Intelligent control method and device for stratified casting process, terminal equipment and medium

Abstract

The application is suitable for the field of casting technology, and provides an intelligent control method and device for a layered casting process, a terminal device and a medium. Through real-time collection of multi-source data such as liquid-phase surface temperature, mold temperature and liquid level height of a pouring site, the residual liquid-phase volume fraction and the solid-liquid interface position of the current layer are dynamically calculated, the macrosegregation risk is evaluated based on this, and then the process execution parameters such as the pouring start time, the target pouring temperature and the pouring speed of the subsequent ladle are instantaneously corrected and closed-loop controlled, thereby effectively overcoming the hysteresis and blindness of the traditional experience-based pouring, significantly improving the precise control ability of the layered casting process on the solidification state, ensuring the reliability of the interlayer metallurgical bonding, and greatly reducing the macrosegregation of the ingot and improving the overall composition uniformity and process stability.

Description

Technical Field

[0001] This application belongs to the field of casting technology, and in particular relates to an intelligent control method, device, terminal equipment and medium for a layered casting process. Background Technology

[0002] Large metal ingots are prone to defects such as macroscopic segregation, shrinkage cavities, porosity, and uneven microstructure during solidification. This is particularly true for high-alloy aluminum alloys, steel, and other multi-element alloy systems, where solute redistribution and liquid phase flow can lead to negative segregation at the bottom and positive segregation at the top, severely impacting product quality. To address this issue, existing technologies typically employ layered casting (i.e., intermittent pouring in multiple ladles), where molten metal is poured in multiple ladles sequentially, promoting layer-by-layer solidification to suppress large-scale solute migration. However, current layered casting implementations generally rely on the experience of technicians to determine the number of ladles, the interval between ladles, and the pouring temperature, lacking quantitative basis for process parameter design. More importantly, the lack of real-time sensing and feedback mechanisms for the current solidification state during pouring means that once process parameters are set, they cannot be dynamically adjusted based on actual on-site conditions (such as temperature fluctuations and compositional deviations). This easily leads to unstable interlayer bonding quality and difficulty in effectively controlling the solidification process of each layer, ultimately limiting the suppression of macroscopic segregation and making it difficult to consistently achieve high standards of ingot compositional uniformity. Therefore, how to solve the problems of poor segregation control stability and low quality consistency in the existing layered casting process has become a technical bottleneck that urgently needs to be overcome in this field. Summary of the Invention

[0003] In view of this, embodiments of this application provide an intelligent control method, device, terminal equipment and medium for layered casting process, which can solve the problems of poor segregation control stability and low quality consistency in existing layered casting processes.

[0004] In a first aspect, embodiments of this application provide an intelligent control method for a layered casting process, comprising: Obtain the casting parameters of the target ingot; A layered casting process plan is generated based on the ingot parameters. The process plan includes one or more process execution parameters. Real-time data collection during the implementation of the layered casting process; Calculate the solidification state of the current layer based on the on-site data; Assess the macrosegregation risk of the current and subsequent layers based on the solidification state; Based on the solidification state and macrosegregation risk, the process execution parameters for subsequent batches are dynamically modified, and modification instructions are generated and sent to the actuators based on the modified parameters.

[0005] Secondly, embodiments of this application provide an intelligent control device for a layered casting process, comprising: The acquisition module is used to acquire the casting parameters of the target ingot; The generation module is used to generate a layered casting process plan based on the ingot parameters. The process plan includes one or more process execution parameters. The data acquisition module is used to collect real-time field data when implementing the layered casting process. The calculation module is used to calculate the solidification state of the current layer based on the on-site data; The evaluation module is used to assess the macrosegregation risk of the current and subsequent layers based on the solidification state. The correction module is used to dynamically correct the process execution parameters of subsequent batches based on the solidification state and macrosegregation risk, and generate correction instructions based on the corrected parameters and send them to the actuator.

[0006] Thirdly, embodiments of this application provide a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the intelligent control method for the layered casting process described above.

[0007] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the intelligent control method for the layered casting process described above.

[0008] Fifthly, embodiments of this application provide a computer program product that, when run on a terminal device, enables the terminal device to execute the aforementioned intelligent control method for the layered casting process.

[0009] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment collects multi-source data such as liquid phase surface temperature, mold temperature and liquid level height at the pouring site in real time, dynamically calculates the remaining liquid phase volume fraction and solid-liquid interface position of the current layer, and assesses the risk of macrosegregation based on this. Then, it performs real-time correction and closed-loop control on process execution parameters such as pouring start time, target pouring temperature and pouring speed of subsequent batches. This effectively overcomes the lag and blindness of traditional experience-based pouring, significantly improves the ability to accurately control the solidification state in the layered casting process, ensures the reliability of interlayer metallurgical bonding, and thus greatly reduces macrosegregation of ingots, improves overall composition uniformity and process stability. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic diagram illustrating the implementation process of the intelligent control method for the layered casting process provided in this application embodiment.

[0012] Figure 2 This is a schematic diagram of the intelligent control device for the layered casting process provided in the embodiments of this application.

[0013] Figure 3 This is a schematic diagram of the structure of the terminal device provided in the embodiments of this application. Detailed Implementation

[0014] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are protected by this application.

[0015] It should be noted that the terms "comprising," "including," and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application, are intended to cover non-exclusive inclusion. For example, a process, method, terminal, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices. Terms such as "first" and "second" in the claims, specification, and accompanying drawings of this application, as well as relational terms, are used merely to distinguish one entity / operation / object from another entity / operation / object, and do not necessarily require or imply any such immediate relationship or order between these entities / operations / objects.

[0016] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0017] Large metal ingots are prone to defects such as macroscopic segregation, shrinkage cavities, porosity, and uneven microstructure during solidification. This is particularly true for high-alloy aluminum alloys, steel, and other multi-element alloy systems, where solute redistribution and liquid phase flow can lead to negative segregation at the bottom and positive segregation at the top, severely impacting product quality. To address this issue, existing technologies typically employ layered casting (i.e., intermittent pouring in multiple ladles), where molten metal is poured in multiple ladles sequentially, promoting layer-by-layer solidification to suppress large-scale solute migration. However, current layered casting implementations generally rely on the experience of technicians to determine the number of ladles, the interval between ladles, and the pouring temperature, lacking quantitative basis for process parameter design. More importantly, the lack of real-time sensing and feedback mechanisms for the current solidification state during pouring means that once process parameters are set, they cannot be dynamically adjusted based on actual on-site conditions (such as temperature fluctuations and compositional deviations). This easily leads to unstable interlayer bonding quality and difficulty in effectively controlling the solidification process of each layer, ultimately limiting the suppression of macroscopic segregation and making it difficult to consistently achieve high standards of ingot compositional uniformity. Therefore, how to solve the problems of poor segregation control stability and low quality consistency in the existing layered casting process has become a technical bottleneck that urgently needs to be overcome in this field.

[0018] In view of this, the embodiments of this application provide an intelligent control method for layered casting process. By collecting multi-source data such as liquid phase surface temperature, mold temperature and liquid level height at the pouring site in real time, the remaining liquid phase volume fraction and solid-liquid interface position of the current layer are dynamically calculated. Based on this, the risk of macrosegregation is assessed, and the process execution parameters such as pouring start time, target pouring temperature and pouring speed of subsequent pours are corrected and controlled in real time. This effectively overcomes the lag and blindness of traditional experience-based pouring, significantly improves the precise control of solidification state in layered casting process, ensures the reliability of interlayer metallurgical bonding, and thus greatly reduces macrosegregation of ingots, improves overall compositional uniformity and process stability.

[0019] To illustrate the technical solution of this application, specific embodiments are described below.

[0020] Figure 1 This illustration shows a flowchart of an intelligent control method for a layered casting process according to an embodiment of this application. This method can be applied to terminal devices. Terminal devices can be servers, service clusters, mobile phones, tablets, laptops, ultra-mobile personal computers (UMPCs), netbooks, etc.

[0021] Specifically, the intelligent control method for the above-mentioned layered casting process may include the following steps S101 to S106.

[0022] Step S101: Obtain the casting parameters of the target ingot.

[0023] The target ingot is the final metal ingot billet that is planned to be produced through this layered casting process.

[0024] Casting parameters are input parameters that describe the characteristics of the target ingot itself and the basic process requirements. They may include ingot geometry, target chemical composition, allowable segregation range, mold parameters, total pouring mass, single ladle capacity range, equipment capacity boundaries, and process constraints.

[0025] In the embodiments of this application, the terminal device can receive production order information from the upstream production management system, or the operator can input the core specifications of the casting task through a human-machine interface. The parameters acquired by the terminal device mainly include the alloy grade of the target ingot (corresponding to a specific chemical composition range), the target cross-sectional dimensions and length of the ingot, and the initial basic cooling mode required by the process standard. These casting parameters are stored in the control database of the terminal device, serving as the original basis for all subsequent intelligent decisions.

[0026] Step S102: Generate a layered casting process scheme based on the ingot parameters.

