A high-efficiency casting process based on high-strength casting cup outer frame

Abstract

This invention relates to the field of casting technology, and more particularly to an efficient casting process based on a high-strength cast pouring cup outer frame. The casting process includes acquiring process state parameters at detection points, obtaining outer frame thermal field parameters and interface pressure field parameters, fusing them to generate a comprehensive pouring feature vector, inputting it into a process state recognition model, determining the current process state category and outputting the category confidence score, thereby obtaining a process category selection set and an accurate recognition rate to determine whether the recognition result meets the standard; selecting the corresponding process state category from the selection set and determining the corresponding compensation degree, thereby performing compensation control operations; finally, determining the control efficiency index based on the re-acquired process state parameters after the compensation control operation, and determining whether to adjust the preset compensation degree until the control effect meets the standard. This invention solves the problem of isolated process parameters and inability to adaptively adjust in existing technologies, effectively improving the demolding success rate and reducing the casting defect rate.

Description

Technical Field

[0001] This invention relates to the field of casting technology, and in particular to a high-efficiency casting process based on a high-strength cast pouring cup outer frame. Background Technology

[0002] In the casting process, the pouring cup, as a key component that receives molten metal, directly affects demolding efficiency and casting quality through its outer frame structure. With the development of casting automation technology, automatic pouring devices have been gradually applied in production practice, effectively improving pouring efficiency and operational safety by replacing manual pouring with mechanized and automated methods.

[0003] Chinese Patent Application Publication No. CN111940711A discloses an automatic casting device, including a safety sliding mechanism, a weighing frame, a rotating seat, a casting bucket, a connecting rod lifter, a material hopper, and a visual inspection camera. The safety sliding mechanism includes a railcar, a lifting rail, and a limit lock. The railcar is connected to a first limit post on a fixed frame via rollers, enabling automatic movement along the mold arrangement direction. The working end of the visual inspection camera is vertically downward-facing at the bottom of the railcar, used to detect the position of the casting opening at the top of the mold and guide the railcar to align. The weighing frame is located at the top of the railcar, used to detect changes in the weight of the casting bucket. When the railcar moves directly above the casting opening of the mold, the limit locker locks the traveling rollers, and the connecting rod lifter raises one side of the bottom of the casting bucket to tilt it, injecting molten steel into the mold through the material hopper. The weighing frame monitors the casting weight in real time. Once the preset weight is reached, the connecting rod lifter resets and releases the lock, allowing the railcar to continue moving to the next mold for continuous casting. The relevant technical solutions have automated the pouring process through visual positioning, weight feedback, and mechanical locking, solving the problems of low efficiency and poor safety associated with manual pouring. However, the aforementioned existing technologies still have the following problems:

[0004] Existing technical solutions only detect the location of the pouring gate and the pouring weight, without addressing the thermodynamic and mechanical parameters of the interaction between the molten metal and the outer frame of the pouring cup. This makes it difficult to identify problems such as thermal shock, turbulence, and interface adhesion during the pouring process in real time. Their control relies on a pre-programmed, open-loop approach with simple weight feedback, which cannot adaptively adjust pouring parameters according to real-time process conditions, resulting in insufficient adaptability to process fluctuations. Furthermore, these solutions only focus on the pouring process itself, failing to establish parameter correlations with subsequent processes such as molding, solidification, and demolding, thus failing to achieve full-process collaborative optimization from pouring to demolding. They also lack self-optimization capabilities, with fixed pre-programmed parameters that cannot be dynamically adjusted based on the actual production process's control effects, making it difficult to maintain stable and reliable control performance when production conditions change.

[0005] Therefore, there is an urgent need for an intelligent casting process that can achieve cross-process collaborative control based on real-time detection of the outer frame of the casting cup and has adaptive compensation capabilities, in order to solve the problems of difficult demolding, isolated processes, and inability to adaptively adjust in the existing technology. Summary of the Invention

[0006] Therefore, this invention provides an efficient casting process based on a high-strength cast pouring cup outer frame to overcome the problems in the prior art, such as difficulty in demolding the pouring cup outer frame, isolated process parameters, inability to adaptively adjust according to the process state, and reliance on manual sand vibration for demolding, which leads to easy deformation of the outer frame.

[0007] To achieve the above objectives, the present invention provides an efficient casting process based on a high-strength cast pouring cup outer frame, comprising:

[0008] The process state parameters corresponding to the detection points of the outer frame of the pouring cup, the detection point of the sprue, and the detection point of the casting cavity during the pouring process are obtained and feature extraction is performed to obtain the thermal field parameters of the outer frame and the interface pressure field parameters.

