the process parameter table as execution criteria; when the tracking error meets the continuous over-limit condition, or an abnormal response of the actuator is detected, or an abnormal equipment status parameter is detected, the injection mechanism triggers the abnormal handling strategy; among which the abnormal The standard handling strategy includes at least degrading execution, switching to a safe speed curve, or triggering an interlock shutdown. The exception code, triggering condition, trajectory stage at the time of the exception, switching condition status, and execution context are written to the execution log. Simultaneously, the exception status is sent back to the control system to trigger a re-solution or manual intervention. The safe speed curve is a pre-stored conservative trajectory or a limited version of the previous stable cycle trajectory. Its selection rules and activation conditions are configured by the process parameter table. When an exception is triggered, the injection mechanism activates the safe speed curve according to the selection rules and records the activation reason to ensure that the process remains controllable under abnormal conditions. After the steps are completed, the injection mechanism sends a structured execution log back to the control system. The execution log includes at least the instruction number and version field, execution start and end time identifier, target speed curve summary, actual speed curve and plunger displacement curve records, switching condition trigger records, amplitude limiting and weight attenuation effect records, tracking error statistics results, communication and execution exception codes and safety curve activation records.
[0023] S5: During the adjustment process, the changes in melt parameters continue to be fed back in real time through sensors. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment. Specifically, the implementation is as follows: After the control system executes the adjustment command and sends back the execution log, it continuously receives melt parameters fed back from the sensor network within the observation window where the command takes effect. It then compares and analyzes the feedback data with the target equilibrium state and outputs the convergence determination, persistence determination, execution consistency determination, and subsequent control decisions for this adjustment. The observation window is determined by the execution start and end time identifier, the stage switching trigger record, and the command number, and is used to align the feedback data with the solution record and actual execution trajectory of this command. Within the observation window, the sensor network synchronously acquires data from key measuring points according to the sampling period specified in the sampling configuration version. The acquired raw signals are then conditioned, converted from analog to digital, and timestamped before being transmitted back to the control system. The control system processes the feedback data using established data standardization rules to form a comparable standardized sequence and records threshold version labels, sampling configuration version labels, and measuring point health status labels to ensure consistent judgment and traceability. The standardized sequence includes at least the instantaneous pressure, instantaneous temperature, and corresponding timestamp of each measuring point. Furthermore, it extracts the pressure drop characteristics from the main channel inlet to the branch node, the pressure difference characteristics from the branch node to each cavity inlet, the pressure differences and their trends between each cavity inlet, and the temperature differences and their trends corresponding to the pressure differences. The control system also writes an instruction number and version field into the data for this step to establish a correlation with the execution record and solution record. The control system compares and analyzes the standardized sequence based on the target equilibrium state and threshold set defined in the threshold version. The comparison and analysis includes at least convergence determination, persistence determination, and execution consistency determination. When performing convergence determination, the control system selects the pressure difference index and pressure drop index corresponding to key nodes or key flow channels, calculates the deviation of each index relative to the target equilibrium state, and compares the deviation with the deviation threshold set. When all key deviations meet the deviation threshold set, the adjustment is determined to have reached equilibrium convergence; otherwise, residual entries are generated and summarized to form a residual input set. The residual entries include at least the node or flow channel segment identifier corresponding to the residual, the residual deviation, the residual trend characteristics, and the confidence level. The control system aggregates the residuals of the same type of deviation or adjacent position segments according to the residual merging rules to obtain the aggregated residuals for the next round of solution, so as to avoid residual fragmentation affecting the determination of subsequent adjustment strategies. During persistence determination, the control system statistically analyzes the trend of deviation changes within the persistence determination window. The persistence determination window is jointly determined by the sampling period defined by the sampling configuration version and the stage switching trigger record, and selects the stable interval of the instruction effective stage to avoid the transient interval caused by stage switching. The control system determines whether the deviation is in a stable convergence, oscillation, or rebound state based on the persistence threshold configured in the threshold version. When the persistence requirement is not met, the control system adds a persistence mark to the corresponding residual entry and retains the trend characteristics for adaptive adjustment of the adjustment amplitude and adjustment stage during subsequent strategy updates or re-solution. During consistency determination, the control system reads the actual speed curve summary, switching condition trigger record, tracking error statistics, anomaly code, and safety curve activation record from the execution side, and aligns the information with the key points of the target trajectory corresponding to the current instruction. When the execution log shows execution