[0027] The process scheme includes one or more process execution parameters. These process execution parameters are specific physical quantity settings used to directly control the operation of the casting equipment in the layered casting process scheme, such as the pouring temperature, pouring speed, and cooling water flow rate and temperature for each batch of molten metal.

[0028] Layered casting process scheme refers to a complete work plan that plans multiple casting ladles (pouring batches) to be executed in sequence in order to complete the production of target ingots. Each ladle corresponds to a physical layer of the ingot.

[0029] In the embodiments of this application, the terminal device can utilize an internal process model to automatically perform calculations and inferences based on the input casting parameters. The model can combine the thermophysical parameters of the alloy, the ingot size, and production efficiency requirements to plan the total number of casting ladles, the approximate weight distribution of each ladle, and initialize a set of process execution parameters for each planned casting ladle.

[0030] Step S103: Real-time data collection during the implementation of the layered casting process.

[0031] Among them, the field data are physical quantities that reflect the state of the casting process, collected in real time by field sensors during the actual execution of the layered casting process, such as the temperature at different heights in the crystallizer, the inlet and outlet temperatures of the cooling water, the casting speed, and the liquid level.

[0032] In the embodiments of this application, once the casting process starts according to the generated process plan, key sensors arranged around the crystallizer, in the secondary cooling zone, and in the relevant water circuits begin to operate continuously. A temperature sensor array measures the temperature at different heights of the copper plate in the crystallizer and the surface temperature of the ingot; thermocouples measure the inlet and outlet temperatures of the cooling water; a casting speed encoder measures the real-time casting speed; and a liquid level detector monitors the metal level height within the crystallizer. The measurement signals from all these sensors are read, converted, and transmitted in real-time to the terminal equipment by a high-speed data acquisition system, forming a continuous time-series data stream.

[0033] Step S104: Calculate the solidification state of the current layer based on the on-site data.

[0034] Among them, the solidification state refers to the thermal state and phase transformation process inside the ingot layer that is currently solidifying. Its core is to describe the position and shape of the solidification front and the temperature field distribution in the molten pool.

[0035] In the embodiments of this application, the terminal device can invoke a built-in mathematical model or data-driven model of the solidification process, using real-time collected field data, especially temperature field data, as boundary conditions or input parameters of the model. The model can infer key states inside the ingot that cannot be directly measured at the current moment through numerical calculations or state estimation. This outputs the position and shape of the solidification front of the current casting layer, i.e., the interface between the solid and liquid phases, and the three-dimensional temperature distribution within the molten pool.

[0036] Step S105: Assess the macrosegregation risk of the current layer and subsequent layers based on the solidification state.

[0037] Among them, macrosegregation risk is the possibility and severity of uneven distribution of chemical composition on the macro scale of the ingot during the solidification process of the ingot due to factors such as the different distribution coefficients of solute elements in the solid and liquid phases and melt flow.

[0038] In the embodiments of this application, the terminal device can utilize the current layer solidification state information, combined with the alloy's solute distribution coefficient, interdendritic flow, etc., to perform analysis through a segregation prediction model. This model can assess the degree of segregation that may have formed in the currently solidified portion, and can also predict the trend of solute element redistribution in the subsequent solidification process (including the unsolidified portion of the current layer and the subsequent layers to be poured) based on the current molten pool shape and temperature gradient, thereby quantitatively assessing the risk level of macroscopic compositional inhomogeneity in the final ingot.

[0039] Step S106: Based on the solidification state and the macrosegregation risk, dynamically modify the process execution parameters for subsequent packages, and generate a modification command based on the modified parameters and send it to the actuator.

[0040] Among them, the actuator is a device that receives control commands and drives the physical equipment to move, such as a servo mechanism that controls the opening degree of the intermediate baffle rod or slide plate, an actuator that adjusts the opening degree of the water-cooled valve, and a motor driver that controls the casting traction speed.

[0041] In the embodiments of this application, the terminal device can comprehensively consider the current process state (solidification state) and future quality risks (macro-segregation risk). If the assessment deems the risk to be within an acceptable range, the original process plan can be maintained. If the risk exceeds a threshold, a parameter optimization algorithm can be initiated to reduce segregation risk and improve the solidification state. Within the process constraints, the optimal process execution parameters for one or more subsequent batches can be recalculated, such as adjusting the pouring temperature of subsequent batches, modifying the cooling intensity distribution, or fine-tuning the casting speed. After calculating the corrected parameter set, the terminal device can convert it into specific, time-sequential equipment control commands and send them to the corresponding actuators in real time through the industrial network, such as adjusting valve openings or modifying heating power settings, thereby actively suppressing the generation of defects.

[0042] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment collects multi-source data such as liquid phase surface temperature, mold temperature and liquid level height at the pouring site in real time, dynamically calculates the remaining liquid phase volume fraction and solid-liquid interface position of the current layer, and assesses the risk of macrosegregation based on this. Then, it performs real-time correction and closed-loop control on process execution parameters such as pouring start time, target pouring temperature and pouring speed of subsequent batches. This effectively overcomes the lag and blindness of traditional experience-based pouring, significantly improves the ability to accurately control the solidification state in the layered casting process, ensures the reliability of interlayer metallurgical bonding, and thus greatly reduces macrosegregation of ingots, improves overall composition uniformity and process stability.

[0043] In some specific embodiments of this application, the step of generating a layered casting process scheme based on the ingot parameters may specifically include steps S401 to S407.

[0044] Step S401: Based on the total casting mass and single-bundle capacity range in the ingot parameters, determine the candidate subcontracting quantity range, and based on the candidate subcontracting quantity range combined with the target segregation control requirements and equipment capacity boundaries, select the target subcontracting quantity.

[0045] The total casting mass is the total weight of molten metal required to complete the target ingot.

[0046] The single-ladle capacity range refers to the lower and upper limits of the weight of molten metal that can be safely and effectively contained and poured in ladles (such as steel ladles and tundishes) used in the foundry.

[0047] The candidate subcontractor quantity range is the set of all mathematically possible subcontractor quantities (i.e. how many packages are needed to complete the pouring) under the constraints of the total pouring mass and the single package capacity range.

[0048] The target segregation control requirements are quantitative or qualitative limits imposed by process standards or customers on the degree of macroscopic segregation in the final ingot (such as compositional range and depth of positive segregation zone).

[0049] Equipment capacity boundaries are the equipment and process factors that limit the number of subcontractors, other than the capacity of a single ladle, such as the total number of ladles available in the workshop, the overhead crane scheduling cycle, and the steel tapping rhythm of the refining furnace.

[0050] The target number of subcontractors is the total number of casting packages to be used, selected from the candidate range.

[0051] In the embodiments of this application, the terminal device can first perform mathematical calculations, dividing the total casting mass by the upper and lower limits of the single-bundle capacity range to obtain two values. These two values ​​are then rounded up and down to obtain a continuous set of integers as the candidate sub-bundle quantity range. For example, with a total mass of 100 tons and a single-bundle range of 20-30 tons, the candidate range is 4 to 5 bundles. The terminal device can then access a process knowledge base, which stores statistical data or model predictions on the impact of different sub-bundle quantities on the macroscopic segregation of typical ingots. The terminal device can compare each sub-bundle quantity within the candidate range with the target segregation control requirements, eliminating sub-bundle quantity options that historical data indicates are difficult to meet the segregation requirements. Simultaneously, the terminal device can also query the equipment status data of the production execution system to evaluate the available ladle quantity, crane operation cycle, and other capacity boundaries, further filtering out feasible sub-bundle quantities in actual production. Finally, through multi-objective trade-offs, an optimal value is determined from the remaining options as the target sub-bundle quantity.

[0052] Step S402: Based on the target number of sub-packages and the total casting mass, allocate the mass of each package of molten metal under the constraint of satisfying the single package capacity range.

[0053] The mass of each package of molten metal is the weight of molten metal allocated to each package in the target sub-package quantity.

[0054] In the embodiments of this application, the terminal equipment can construct an optimization problem with the target number of sub-packages as a fixed value, the total casting mass as an equality constraint that must be satisfied, and the single-package capacity range as an inequality constraint. The optimization objective can be to make the mass of each package as uniform as possible to reduce production fluctuations, or to allocate a specific proportion according to the subsequent composition design requirements. The terminal equipment can use linear programming or a simple heuristic algorithm to solve the problem. For example, in the case of pursuing uniformity, the terminal equipment will try to distribute the total mass evenly and check whether it exceeds the single-package capacity range; if it does, the excess package weight is set as the upper limit, the insufficient weight is set as the lower limit, and the weight of other packages is adjusted to compensate until all constraints are satisfied. After the solution is completed, a mass array containing the target number of sub-packages is output, that is, the mass of each package of molten metal.

[0055] Step S403: Based on the mass of each package of molten metal, the target chemical composition in the ingot parameters, and the allowable segregation range, the target composition of each package is set differently, while satisfying the overall average composition conservation constraint.