anomaly markers such as communication anomaly, tracking error meeting continuous over-limit conditions, triggering anomaly rollback, or activating safety curve, the control system downgrades the comparison result. The downgrade process includes at least reducing the weight of the feedback data in the next round of solution, freezing the priority of residual entries highly correlated with execution anomalies, and prioritizing link and equipment status verification. After the execution anomaly is resolved, the normal solution path is restored to avoid misjudging the execution anomaly as a process deviation. After the judgment is completed, the control system further performs correlation consistency verification to confirm the consistency between the feedback results and the influence factor labels. The control system performs correlation analysis on the improvement trend of pressure difference and the change trend of temperature difference, and determines whether the improvement occurs in the target location segment in combination with the positioning results. If the influence factor label indicates that thermal factors are dominant, it verifies whether the temperature difference trend and the convergence of pressure difference have a consistent improvement relationship, and checks whether the temperature-related indicators meet the thermal stability threshold defined by the threshold version. If the influence factor label indicates that non-thermal factors are dominant, it verifies whether the pressure drop characteristics and the improvement of pressure difference are concentrated in the positioning flow channel segment or gate segment, and checks whether the temperature difference is within the acceptable range defined by the threshold version. When there are contradictory situations such as the pressure difference has converged but the pressure drop characteristics of the target location segment have not improved, the temperature difference trend is inconsistent with the influence factor label, or the residual entries jump across segments between consecutive cycles, the control system generates consistency anomaly entries and sets a review mark. The consistency anomaly entries include at least the anomaly type, a summary of correlation evidence, and the identifier of the location segment involved, and are entered into subsequent iteration processing together with the residual input set. The control system encapsulates the output of this step into a structured verification report and writes it into the feedback buffer. The verification report includes at least the instruction number, model version, threshold version, sampling configuration version, observation window identifier, key node or flow segment identifier, deviation and its trend characteristics, convergence determination result, persistence determination result, execution consistency determination result, verification flag, anomaly code reference, and residual input set. The control system generates subsequent control decisions based on the verification report: when both convergence and persistence determinations pass and there are no execution anomalies, the current control strategy is maintained; when there is a residual input set or the persistence determination fails, the residual input set is used for the next round of solution and parameter update; when there are execution anomalies or consistency anomalies, the system enters the verification path and triggers a re-solution or execution degradation strategy after verification is completed.
[0024] S6: Repeat the feedback and adjustment process until the entire die-casting cycle ends, and record all parameter changes to optimize the initial settings for the next cycle. Specifically, this is implemented as follows: In the multi-cavity die casting process, after the control system confirms the results of the current round of adjustments and generates a confirmation report, it decides whether to continue the feedback and adjustment process in the remaining stage of the cycle based on the judgment results of the confirmation report and the residual input set. At the end of the cycle, the confirmation report, execution log and residual input set are summarized into a cycle-level file for optimization of the initial process parameters, model parameters and threshold settings of subsequent die casting cycles. Within a single die-casting cycle, the control system reads the convergence judgment result, persistence judgment result, execution consistency judgment result, residual input set, and verification mark from the latest confirmation report. If the confirmation report indicates that the deviation of the critical position segment meets the deviation threshold set, the persistence judgment passes, and the execution consistency judgment has no abnormal impact, and the residual input set is empty or only contains residual items marked as low priority and not required to be processed in the current cycle, the control system marks the current stage as the adjustment completed state. No new adjustment instructions will be issued in subsequent observation windows; only the acquisition and recording of monitoring data will be maintained. When the confirmation report indicates that further adjustment is needed and the current die-casting cycle has not yet met the end conditions, the control system restarts the feedback and adjustment process in subsequent observation windows. The need for further adjustment includes at least one of the following: residual items that need to be processed first, persistence judgment that has not passed, execution consistency in a downgraded state, or items with consistency anomalies. This will cause the current die-casting cycle to form several rounds of acquisition, difference analysis, instruction generation, execution, and confirmation in subsequent stages. To facilitate the selective repetition of the above cycle within a single die-casting cycle, the control system divides the die-casting cycle into sub-stages based on the time marker of the observation window, the plunger displacement range, and the process stage marker, and configures one or more corresponding observation windows for each sub-stage. For sub-stages determined to require further adjustment, the control system repeats the acquisition and analysis process within its corresponding observation window and generates a new confirmation report after the adjustment command is executed. The control system associates the confirmation reports with existing confirmation reports according to the cycle marker, forming a confirmation report sequence corresponding to the die-casting cycle. For sub-stages determined not to require further adjustment, the control system stops generating new adjustment commands within the corresponding observation window and only continues to record melt parameters and execution status. By selectively repeating the process by sub-stage, the adjustment behavior is concentrated on stages with significant deviations or insufficient stability, avoiding unnecessary intervention in stages that have reached equilibrium. During the die-casting cycle, the control system records data and events hierarchically. During each round of adjustment, the control system continuously generates execution logs and confirmation reports associated with the instruction number. Simultaneously, it associates the instruction number, observation window identifier, key node or flow channel segment identifier, deviation amount and its trend characteristics, anomaly code reference, verification mark, and residual input set corresponding to the adjustment round with the cycle identifier of the current die-casting cycle, forming an intermediate record set for this cycle. The intermediate record set is used to support the sorting and statistics after the cycle ends, so that the generated cycle-level archive can reflect the adjustment behavior and deviation evolution process of each observation window within the cycle. The determination of the end of the die casting cycle is jointly made by the process status signal and the stability determination result of the melt parameters. The control system reads the equipment and process status signals. When the signal indicates that the current injection and holding pressure process is completed and the mold enters the mold opening, cooling or other termination-related stages, the control system performs stability determination on the melt pressure, melt temperature and their derived indicators in the corresponding observation window, and compares the determination result with the termination threshold and stability threshold to determine whether the current die casting cycle has ended. The termination threshold and stability threshold are pre-configured by the threshold version and process parameter table, and the corresponding threshold version label is recorded in the cycle-level file to ensure traceability between the cycle end conditions and the threshold settings. When the process status signal meets the termination stage conditions and the stability determination is passed, the control system marks the current die casting cycle as ended, stops triggering new adjustment cycles, and completes the sorting and archiving of the records for this cycle. At the end of the die-casting cycle, the control system generates a corresponding cycle-level file in the built-in storage module. The cycle-level file includes at least the cycle identifier, model version, threshold version, and sampling configuration version, as well as the set of confirmation reports, execution logs, and residual inputs generated within the cycle. It also records statistical information on communication and execution anomalies, deviation statistical characteristics of key nodes or flow channel sections, and process parameter adjustment events. The cycle-level file is stored in structured data format and supports querying and statistics by cycle, location segment, anomaly type, and version dimension. It is used to identify long-term deviation patterns and equipment status change trends, and to evaluate the process performance corresponding to different version parameter configurations. After generating the cycle-level profile, the control system or its connected parameter optimization module performs offline or online analysis on the cycle-level profile across multiple die-casting cycles to optimize the initial settings for subsequent die-casting cycles. When the profile statistics show that a certain cavity inlet or a certain flow channel section repeatedly exhibits deviation entries within multiple cycles, and the influence factor labels, convergence judgment results, and execution consistency judgment results corresponding to the deviation entries indicate that the deviation mainly originates from process factors or structural factors rather than occasional execution anomalies, the control system updates the parameters of the position segment in the balanced flow channel model and simultaneously updates the deviation threshold set or adjusts the weights of relevant factors in the solution and command generation process. The system is reconfigured to match model parameters and threshold settings with actual flow characteristics. When the cycle-level file shows that there is insufficient thermal stability, temperature gradient deviation from the target range, or abnormal solidification-related indicators in a certain process stage, the control system adjusts the mold preheating parameters, melt initial temperature setting, or stage parameters in the initial trajectory of injection speed according to the statistical results, so that subsequent die-casting cycles can compensate for systematic deviations at the starting conditions. When the cycle-level file shows that abnormal codes related to equipment status occur frequently under specific operating conditions, the control system improves the operational stability of subsequent cycles from both the process parameters and equipment status by updating safety thresholds, optimizing adjustment strategies, or triggering maintenance prompts. The control system employs a version-based management approach for adjusting model parameters, deviation threshold sets, thermal stability thresholds, and safety thresholds. Each time a parameter or threshold update is completed, a corresponding model version identifier and threshold version identifier are generated, and this version information is recorded in the cycle-level archive for subsequent cycles. This ensures that the evolution of parameters and thresholds is traceable in both time and operating condition dimensions. Within a single die-casting cycle, the control system performs selective, repetitive adjustments based on confirmation reports and residual input sets. At the cross-cycle level, it iterates initial parameters and models based on cycle-level archives. This maintains balanced melt distribution and stable operation in the multi-cavity die-casting process across cycles, and gradually optimizes initial settings and model parameters during continuous production, improving adaptability to process disturbances and equipment status changes.