[0056] Among them, the target chemical composition and allowable segregation range refer to the average target values ​​of each major element specified in the target ingot alloy grade, as well as the upper and lower limits of the standard allowable composition fluctuation.

[0057] The target composition for each package is a chemical composition target value set individually for each package of molten metal.

[0058] The overall average composition conservation constraint means that the mass-weighted average of all target components in each package equals or infinitely approaches the target chemical composition of the ingot.

[0059] In the embodiments of this application, the terminal device can initiate a composition inverse design algorithm based on metallurgical principles, particularly solute redistribution theory. This algorithm aims to homogenize the composition of the final ingot, taking into account the inherent differences in composition between the first and later solidified portions during sequential solidification. Within the defined boundaries of the allowable segregation range, the algorithm can set different target compositions for different ladles (corresponding to different height positions of the ingot). For example, to compensate for solute enrichment in the later stages of solidification, a composition slightly lower than the target average might be set for the last pour. The specific design process strictly adheres to the overall average composition conservation constraint, meaning that the target composition of all ladles, after being weighted by mass, must be precisely equal to the target chemical composition in the ingot parameters.

[0060] Step S404: Set the target pouring temperature for each package according to the target composition of each package and the safe temperature window in the ingot parameters.

[0061] The safe temperature window refers to the minimum and maximum values ​​of the molten metal pouring temperature allowed to ensure a smooth casting process (such as good fluidity and avoidance of secondary oxidation) and the quality of the ingot (such as avoidance of hot cracking).

[0062] The target pouring temperature for each package is the planned temperature set for each package of molten metal when it is poured into the crystallizer.

[0063] In embodiments of this application, the terminal device can store a liquidus temperature calculation model for the alloy system. The terminal device can first calculate the theoretical liquidus temperature for each batch of molten metal in real time based on the target composition of each batch. Since the pouring temperature must be higher than the liquidus temperature to prevent premature solidification, the terminal device can add a superheat value determined based on alloy characteristics and process experience to the theoretical liquidus temperature, thereby obtaining the preliminary pouring temperature setting value for each batch. Subsequently, the terminal device can compare and trim these setting values ​​with the global safe temperature window specified in the ingot parameters to ensure that the setting value for each batch falls within the safe window. For example, if the calculated temperature of a batch of molten metal with a special composition is lower than the lower limit of the safe window, the terminal device can automatically set it to the lower limit value and issue a warning.

[0064] Step S405: Set the pouring speed for each package based on the mass of each package of molten metal, the geometry of the mold, and the allowable rate of liquid level rise.

[0065] Among them, the mold geometry refers to the key dimensions such as the cross-sectional shape, area, and height of the crystallizer or casting.

[0066] The allowable liquid level rise rate is the maximum speed at which the metal liquid level rises in the crystallizer, as allowed by the process, to prevent slag entrapment and ensure uniform initial solidification.

[0067] The pouring rate of each batch is the mass or volume of molten metal injected into the crystallizer per unit time during the pouring of each batch of molten metal, and is usually expressed as a curve that changes over time.

[0068] In the embodiments of this application, the terminal equipment can calculate the cross-sectional area of ​​the crystallizer based on the geometry of the mold. For each specified mass of molten metal, the terminal equipment can calculate the theoretical ingot height formed in the crystallizer after pouring. To ensure smooth pouring and meet the critical process constraint of the allowable liquid level rise rate, the terminal equipment can divide the mass of the molten metal by the cross-sectional area of ​​the crystallizer and the maximum allowable liquid level rise rate to obtain a theoretically shortest pouring time. Based on this, the terminal equipment can set an average pouring speed. Furthermore, the terminal equipment can use a standard pouring speed curve template, such as a "slow-fast-slow" curve, scale the template according to the mass of the package and the shortest time, generate a pouring speed setting curve for each package that varies with time, and ensure that the transient rise rate corresponding to any moment does not exceed the allowable value.

[0069] Step S406: Based on the target number of packages, the mass of each package of molten metal, and the heat dissipation conditions of the mold, predict the solidification process of each layer and generate an initial value for the interval between packages.

[0070] The initial value of the interval between the two packages is the waiting time initially planned by the system between the end of the previous package pouring and the start of the next package pouring.

[0071] In the embodiments of this application, the terminal device can utilize the pre-set target number of sub-packages, the mass of each package of molten metal, and temperature and velocity parameters, combined with the thermophysical parameters of the mold and the initial conditions of the cooling water, to call a solidification heat transfer mathematical model for offline simulation calculations. The model can simulate the complete process from the first package to the last package, predicting the temperature field changes, solidification layer thickness growth, and molten pool depth changes after each layer of molten metal is poured. By analyzing the simulation results, the terminal device can determine how long the current layer needs to solidify to a sufficient strength to withstand the impact of the next package of molten metal before pouring the next package, or whether the molten pool depth will decrease to a safe level. Based on this determination, the terminal device can calculate an initial waiting time for each pair of adjacent packages, i.e., an initial value for the package interval.

[0072] Step S407: Based on the solidification process and the allowable segregation range, generate the next pouring trigger condition and backup parameters for abnormal conditions, and generate the layered casting process scheme based on the target number of sub-packages, the mass of each package of molten metal, the target composition of each package, the target pouring temperature of each package, the pouring speed of each package, the initial value of the interval between packages, the next pouring trigger condition and the backup parameters for abnormal conditions.

[0073] The next pouring trigger condition is the real-time process state condition that must be met to determine whether the next pouring of molten metal can begin, such as the current liquid level dropping to a specific height or the solidified layer reaching a specific thickness.

[0074] The backup parameters for abnormal operating conditions are pre-set sets of backup process parameters that the system automatically switches to when the production process detects deviations from normal conditions (such as excessively low temperature or abnormal speed) to stabilize the process.

[0075] In the embodiments of this application, the terminal device can extract key state variables as pouring trigger conditions based on the solidification process. For example, "the thickness of the current solidified shell layer reaches a set threshold" or "the liquid level in the crystallizer drops to a specific height" can be set as the trigger conditions for the next pour. At the same time, the terminal device can preset a set of backup parameters (such as backup pouring speed and backup cooling intensity) for common abnormal conditions that may occur (such as the actual temperature being lower than the target value by a certain range, pouring interruption, etc.) based on the process experience database. The terminal device can structurally integrate all the parameters and conditions generated in the aforementioned steps, including the target number of sub-bundles, the mass of each bundle, composition, temperature, speed curve, recommended value of inter-bundle interval, dynamic trigger conditions, and backup parameters, and encapsulate them into a complete digital file of layered casting process scheme containing temporal logic and conditional branches.

[0076] This application's implementation method integrates composition design with process parameters such as temperature, speed, and rhythm, and verifies and optimizes the scheme through solidification simulation, generating dynamic triggering conditions and contingency plans in advance. It changes the traditional process design mode that relies on empirical formulas and static procedures, generating an optimized process package that fully considers quality objectives, equipment constraints, process physics, and possesses adaptive capabilities, thus improving the accuracy and reliability of layered casting process design.

[0077] In some specific embodiments of this application, the step of calculating the solidification state of the current layer based on the on-site data may specifically include steps S501 to S507.

[0078] Step S501: Preprocess the field data to obtain the current state input vector.

[0079] The state input vector is an ordered set of values ​​used to describe the complete state of the casting system at the current moment.

[0080] In the embodiments of this application, the terminal device can acquire raw readings from multiple field sensors in real time. This data may include temperatures from multiple thermocouples in the crystallizer, liquid level height from a laser rangefinder, package mass from a weighing system, and cooling water flow rate and temperature. This raw data typically contains measurement noise, transient interference, and possible communication anomalies. The terminal device can apply digital filters, such as low-pass filters, to each data stream to smooth high-frequency noise. It then checks the reasonableness of the data, such as whether the temperature is within the physically possible range and whether the liquid level is non-negative, and removes or marks obvious outliers. The terminal device can then align the data from all sensors according to their timestamps to ensure they describe the system state at the same moment. The aligned and cleaned multidimensional data is arranged into an array in a predetermined order; this array is the state input vector for the current moment.

[0081] Step S502: Calculate the geometric height and total metal volume of the current layer of molten metal based on the liquid level height, mold size, and package mass in the state input vector.

[0082] Among them, geometric height refers to the physical height of the liquid metal column that has been formed in the crystallizer for the current pouring batch.

[0083] Total metal volume refers to the total volume occupied by the current layer of molten metal in the crystallizer, which is determined by the mass and density of the molten metal.

[0084] In the embodiments of this application, the terminal device can read the measured value of the liquid level sensor from the state input vector. This value directly reflects the absolute height of the molten metal surface in the crystallizer. Combined with the mold geometry stored in the system, particularly the cross-sectional area of ​​the crystallizer, the terminal device can calculate the geometric height occupied by the current molten metal column. Simultaneously, the terminal device can read the mass data of the current ladle from the state input vector, or extrapolate it based on the cumulative mass already poured. Using the average density of the alloy near the liquidus temperature, the system converts the mass of the current ladle of molten metal into volume, which is the total metal volume of the current layer of molten metal. The calculated geometric height is used for spatial positioning, while the total metal volume is the benchmark for calculating the solidification fraction.

[0085] Step S503: Estimate the volume of solidified solids at the current moment based on the liquid surface temperature, mold temperature, cooling time, and mold heat dissipation conditions in the state input vector.

[0086] The solidified volume refers to the volume of the current layer of molten metal that has undergone a phase transition from liquid to solid from the start of casting to the present moment.

[0087] In the embodiments of this application, the terminal device can obtain the liquid phase surface temperature, which characterizes the thermal state of the molten metal, and multiple mold wall temperatures, which characterize the cooling intensity of the mold, from the state input vector. The terminal device also records or calculates the cumulative cooling time experienced from the start of pouring at the current layer to the current moment. Combining pre-stored mold material thermophysical parameters, interfacial heat transfer coefficients, and other mold heat dissipation conditions, a simplified one-dimensional or two-dimensional transient heat transfer model is constructed. This model uses the liquid phase surface temperature and mold temperature as boundary conditions and cooling time as the process variable to simulate the heat loss process from the molten metal through the solidified shell to the mold. Through model calculations, the terminal device can estimate the thickness of the solidified shell from the ingot surface towards the center. Combining the geometry of the molten pool, the terminal device can integrate the solidified shell thickness in three-dimensional space to estimate the total volume of solidified metal at the current moment.

[0088] Step S504: Calculate the remaining liquid phase volume fraction in the current layer based on the total metal volume and the solidified solid volume.

[0089] The remaining liquid volume fraction refers to the ratio of the volume of the unsolidified liquid portion in the current layer of molten metal to the total metal volume, and is a key parameter characterizing the solidification process.

[0090] In the embodiments of this application, the remaining liquid volume fraction can be calculated using the following formula: subtract the volume of solidified solids from the total metal volume to obtain the remaining liquid volume, and then divide the remaining liquid volume by the total metal volume to obtain the remaining liquid volume fraction. This fraction is a dimensionless number between 0 and 1, where a value of 1 indicates complete liquid state and a value of 0 indicates complete solidification.

[0091] Step S505: Calculate the position of the solid-liquid interface based on the remaining liquid phase volume fraction and the geometric height.

[0092] The solid-liquid interface position refers to the spatial coordinate description of the interface between the solid and liquid regions within the solidifying metal layer. It can usually be represented as a function related to the ingot height or an equivalent contour position.

[0093] In embodiments of this application, the terminal device can estimate the spatial location of the solidification front by utilizing the remaining liquid volume fraction and combining reasonable assumptions about its shape. When vertical solidification is dominant and lateral solidification is ignored, the solid-liquid interface can be considered a horizontal plane. In this case, the height of the solid-liquid interface can be directly estimated by multiplying the geometric height by the remaining liquid volume fraction. For example, if the geometric height is 1 meter and the remaining liquid volume fraction is 0.4, the solid-liquid interface can be estimated to be approximately 0.4 meters above the bottom. If more complex shapes are considered, the terminal device can use a parameterized interface shape function, using the remaining liquid volume fraction as a constraint, to solve for the function parameters and thus determine the interface profile.

[0094] Step S506: Calculate the solidification rate based on the change in the solid-liquid interface position during continuous sampling, and calculate the local temperature gradient based on the temperature difference between different temperature measurement points in the state input vector.

[0095] Among them, the solidification rate refers to the linear velocity of the solid-liquid interface moving forward in a direction perpendicular to its own normal, and is a dynamic parameter characterizing the speed of solidification.

[0096] Local temperature gradient refers to the rate of temperature change per unit distance at a specific location or in a specific direction within an ingot. It is a key thermodynamic parameter that affects the solidification morphology and solute transport.

[0097] In the embodiments of this application, the terminal device can record and cache the solid-liquid interface position calculated at multiple consecutive sampling times. By calculating the difference between the interface position at the current time and the interface position at the previous time and dividing it by the sampling time interval, the average propagation velocity of the solid-liquid interface during this time interval, i.e., the solidification rate, can be obtained. Simultaneously, the terminal device can select readings from two or more spatially adjacent temperature measurement points from the state input vector. For example, the temperature values ​​of two thermocouples arranged along the height direction on the crystallizer wall can be selected. Calculating the temperature difference between these two points and dividing it by the physical distance between them yields the local temperature gradient of the region in a specific direction. The terminal device can perform similar calculations at different locations and in different directions to obtain the gradient distribution characteristics of the temperature field.

[0098] Step S507: Determine the solidification state of the current layer based on the remaining liquid phase volume fraction, the solid-liquid interface position, the solidification rate, and the local temperature gradient.

[0099] In the embodiments of this application, the terminal device can package and structure all key parameters, namely the remaining liquid volume fraction, solid-liquid interface position, solidification rate, and local temperature gradient. These parameters together constitute a multi-dimensional, quantitative description of the current layer's solidification state. For example, the terminal device can output a status report indicating that 60% of the metal in the current layer has solidified, the solid-liquid interface is approximately at a certain horizontal position above the bottom, the interface is advancing upwards at a rate of several millimeters per minute, and a certain temperature gradient exists near the mushy region. This comprehensive state description is the final determined solidification state of the current layer.

[0100] In some specific embodiments of this application, the macrosegregation risk includes the risk of local solute enrichment, the risk of interlayer composition fluctuation, the risk of negative segregation at the bottom, and the risk of positive segregation at the top. The assessment of the macrosegregation risk of the current layer and subsequent layers based on the solidification state may specifically include steps S601 to S604.

[0101] The risk of localized solute enrichment is a risk that occurs at the micro or macro scale of the ingot, due to the uneven distribution of solute elements between the solid and liquid phases during solidification, and their continuous enrichment in the interdendritic space or the liquid phase at the solidification front, ultimately leading to a significant increase in the actual chemical composition of the local area compared to the average composition.

[0102] Interlayer compositional fluctuation risk refers to the possibility that, in layered casting, the final ingot may exhibit abrupt changes in chemical composition at the interface between layers due to differences in the composition of molten metal between adjacent pours, abnormal solidification behavior at the interlayer junction, or remelting.

[0103] The risk of negative segregation at the bottom of the ingot is that the actual chemical composition in the bottom region is lower than the average composition due to the sinking of fine equiaxed grains formed by rapid cooling in the early stage of solidification or the solute trapping effect.

[0104] The risk of top positive segregation is that at the top of the ingot or in the central region of final solidification, the actual chemical composition of the region is significantly higher than the average composition because the solute-rich liquid phase gathers and solidifies here at the end of solidification.

[0105] Step S601: Receive the solidification state and the actual component detection value of the current batch and the actual component deviation of the previous batch from the field data, and construct a segregation risk assessment input set.

[0106] In embodiments of this application, the terminal device can receive the solidification state. It can also acquire the actual chemical composition analysis values ​​of the currently solidifying batch obtained through online spectrometer or rapid sampling detection. Furthermore, it can query a historical database for deviation records between the actual composition detection values ​​of the previous batch or several batches of molten metal that have already been poured and their respective target composition settings. The terminal device can encapsulate these data from different sources and of different types according to a predefined data structure to create a dedicated segregation risk assessment input set for the current moment. This set ensures that the assessment model can simultaneously acquire all the necessary information describing the current thermal state, current material composition, and historical composition fluctuations.

[0107] Step S602: Based on the remaining liquid phase volume fraction, local temperature gradient, and solidification rate in the solidified state, combined with the actual component detection value of the current batch, assess the solute redistribution intensity and local solute enrichment risk of the current layer.

[0108] The solute redistribution intensity refers to the strength of the tendency for solute to be displaced from the solidification front into the liquid phase at the solid-liquid interface, driven by the difference in equilibrium concentration of solute elements in the solid and liquid phases. It is influenced by the partition coefficient, local solidification rate, and diffusion conditions, and is an intrinsic factor for local enrichment.

[0109] In the embodiments of this application, the terminal device can use the residual liquid volume fraction, local temperature gradient, and solidification rate as key process variables, and the actual composition detection value of the current batch as the initial solute conditions. First, based on classical solidification theory formulas, the local solidification morphology parameters are calculated using the ratio of the local temperature gradient to the solidification rate, and the thickness of the mushy region is estimated by combining the solute diffusion coefficient. Then, according to the solute balance equation, at a given solidification rate, the amount of solute displaced per unit volume of metal during solidification is calculated to quantify the solute redistribution intensity. Next, the transport and accumulation process of these displaced solutes in the mushy region or molten pool is simulated, and the maximum local solute concentration that may form before the end of solidification is predicted by combining the enrichment space size reflected by the residual liquid volume fraction. By comparing this predicted concentration with the allowable segregation range, a quantitative level or probability value of the local solute enrichment risk of the current layer is output.