[0025] In this embodiment, the die-casting machine first enters a preheating stage after startup, allowing the mold to reach a stable operating temperature range. Then, pressure sensors and thermocouples are placed at key nodes in the flow channel and at the inlets of each cavity to synchronously collect data on the pressure distribution, temperature changes, and gradient characteristics during the initial melt injection, forming initial flow state data. The control system receives this data and compares it with a preset balanced flow channel model, calculating the deviation and trend characteristics between each node and cavity, locating potential unbalanced flow positions, and outputting the difference analysis results. Based on the difference analysis results, the control system combines the melt viscosity model and the mold resistance calculation module to solve for constraints, generating an injection speed adjustment command, which is then sent to the injection mechanism. The injection mechanism adjusts the injection speed according to the command. The plunger's speed trajectory is adjusted in stages to smoothly transition the speed curve from the low-speed venting and stable filling stages to the rapid filling stage, matching the flow requirements under the current operating conditions. During execution, the sensor network continuously feeds back updated melt parameters. The control system compares the feedback data with the target equilibrium state within the observation window to determine convergence, sustainability, and execution consistency, and decides whether to continue fine-tuning or maintain the current strategy accordingly. After the die-casting cycle ends, the system structurally archives the confirmation reports, execution logs, and parameter curves of this cycle, forming a cycle-level archive. Based on the cycle archive, the initial settings and model parameters for the next cycle are updated in a versioned manner, thereby achieving closed-loop control and cross-cycle adaptive optimization in continuous production.
[0026] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.
[0027] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.
[0028] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.
[0029] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.
[0030] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0031] In addition, the functional modules in the embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.
[0032] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0033] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
[0034] 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 for controlling the balanced flow channel and injection speed of a multi-cavity mold, characterized in that, include: S1: After the die-casting machine is started, the key parameters of the melt flow are collected in real time by sensors arranged in the flow channels and cavities of the multi-cavity mold to obtain the initial flow state data. S2: Transmit the parameters collected in real time to the control system and compare them with the preset balanced flow channel model. Analyze the differences between parameters to identify potential areas of uneven melt flow and provide a basis for subsequent adjustments. S3: Based on the difference analysis results, generate injection speed adjustment commands and optimize the flow requirements between multiple cavities according to the influence of melt viscosity and mold resistance; S4: The adjustment command is sent to the injection mechanism, which dynamically adjusts the speed curve of the injection plunger according to the command and transitions from the slow stage to the fast stage to adapt to the real-time flow conditions. S5: During the adjustment process, the changes in melt parameters continue to be fed back in real time through sensors. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment. S6: Repeat the feedback and adjustment process until the end of the entire die-casting cycle, and record all parameter changes to optimize the initial settings for the next cycle.
2. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, After the die-casting machine is started, sensors arranged in the flow channels and cavities of the multi-cavity mold collect key parameters of the melt flow in real time to obtain initial flow state data, including: Before injection, thermal equilibrium is determined based on the temperature measurement point sequence, and the temperature difference threshold, fluctuation threshold and stable determination period are solidified into parameter versions. Pressure and temperature sensors are flush-mounted at the main channel inlet, branch nodes, and inlet of each cavity. High-temperature resistant sealing and limiting fixation are used, and the sensor link is calibrated and self-tested. Degradation compensation is performed for abnormal measuring points. During the injection phase, each measuring point is sampled synchronously under a unified acquisition clock and a timestamp is added. The sampling configuration is managed by version, and the start and stop of the acquisition window are determined by de-jittering based on the plunger status and the pressure threshold at the main channel inlet. After signal conditioning of the acquired signals, outlier removal, short-term jitter suppression and drift compensation are performed to generate a dataset with pressure attenuation characteristics, pressure difference characteristics and their changing trends, and with version labels.
3. The method for balanced flow channels and injection speed control of a multi-cavity mold according to claim 1, characterized in that, The parameters acquired in real time are transmitted to the control system, including: The data within the acquisition window is encapsulated into data packets and sent to the control system via a wired or wireless link. The data packets carry a unified timestamp sequence, a health status label for the measurement point, a threshold version label, and a sampling configuration version label. Perform field integrity checks and timing consistency checks on the received data packets. If the checks fail, the data packets are either discarded or retransmitted. In cases where padding is allowed, the missing intervals are padded according to the packet loss tolerance rules, and a padding marker is generated.
4. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, By comparing the model with a pre-defined balanced flow channel model, the differences between parameters are analyzed to identify potential areas of uneven melt flow, providing a basis for subsequent adjustments, including: Data of measuring points are filtered based on the health status labels of the measuring points. Data of abnormal measuring points are replaced by data of neighboring measuring points, or missing data are estimated based on the flow channel topology constraints. The source of replacement and estimation method are recorded. Establish a mapping relationship between the measurement points and the nodes or flow segments of the balanced flow channel model, and align the measurement point data to a unified time grid; Calculate pressure deviation, temperature deviation and their changing trends, and determine the limits based on the deviation threshold and the persistence judgment window; The system locates the flow path based on the flow channel topology and the propagation relationship of deviations between adjacent nodes, and outputs structured difference analysis data.
5. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, Based on the difference analysis results, injection speed adjustment commands are generated, including: Read the version tag from the difference analysis results and generate an adjustment task based on the node or flow channel segment identifier. The adjustment task must have at least the fields of deviation amount, persistence index and confidence index. Mediation tasks are diverted according to confidence level rules. Low-confidence mediation tasks are subject to amplitude limiting and weight decay and are marked for review. The remaining mediation tasks are processed in a preset priority order.
6. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, Based on the influence of melt viscosity and mold resistance, optimize the flow requirements between multiple cavities, including: The control system determines the viscosity correction factor based on the temperature deviation and the resistance correction factor based on the pressure drop deviation. The viscosity correction factor and the drag correction factor are fused according to the influence factor weights to obtain the comprehensive correction amount; Solve for the segmented velocity setpoint, segment switching condition, and velocity change slope parameter within the feasible region defined by the constraint set; Based on the target node or target flow channel segment identifier, the position segment association consistency is checked, and a structured injection speed adjustment command is generated. The command carries the command number, version identifier, target segment identifier and speed trajectory parameters, and sets the validity conditions.
7. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, The adjustment command is sent to the injection mechanism, including: Command frames are sent via industrial field communication links or isolated input / output links. The command frames carry command number, sequence number, version field, check field, speed trajectory parameter set and switching condition flag; Pre-verification is performed based on the validity conditions of the instruction, and a handshake mechanism is used to complete the issuance confirmation; When a response timeout or verification failure occurs, the control system performs a retransmission based on the retransmission threshold configured in the process parameter table. When the injection mechanism returns a rejection code, the control system records the rejection code and triggers a re-evaluation or a switch to a safety strategy.
8. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, The injection mechanism dynamically adjusts the injection plunger speed profile according to instructions, transitioning from a slow phase to a fast phase, including: The instruction frame is checked for integrity and version consistency, and the instruction number and sequence number are checked for deduplication and order. When the check passes, the velocity trajectory parameters are written into the execution buffer, and the segmented velocity trajectory is shaped by limiting the slope and acceleration according to the constraint set. The stage switching is triggered based on displacement conditions, timing conditions, or pressure conditions, causing the speed curve to switch from a slow stage to a fast stage; Closed-loop tracking control is performed based on speed and displacement feedback, and tracking error is monitored by criteria. When the error criteria are met, the injection mechanism performs degraded control or activates the safety speed curve, and generates an execution log to be sent back to the control system.
9. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, During the adjustment process, sensors continue to provide real-time feedback on melt parameter changes. The control system compares the feedback data with the target equilibrium state to confirm the continuity and accuracy of the adjustment, including: The observation window is determined by the execution start and end time identifiers, the stage switch trigger record, and the instruction number; Within the observation window, data is collected synchronously according to the sampling configuration and standardized, and then written to the threshold version label, sampling configuration version label, and measurement point health status label. Based on the target equilibrium state, convergence determination, persistence determination, and execution consistency determination are performed respectively to obtain the corresponding determination results; Generate a residual input set and a verification flag, and encapsulate the judgment result into a confirmation report and write it into the feedback buffer.
10. The method for controlling the balanced flow channel and injection speed of a multi-cavity mold according to claim 1, characterized in that, The feedback and adjustment process is repeated until the entire die-casting cycle ends. The system records all parameter changes to optimize initial settings for the next cycle, including: The control rounds are determined based on the confirmation report and residual set, and the data collection, difference analysis, solution and command issuance are repeated within each round; At the end of the die-casting cycle, the control system generates a cycle-level file and records the model version, threshold version, and sampling configuration version. The control system updates model parameters based on multi-period periodic archives and generates corresponding threshold version identifiers.
Citation Information
Patent Citations
Mold runner structure
CN102717488A
Method for producing aluminum alloy steering knuckle through horizontal squeeze casting machine
CN106378432A