[0110] Step S603: Based on the solid-liquid interface position in the solidified state and the actual composition deviation of the previous batch, assess the interlayer metallurgical bonding stability and the risk of interlayer composition fluctuation.

[0111] Interlayer metallurgical bonding stability refers to the ability of subsequent molten metal to achieve good metallurgical bonding and avoid cold shuts, oxide inclusions or abnormal segregation zones in layered casting.

[0112] In the embodiments of this application, the location of the solid-liquid interface determines the temperature difference and thermal state between the subsequent batch of molten metal and the solidified surface of the previous batch. The terminal equipment can use the interface location to calculate the thickness and surface temperature of the solidified shell of the previous batch. Combined with the pouring temperature of the subsequent batch, the thermal shock and remelting depth at the contact of the two metal layers can be assessed. Simultaneously, actual compositional deviation data from previous batches reveals inherent compositional differences between the layers. The terminal equipment can combine the thermal state assessment with the compositional difference assessment, using a statistical model or empirical rule base trained on a large amount of historical production data to determine the likelihood of achieving good metallurgical bonding between the two layers under current conditions. If the predicted remelting depth is too deep, it may dilute the interlayer composition but increase the risk of hot cracking; if the bonding is poor, it may form a weak interface with abrupt compositional changes. The terminal equipment can integrate these factors and output an assessment result of the risk of interlayer compositional fluctuations, such as "low risk - expected good bonding," "medium risk - possible slight compositional step," and "high risk - beware of cold shuts or abnormal segregation bands."

[0113] Step S604: Based on the distribution trend of the remaining liquid phase volume fraction in the ingot height direction, assess the risk of negative segregation at the bottom and positive segregation at the top.

[0114] In embodiments of this application, the terminal device can analyze the calculated or predicted distribution trend of the residual liquid phase volume fraction at different heights of the ingot. For the risk of bottom negative segregation, the terminal device can focus on the early solidification stage. By analyzing the trend of the residual liquid phase fraction rapidly decreasing to zero in the bottom region, combined with the high solidification rate in this region, the terminal device can determine the strength of the quench zone formation. Strong bottom quenching is usually accompanied by a significant tendency for negative segregation. The terminal device can quantify this risk by matching or comparing the current trend with solidification curves known in historical databases that lead to severe bottom negative segregation. For the risk of top positive segregation, the terminal device can focus on the late solidification stage. By analyzing the trend of a slower rate of decrease in the residual liquid phase fraction and a widening of the mushy zone in the top region, the terminal device identifies the formation characteristics of a "V"-shaped or inverted conical molten pool at the end, which are typical precursors to top positive segregation. The terminal device can assess the severity of final shrinkage cavities and positive segregation based on the volume of the mushy zone at the end and the estimated solute enrichment in the residual liquid phase.

[0115] This application's implementation methods address three different scales and mechanisms of segregation phenomena: localized solute enrichment within the solidification process, interlayer compositional fluctuations, and overall compositional deviations at the beginning and end of the ingot. Specific evaluation logics are designed for each. Localized evaluation delves into the physical essence of solidification, interlayer evaluation considers both thermal and compositional matching, and beginning-and-end evaluation grasps the macroscopic solidification sequence. This allows for real-time disclosure of potential chemical compositional inhomogeneities that may arise during the casting process from multiple key levels, transforming previously difficult-to-quantify quality issues, which relied heavily on post-cast analysis, into process indicators that can be monitored and quantified in real time.

[0116] In some specific embodiments of this application, the method of dynamically correcting the process execution parameters of subsequent batches based on the solidification state and the macrosegregation risk may further include steps S701 to S704.

[0117] Step S701: Receive the solidification state and the macrosegregation risk, and determine the current deviation type by combining the actual component detection values ​​of the previous batch in the field data.

[0118] The deviation types include at least one of temperature deviation, composition deviation, and solidification process deviation.

[0119] Composition deviation refers to the difference between the actual chemical composition of the molten metal obtained through rapid testing and the target composition value set for that batch in the process plan.

[0120] Temperature deviation refers to the difference between the measured molten metal pouring temperature or the molten pool temperature in the crystallizer and the corresponding target pouring temperature or ideal temperature curve.

[0121] In the embodiments of this application, the terminal device can simultaneously receive solidification state data packets and macrosegregation risk assessment results. It can also acquire the latest field data, particularly a rapid analysis report of the actual chemical composition of the most recently completed pour. The terminal device can perform multi-dimensional comparisons between real-time data and the preset process targets for the current layer. It calculates the difference between the measured pouring temperature and the target temperature; if it exceeds a threshold, it marks a temperature deviation. It compares the actual composition of previous pours with the target composition; if any key element deviation exceeds the allowable range, it marks a composition deviation. Furthermore, the terminal device can perform matching degree analysis between the actual change curve and the predicted curve of key parameters in the solidification state, such as the residual liquid phase volume fraction, or check whether the solidification rate and temperature gradient deviate from the safety window; if they deviate, it marks a solidification process deviation. This yields one or more labels indicating the main types of deviations and their severity.

[0122] Step S702: When the deviation type is composition deviation, calculate the amount of alloy element to be added in subsequent batches based on the difference between the actual composition detection value and the target composition of the previous batch, and perform online compensation correction on the target composition in the process execution parameters of the subsequent batches based on the amount of alloy element to be added.

[0123] The amount of alloying element added refers to the mass of a specific alloying element that needs to be added to the ladle to correct the identified compositional deviations and ensure that the overall composition of the molten metal in subsequent ladles approaches the target value after considering the addition.

[0124] In the embodiments of this application, when the current major deviation is determined to be a composition deviation, the terminal device can activate the composition compensation algorithm. First, it reads the actual composition detection value of the preceding ladle that caused the deviation and calculates the difference between each element and the originally set target composition. Then, based on the principle of metallurgical material balance, considering the original target composition of the subsequent ladle to be corrected, its molten metal mass, and the recovery rate of alloying elements, it calculates in reverse how much of a specific type of alloying additive needs to be added to the molten steel of the subsequent ladle so that the corrected expected composition of the ladle can, to a certain extent, offset the average composition deviation caused by the preceding ladle, ensuring that the average composition of the entire furnace ingot still tends towards the overall target. The calculated amount of alloying element to be added is a precise quality instruction. The terminal device can update the process execution parameters of the subsequent ladle online, modifying its target composition setting value to the new value after compensation calculation.

[0125] Step S703: When the deviation type is temperature deviation or solidification process deviation, determine whether the current layer meets the metallurgical bonding requirements and segregation control requirements based on the remaining liquid phase volume fraction in the solidification state, the solid-liquid interface position, and the macrosegregation risk index.

[0126] Among them, solidification process deviation refers to the difference between solidification state parameters obtained by actual monitoring or calculation, such as residual liquid volume fraction, solid-liquid interface position, solidification rate, etc., and the process trajectory predicted based on ideal conditions.

[0127] Metallurgical bonding requirements refer to the thermodynamic and kinetic conditions, such as interface temperature, solidified layer thickness, and remelting depth, required to ensure a strong metallurgical bond between layers when two metal layers come into contact.

[0128] Segregation control requirements refer to the constraints set on key parameters of the solidification process, such as temperature gradient, solidification rate, and size of the pasty region, in order to suppress the risk of macrosegregation to an acceptable level.

[0129] In the embodiments of this application, when the deviation is determined to be a temperature deviation or a solidification process deviation, the terminal device can use the remaining liquid phase volume fraction and solid-liquid interface position in the solidified state as core inputs to accurately assess the actual thermal state and solidified shell thickness of the current layer. This real-time data is then substituted into a pre-stored metallurgical bonding model, which defines the interface temperature conditions and minimum solidified shell thickness required to achieve good interlayer bonding. Simultaneously, a risk index output from the macrosegregation risk assessment module is used, quantifying the level of risk for current state-of-the-art segregation, such as downward segregation and center segregation. The actual state is synchronously verified against metallurgical bonding requirements and segregation control requirements. For example, it determines whether the current solidified shell thickness is sufficient to withstand the impact of the next pour and whether the current molten pool shape will lead to severe end segregation. The verification result is output as a binary logic judgment, indicating whether the current layer state "meets" or "does not meet" the quality and process requirements.

[0130] Step S704: If the current layer does not meet the metallurgical bonding requirements and the segregation control requirements, then the pouring start time, target pouring temperature, or pouring speed in the process execution parameters of the subsequent batches shall be adjusted.

[0131] In some more specific embodiments of this application, adjusting the pouring start time, target pouring temperature, or pouring speed in the process execution parameters of the subsequent packages may specifically include steps S801 to S804.

[0132] Step S801: Based on the remaining liquid phase volume fraction in the solidified state, the solid-liquid interface position, and the macrosegregation risk index, determine whether the current layer meets the preset metallurgical bonding window and segregation control window.

[0133] The metallurgical bonding window refers to an allowable range of thermal states that the current layer should be in when casting the next batch of metal to ensure a good, defect-free weld between adjacent metal layers. This range is usually quantified by the remaining liquid phase volume fraction and the solid-liquid interface position of the current layer. For example, it may require that the remaining liquid phase fraction be below a certain value and the solidified shell thickness be greater than a certain value when casting the next batch.

[0134] The segregation control window refers to a safe operating range for key state parameters in the solidification process, defined to keep the risk of macrosegregation within an acceptable level. This range is quantified by a macrosegregation risk index, which is required to be below a preset safety threshold.

[0135] In the embodiments of this application, the terminal device can obtain the latest values ​​of the remaining liquid phase volume fraction and the solid-liquid interface position from the real-time data stream, and simultaneously obtain the latest value of the macroscopic segregation risk index output by the segregation risk assessment module. The terminal device internally stores a pre-set process specification database, which defines the requirements of the metallurgical bonding window for the remaining liquid phase volume fraction and the solid-liquid interface position. For example, it requires that when starting the next batch of casting, the remaining liquid phase fraction of the current layer must be less than 5% and the solid-liquid interface position must be higher than a certain height threshold. It also defines the requirements of the segregation control window for the macroscopic segregation risk index, for example, requiring it to be lower than risk level two. The terminal device can logically compare the obtained real-time values ​​with these preset window boundary conditions one by one. The result of the judgment generates a set of Boolean logic outputs, respectively indicating whether the remaining liquid phase fraction is within the window, whether the solid-liquid interface position is within the trigger interval, and whether the segregation risk index is within the safe range.

[0136] Step S802: If the remaining liquid phase volume fraction is higher than the upper limit of the preset range, or the solid-liquid interface position has not yet reached the preset trigger interval, or the segregation risk index is higher than the upper limit of the preset range, then it is determined that the current layer does not meet the metallurgical bonding window and segregation control window, and the waiting time correction amount is calculated based on the difference between the remaining liquid phase volume fraction and the upper limit of the preset range, and the difference between the segregation risk index and the upper limit of the preset range.

[0137] The waiting time correction refers to the time value that needs to be added or reduced from the original planned time in order to adjust the pouring start time.

[0138] In the embodiments of this application, when any condition is not met—that is, if the remaining liquid fraction is too high (indicating insufficient solidification), the interface position is too low (indicating insufficient shell thickness), or the risk index is too high (indicating a high tendency for segregation)—the terminal device can determine that the current layer state is not up to standard. In this case, the terminal device can first calculate the deviation of the parameters that do not meet the conditions. For example, it can calculate the difference between the measured value of the remaining liquid fraction and the preset upper limit, and the difference between the measured value of the segregation risk index and the preset upper limit. Then, it calls an internally preset or self-learned response relationship model, which defines the waiting time required to compensate for a unit deviation. For example, for every 2% increase in the remaining liquid fraction above the upper limit, an additional 30 seconds of waiting is required; for every level increase in the segregation risk index above the upper limit, an additional one minute of waiting is required. The terminal device can multiply each deviation by the corresponding response coefficient and sum all the calculated compensation times to obtain the total waiting time correction. This correction is usually a positive number, meaning that a delay is needed based on the originally planned start time of the next batch of pouring.

[0139] Step S803: If the remaining liquid volume fraction is lower than the lower limit of the preset range, or the solid-liquid interface position has exceeded the upper limit of the preset trigger interval, or the segregation risk index is lower than the lower limit of the preset range, then it is determined that the current layer has the risk of overcooling or insufficient interlayer bonding, and the temperature correction amount or flow correction amount is calculated based on the difference between the remaining liquid volume fraction and the lower limit of the preset range.

[0140] The temperature correction amount refers to the Celsius temperature value that needs to be increased or decreased based on the original set temperature in order to adjust the target pouring temperature.

[0141] Flow correction refers to the unit time mass flow rate value that needs to be increased or decreased based on the original set flow rate in order to adjust the pouring speed.

[0142] In the embodiments of this application, when certain parameters are abnormally low, such as a low residual liquid fraction indicating insufficient overheating of the molten metal that may be nearing solidification, a high interface position indicating excessive waiting leading to low temperatures in the lower layers, or an abnormally low risk index potentially accompanied by abnormal solidification patterns, the terminal device can determine that the current layer is at risk of undercooling or insufficient subsequent bonding heat. In this case, the terminal device can calculate the negative deviation between the measured residual liquid fraction value and a preset lower limit. Based on the magnitude of this negative deviation, and in conjunction with other parameters (such as the current temperature), a built-in rule determines whether to adjust the temperature or the flow rate. Typically, when there is a large negative deviation and the current temperature is low, a temperature correction is tended to be calculated to increase the pouring temperature of the next batch, providing more heat for interlayer bonding. Specifically, the calculation method can be to consult a temperature compensation coefficient table based on the negative deviation value to determine the required temperature correction. If the negative deviation is mainly due to an error in estimating the waiting time, causing the process to advance, a flow rate correction may be calculated, i.e., appropriately increasing the initial pouring speed of the next batch to quickly provide heat.

[0143] Step S804: Adjust the pouring start time according to the waiting time correction amount, adjust the target pouring temperature according to the temperature correction amount, and adjust the pouring speed according to the flow rate correction amount.

[0144] In the embodiments of this application, the terminal device can update the process execution parameters of subsequent affected batches online in real time based on the calculated specific correction amount. If a waiting time correction amount is calculated, the terminal device can add the correction amount to the originally planned next batch pouring start time to obtain the adjusted pouring start time, update the corresponding time node in the process plan, and simultaneously start the corresponding delay waiting timer. If a temperature correction amount is calculated, the terminal device can add the correction amount to the originally set target pouring temperature for the batch to obtain the new target pouring temperature, update the process parameter database, and simultaneously send the new temperature setpoint to the refining station or heating station. If a flow rate correction amount is calculated, the terminal device can apply the correction amount to the base value or a specific stage of the original pouring speed curve to generate a corrected pouring speed curve and update the instruction sequence issued to the pouring control system.

[0145] This application's implementation method utilizes a metallurgical bonding window and a segregation control window to rigorously verify the real-time state, transforming complex quality requirements into monitorable boundary conditions. Furthermore, for deviations of two different natures—state lag and lead—waiting time correction and temperature / flow rate correction strategies based on difference calculations are designed respectively, ensuring that the adjustment amount is proportional to the degree of deviation, thus achieving refined control response. This effectively suppresses poor interlayer bonding and macroscopic segregation defects caused by process fluctuations, significantly improving the stability of the casting process and the uniformity of ingot quality.

[0146] In some specific embodiments of this application, the method may further include steps S901 and S902.

[0147] Step S901: Obtain the composition distribution, macrosegregation and microstructure uniformity data of the ingot, and compare the results with the casting parameters and the layered casting process scheme to obtain the process deviation analysis results.

[0148] The process deviation analysis results refer to the systematic deviations, regular differences, and their quantitative descriptions identified by systematically comparing the measured data of the final ingot with the initially set casting parameter targets and the actual layered casting process scheme. For example, the composition at a specific height is consistently higher, or the proportion of equiaxed crystal regions is lower under specific process conditions.

[0149] In the embodiments of this application, the terminal device can retrieve the complete result dataset of the current batch of ingots from the manufacturing execution system or quality management system. Simultaneously, it retrieves the original casting parameters used in producing the ingot and the actual layered casting process plan issued and executed from its own historical database. The terminal device can spatially align and quantitatively compare the result data with the target. For example, it compares the measured positive segregation depth at the ingot head with the predicted molten pool depth in the process plan, compares the core composition range with the design allowable range, and compares the proportion of equiaxed crystal regions with historical excellent records. Through statistical analysis and image processing algorithms, it identifies regular deviation patterns and generates a structured process deviation analysis result report, clearly indicating what kind of quality deviation occurred under what process conditions.

[0150] Step S902: Based on the process deviation analysis results, extract the key factors affecting segregation control and their weights, and update them to the process database.

[0151] The key factors include at least one of the following: number of packages, mass of a single package, target composition of each package, target pouring temperature of each package, pouring speed of each package, initial value of interval between packages, and triggering condition for pouring the next package.

[0152] In the embodiments of this application, the terminal device can receive process deviation analysis result reports and associate them with the complete set of process parameters used in the current production to form new sample data. The terminal device can store this new data in a continuously expanding historical case library. When the number of samples in the library reaches a certain amount, the terminal device can initiate data mining analysis. Multivariate statistical methods, such as correlation analysis, principal component regression, or more complex machine learning algorithms, are used to analyze all historical samples to quantify the correlation strength between each process design parameter and each segregation result indicator. By analyzing and identifying which process parameters, such as the number of sub-packages, the set gradient of the target component in each package, and the initial value of the inter-package interval, have the highest explanatory power for the changes in the final component homogeneity or segregation level, these parameters are determined as key factors affecting segregation control. At the same time, the contribution of each key factor to a specific quality result is calculated, which is its weight. For example, the analysis may show that the "number of sub-packages" has the highest weight in terms of the positive segregation depth at the head, while the "difference in the target component of each package" has the highest weight in terms of the fluctuation of the interlayer component. The list of key factors and their weight coefficients identified or updated in this analysis will then be integrated into the central process database to iteratively update the original process knowledge model or rule base.

[0153] This application's implementation method utilizes data mining and machine learning on accumulated process result data to automatically identify the key process factors that truly dominate segregation control and quantifies their influence weights, thereby transforming implicit process experience into explicit decision-making knowledge. This enables the entire intelligent control system not only to optimize process parameters for a single production run but also to continuously revise and enrich its core process design rule base based on long-term quality feedback, resulting in increasingly accurate and reliable process solutions for similar products or quality issues.

[0154] Figure 2 This illustration shows a structural schematic diagram of an intelligent control device for a layered casting process according to an embodiment of this application. The intelligent control device 2 for the layered casting process can be configured on a terminal device. Specifically, the intelligent control device 2 for the layered casting process includes: The acquisition module 201 is used to acquire the casting parameters of the target ingot; The generation module 202 is used to generate a layered casting process scheme based on the ingot parameters, wherein the process scheme includes one or more process execution parameters; The data acquisition module 203 is used to acquire on-site data in real time when implementing the layered casting process. Calculation module 204 is used to calculate the solidification state of the current layer based on the field data; Evaluation module 205 is used to evaluate the macrosegregation risk of the current layer and subsequent layers based on the solidification state; The correction module 206 is used to dynamically correct the process execution parameters of subsequent packages based on the solidification state and the macrosegregation risk, and generate correction instructions based on the corrected parameters and send them to the actuator.

[0155] The beneficial effects of this application embodiment compared with the prior art are as follows: This application embodiment collects multi-source data such as liquid phase surface temperature, mold temperature and liquid level height at the pouring site in real time, dynamically calculates the remaining liquid phase volume fraction and solid-liquid interface position of the current layer, and assesses the risk of macrosegregation based on this. Then, it performs real-time correction and closed-loop control on process execution parameters such as pouring start time, target pouring temperature and pouring speed of subsequent batches. This effectively overcomes the lag and blindness of traditional experience-based pouring, significantly improves the ability to accurately control the solidification state in the layered casting process, ensures the reliability of interlayer metallurgical bonding, and thus greatly reduces macrosegregation of ingots, improves overall composition uniformity and process stability.

[0156] In some embodiments of this application, the generation module 202 is further configured to: determine the candidate subcontracting quantity range based on the total casting mass and single-bundle capacity range in the ingot parameters, and select the target subcontracting quantity based on the candidate subcontracting quantity range combined with the target segregation control requirements and equipment capacity boundaries; Based on the target number of sub-packages and the total casting mass, the mass of each package of molten metal is allocated under the constraint of satisfying the single-package capacity range; Based on the mass of each package of molten metal, the target chemical composition in the ingot parameters, and the allowable segregation range, the target composition of each package is set differently, while satisfying the overall average composition conservation constraint. Based on the target composition of each package and the safe temperature window in the ingot parameters, the target pouring temperature of each package is set; The pouring speed of each package is set according to the mass of each package of molten metal, the geometry of the mold, and the allowable rate of liquid level rise. Based on the target number of sub-packages, the mass of each package of molten metal and the heat dissipation conditions of the mold, the solidification process of each layer is predicted and the initial value of the inter-package interval is generated. Based on the solidification process and the allowable segregation range, the next pouring trigger condition and backup parameters for abnormal conditions are generated. The layered casting process scheme is generated based on the target number of sub-lots, the mass of the molten metal in each lot, the target composition of each lot, the target pouring temperature of each lot, the pouring speed of each lot, the initial value of the interval between lots, the next pouring trigger condition and the backup parameters for abnormal conditions.

[0157] In some embodiments of this application, the above-mentioned calculation module 204 is further configured to: preprocess the field data to obtain the current state input vector; Based on the liquid level height, mold size, and package mass in the state input vector, calculate the geometric height and total metal volume of the current layer of molten metal. Based on the liquid surface temperature, mold temperature, cooling time, and mold heat dissipation conditions in the state input vector, estimate the volume of solidified solids at the current moment; Calculate the remaining liquid phase volume fraction in the current layer based on the total metal volume and the solidified solid volume; The position of the solid-liquid interface is calculated based on the remaining liquid volume fraction and the geometric height. The solidification rate is calculated based on the change in the solid-liquid interface position over continuous sampling time, and the local temperature gradient is calculated based on the temperature difference at different temperature measurement points in the state input vector. The solidification state of the current layer is determined based on the remaining liquid volume fraction, the solid-liquid interface position, the solidification rate, and the local temperature gradient.

[0158] In some embodiments of this application, the macrosegregation risk includes local solute enrichment risk, interlayer composition fluctuation risk, bottom negative segregation risk and top positive segregation risk. The assessment module 205 is also used to: receive the solidification state and the actual component detection value of the current batch and the actual component deviation of the previous batch in the field data, and construct a segregation risk assessment input set. Based on the remaining liquid phase volume fraction, local temperature gradient, and solidification rate in the solidified state, combined with the actual component detection value of the current batch, the solute redistribution intensity and local solute enrichment risk of the current layer are assessed. Based on the solid-liquid interface position in the solidification state and the actual composition deviation of the previous batch, assess the stability of interlayer metallurgical bonding and the risk of interlayer composition fluctuation. Based on the distribution trend of the remaining liquid phase volume fraction along the ingot height direction, assess the risk of negative segregation at the bottom and positive segregation at the top.

[0159] In some embodiments of this application, the above-mentioned correction module 206 is further configured to: receive the solidification state and the macrosegregation risk, and determine the current deviation type by combining the actual component detection values ​​of the previous batch in the field data, wherein the deviation type includes at least one of temperature deviation, component deviation and solidification process deviation. When the deviation type is composition deviation, the amount of alloy element to be added in subsequent batches is calculated based on the difference between the actual composition detection value of the previous batch and the target composition, and the target composition in the process execution parameters of the subsequent batches is compensated and corrected online based on the amount of alloy element to be added. When the deviation type is temperature deviation or solidification process deviation, the current layer is judged to meet the metallurgical bonding requirements and segregation control requirements based on the remaining liquid phase volume fraction in the solidification state, the solid-liquid interface position and the macrosegregation risk index. If the current layer does not meet the metallurgical bonding requirements and the segregation control requirements, the pouring start time, target pouring temperature, or pouring speed in the process execution parameters of the subsequent batches will be adjusted.

[0160] In some embodiments of this application, the above-mentioned correction module 206 is further used to: determine whether the current layer meets the preset metallurgical bonding window and segregation control window based on the remaining liquid phase volume fraction in the solidification state, the solid-liquid interface position and the macrosegregation risk index; If the remaining liquid volume fraction is higher than the upper limit of the preset range, or the solid-liquid interface position has not yet reached the preset trigger interval, or the segregation risk index is higher than the upper limit of the preset range, then it is determined that the current layer does not meet the metallurgical bonding window and segregation control window, and the waiting time correction amount is calculated based on the difference between the remaining liquid volume fraction and the upper limit of the preset range, and the difference between the segregation risk index and the upper limit of the preset range. If the remaining liquid volume fraction is lower than the preset lower limit, or the solid-liquid interface position has exceeded the preset trigger interval upper limit, or the segregation risk index is lower than the preset lower limit, then it is determined that the current layer has the risk of being overcooled or having insufficient interlayer bonding, and the temperature correction amount or flow correction amount is calculated based on the difference between the remaining liquid volume fraction and the preset lower limit. The pouring start time is adjusted according to the waiting time correction amount, the target pouring temperature is adjusted according to the temperature correction amount, and the pouring speed is adjusted according to the flow rate correction amount.

[0161] In some embodiments of this application, the intelligent control device 2 for the layered casting process further includes an update module, used to: acquire data on the composition distribution, macrosegregation and microstructure uniformity of the ingot, and compare the data with the casting parameters and the layered casting process scheme to obtain process deviation analysis results; Based on the process deviation analysis results, key factors affecting segregation control and their weights are extracted and updated to the process database. The key factors include at least one of the following: number of sub-packages, quality of a single package, target composition of each package, target pouring temperature of each package, pouring speed of each package, initial value of interval between packages, and triggering condition for pouring the next package.

[0162] like Figure 3 The diagram shown is a schematic of a terminal device provided in an embodiment of this application. The terminal device 3 may include: a processor 301, a memory 302, and a computer program 303 stored in the memory 302 and executable on the processor 301, such as an intelligent control program for a layered casting process. When the processor 301 executes the computer program 303, it implements the steps in the aforementioned intelligent control embodiments of the layered casting process, for example... Figure 1 Steps S101 to S106 are shown.

[0163] A computer program can be divided into one or more modules / units. One or more modules / units are stored in memory 302 and executed by processor 301 to complete this application. One or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a terminal device.

[0164] The terminal device may include, but is not limited to, processor 301 and memory 302. Those skilled in the art will understand that... Figure 3 This is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, a terminal device may also include input / output devices, network access devices, buses, etc.

[0165] The processor 301 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0166] The memory 302 can be an internal storage unit of the terminal device, such as the hard drive or RAM of the terminal device. The memory 302 can also be an external storage device of the terminal device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, the memory 302 can include both internal and external storage units of the terminal device. The memory 302 is used to store computer programs and other programs and data required by the terminal device. The memory 302 can also be used to temporarily store data that has been output or will be output.

[0167] It should be noted that, for the sake of convenience and brevity, the structure of the terminal device described above can also be referred to the specific description of the structure in the method embodiment, which will not be repeated here.

[0168] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0169] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, can implement the steps in the intelligent control method for the layered casting process described above.

[0170] This application provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to execute the steps in the intelligent control method for the layered casting process described above.

[0171] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0172] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for various specific applications, but such implementations should not be considered beyond the scope of this application.

[0173] In the embodiments provided in this application, it should be understood that the disclosed apparatus / terminal devices and methods can be implemented in other ways. For example, the apparatus / terminal device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units 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 devices or units may be electrical, mechanical, or other forms.

[0174] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0175] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0176] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0177] The embodiments described above are merely illustrative of the technical solutions of this application and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method of intelligent control of a continuous casting process, characterized in that, include: Obtain the casting parameters of the target ingot; A layered casting process scheme is generated based on the ingot parameters, and the process scheme includes one or more process execution parameters. Real-time data collection was performed during the implementation of the layered casting process. Calculate the solidification state of the current layer based on the on-site data; Assess the macrosegregation risk of the current and subsequent layers based on the solidification state; Based on the solidification state and the macrosegregation risk, the process execution parameters for subsequent packages are dynamically modified, and a modification command is generated and sent to the actuator based on the modified parameters.

2. The intelligent control method for the layered casting process as described in claim 1, characterized in that, The step of generating a layered casting process scheme based on the ingot parameters includes: Based on the total casting mass and single-bundle capacity range in the ingot parameters, the candidate subcontracting quantity range is determined, and the target subcontracting quantity is selected based on the candidate subcontracting quantity range combined with the target segregation control requirements and equipment capacity boundaries. Based on the target number of sub-packages and the total casting mass, the mass of each package of molten metal is allocated under the constraint of satisfying the single-package capacity range; Based on the mass of each package of molten metal, the target chemical composition in the ingot parameters, and the allowable segregation range, the target composition of each package is set differently, while satisfying the overall average composition conservation constraint. Based on the target composition of each package and the safe temperature window in the ingot parameters, the target pouring temperature of each package is set; The pouring speed for each package is set based on the mass of each package of molten metal, the geometry of the mold, and the allowable rate of liquid level rise. Based on the target number of sub-packages, the mass of each package of molten metal and the heat dissipation conditions of the mold, the solidification process of each layer is predicted and the initial value of the inter-package interval is generated. Based on the solidification process and the allowable segregation range, the next pouring trigger condition and backup parameters for abnormal conditions are generated. The layered casting process scheme is generated based on the target number of sub-lots, the mass of the molten metal in each lot, the target composition of each lot, the target pouring temperature of each lot, the pouring speed of each lot, the initial value of the interval between lots, the next pouring trigger condition and the backup parameters for abnormal conditions.

3. The intelligent control method for the layered casting process as described in claim 1, characterized in that, The step of calculating the solidification state of the current layer based on the on-site data includes: The on-site data is preprocessed to obtain the current state input vector; Based on the liquid level height, mold size, and package mass in the state input vector, calculate the geometric height and total metal volume of the current layer of molten metal. Based on the liquid surface temperature, mold temperature, cooling time, and mold heat dissipation conditions in the state input vector, estimate the volume of solidified solids at the current moment; Calculate the remaining liquid phase volume fraction in the current layer based on the total metal volume and the solidified solid volume; The position of the solid-liquid interface is calculated based on the remaining liquid volume fraction and the geometric height. The solidification rate is calculated based on the change in the solid-liquid interface position over continuous sampling time, and the local temperature gradient is calculated based on the temperature difference between different temperature measurement points in the state input vector. The solidification state of the current layer is determined based on the remaining liquid volume fraction, the solid-liquid interface position, the solidification rate, and the local temperature gradient.

4. The intelligent control method for the layered casting process as described in claim 1, characterized in that, The macrosegregation risk includes the risk of local solute enrichment, interlayer compositional fluctuation, bottom negative segregation, and top positive segregation. The assessment of the macrosegregation risk of the current and subsequent layers based on the solidification state includes: Receive the solidification state and the actual component detection value of the current package and the actual component deviation of the previous package from the field data, and construct a segregation risk assessment input set; Based on the remaining liquid phase volume fraction, local temperature gradient, and solidification rate in the solidified state, combined with the actual component detection value of the current batch, the solute redistribution intensity and local solute enrichment risk of the current layer are assessed. Based on the solid-liquid interface position in the solidification state and the actual composition deviation of the previous batch, assess the interlayer metallurgical bonding stability and the risk of interlayer composition fluctuation. Based on the distribution trend of the remaining liquid phase volume fraction along the ingot height direction, assess the risk of negative segregation at the bottom and positive segregation at the top.

5. The intelligent control method for the layered casting process as described in claim 1, characterized in that, The step of dynamically adjusting the process parameters for subsequent batches based on the solidification state and the macrosegregation risk includes: The solidification state and the macrosegregation risk are received, and the actual component detection values ​​of the previous batch in the field data are combined to determine the current deviation type, which includes at least one of temperature deviation, composition deviation and solidification process deviation. When the deviation type is composition deviation, the amount of alloy element to be added in subsequent batches is calculated based on the difference between the actual composition detection value of the previous batch and the target composition, and the target composition in the process execution parameters of the subsequent batches is compensated and corrected online based on the amount of alloy element to be added. When the deviation type is temperature deviation or solidification process deviation, the current layer is judged to meet the metallurgical bonding requirements and segregation control requirements based on the remaining liquid phase volume fraction in the solidification state, the solid-liquid interface position and the macrosegregation risk index. If the current layer does not meet the metallurgical bonding requirements and the segregation control requirements, the pouring start time, target pouring temperature, or pouring speed in the process execution parameters of the subsequent batches will be adjusted.

6. The intelligent control method for the layered casting process as described in claim 5, characterized in that, The adjustment of the pouring start time, target pouring temperature, or pouring speed in the process execution parameters of the subsequent batches includes: Based on the remaining liquid phase volume fraction in the solidification state, the solid-liquid interface position, and the macrosegregation risk index, determine whether the current layer meets the preset metallurgical bonding window and segregation control window. If the remaining liquid volume fraction is higher than the upper limit of the preset range, or the solid-liquid interface position has not yet reached the preset trigger interval, or the segregation risk index is higher than the upper limit of the preset range, then it is determined that the current layer does not meet the metallurgical bonding window and segregation control window, and the waiting time correction amount is calculated based on the difference between the remaining liquid volume fraction and the upper limit of the preset range, and the difference between the segregation risk index and the upper limit of the preset range. If the remaining liquid volume fraction is lower than the preset lower limit, or the solid-liquid interface position has exceeded the preset trigger interval upper limit, or the segregation risk index is lower than the preset lower limit, then it is determined that the current layer has the risk of being overcooled or having insufficient interlayer bonding, and the temperature correction amount or flow correction amount is calculated based on the difference between the remaining liquid volume fraction and the preset lower limit. The pouring start time is adjusted according to the waiting time correction amount, the target pouring temperature is adjusted according to the temperature correction amount, and the pouring speed is adjusted according to the flow rate correction amount.

7. The intelligent control method for the layered casting process as described in claim 1, characterized in that, The method further includes: The composition distribution, macrosegregation, and microstructure uniformity of the ingot are obtained, and the results are compared with the casting parameters and the layered casting process scheme to obtain the process deviation analysis results. Based on the process deviation analysis results, key factors affecting segregation control and their weights are extracted and updated to the process database. The key factors include at least one of the following: number of sub-packages, quality of a single package, target composition of each package, target pouring temperature of each package, pouring speed of each package, initial value of interval between packages, and triggering condition for pouring the next package.

8. An intelligent control device for a layered casting process, characterized in that, The device includes: The acquisition module is used to acquire the casting parameters of the target ingot; A generation module is used to generate a layered casting process plan based on the ingot parameters, wherein the process plan includes one or more process execution parameters. The data acquisition module is used to collect on-site data in real time during the implementation of the layered casting process. The calculation module is used to calculate the solidification state of the current layer based on the field data; The evaluation module is used to assess the macrosegregation risk of the current layer and subsequent layers based on the solidification state. The correction module is used to dynamically correct the process execution parameters of subsequent packages based on the solidification state and the macrosegregation risk, and generate correction instructions based on the corrected parameters and send them to the actuator.

9. A terminal device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent control method for the layered casting process as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent control method for the layered casting process as described in any one of claims 1 to 7.

Images