[0009] Based on several outer frame thermal field parameters and several interface pressure field parameters, several corresponding comprehensive casting feature vectors are determined.

[0010] Several casting comprehensive feature vectors are input into the process state recognition model to output several current process state categories and determine the corresponding state category confidence. The process state categories include normal state, casting impact abnormal state, and shrinkage compensation mismatch state.

[0011] The confidence level of the stated state category is matched with the confidence level of a preset state category to output the corresponding process category screening set and determine the corresponding accurate recognition rate;

[0012] Based on the comparison result between the accurate recognition rate and the preset accurate recognition rate, it is determined whether it is necessary to reacquire the process state parameters;

[0013] Based on the process category filter set, the corresponding process state category is selected, and the outer frame temperature compensation deviation or interface pressure compensation deviation is calculated.

[0014] The corresponding compensation degree is determined based on the outer frame temperature compensation deviation or the interface pressure compensation deviation.

[0015] Based on the difference between the compensation degree and the corresponding preset compensation degree, the corresponding compensation control operation is determined and executed. The compensation control operation includes adjusting the pouring speed and starting the pre-demolding vibration.

[0016] The control efficiency index is determined based on the process state parameters re-acquired after performing the corresponding compensation control operation, and compared with the preset efficiency index to determine whether to adjust the preset compensation degree.

[0017] Furthermore, the process of performing feature extraction to obtain the outer frame thermal field parameters and interface pressure field parameters includes:

[0018] The average values ​​of the inner wall temperatures of the outer frame of several pouring cups and the inner wall temperatures of several straight pouring channels are calculated based on the data obtained within a preset time period to obtain the average inner wall temperature of the outer frame and the average inner wall temperature of the straight pouring channel. The difference is then calculated to obtain the temperature difference of heat loss.

[0019] The average temperature of the inner wall of the outer frame and the temperature difference due to heat loss are processed and weighted to obtain the corresponding thermal field parameters of the outer frame.

[0020] The average value of the interface pressure is obtained by averaging the pressures on the inner walls of several casting cavities acquired within a preset time period.

[0021] The variance of the pressure fluctuation amplitude is calculated based on the pressure on the inner wall of the outer frame of several pouring cups obtained within a preset time.

[0022] The average interface pressure and the amplitude of the pressure fluctuation are processed separately and weighted to obtain the corresponding interface pressure field parameters.

[0023] Among them, the outer frame thermal field parameters include the average temperature of the inner wall of the outer frame and the temperature difference due to heat loss, and the interface pressure field parameters include the average interface pressure and the pressure fluctuation amplitude.

[0024] Furthermore, the process of determining the corresponding plurality of casting comprehensive feature vectors includes:

[0025] The outer frame thermal field parameters and the interface pressure field parameters within the adjustment period are aligned according to the time dimension to construct several corresponding casting comprehensive feature vectors, wherein the adjustment period includes several preset times.

[0026] Furthermore, the process of determining the corresponding accurate recognition rate includes:

[0027] The process state categories whose confidence scores are greater than the preset confidence scores are statistically analyzed to obtain the process category filter set.

[0028] The accuracy rate is determined by calculating the ratio between the number of process state categories in the process category filter set and the total number of process state categories output by the process state recognition model.

[0029] Based on the comparison results where the accuracy recognition rate is less than or equal to the preset accuracy recognition rate, the process state parameters are determined to be reacquired.

[0030] Furthermore, the process of determining the corresponding compensation degree based on the outer frame temperature compensation deviation or the interface pressure compensation deviation includes:

[0031] The process state category with the highest confidence level is selected from the process category filter set, and the outer frame temperature compensation deviation or the interface pressure compensation deviation is calculated based on the selected process state category.

[0032] When the process state category is the abnormal state of casting impact, the outer frame temperature compensation deviation is calculated.

[0033] When the process state category is the shrinkage compensation mismatch state, the interface pressure compensation deviation is calculated.

[0034] Furthermore, the temperature difference between the inner wall of the outer frame and the heat loss temperature difference are respectively standardized and calculated with respect to the corresponding preset temperature thresholds to obtain the temperature difference between the inner wall of the outer frame and the temperature difference deviation of the heat loss.

[0035] The outer frame temperature compensation deviation is obtained by weighted calculation based on the temperature difference between the inner wall of the outer frame and the temperature difference deviation due to heat loss.

[0036] The corresponding temperature compensation degree is determined based on the correspondence between the outer frame temperature compensation deviation and the compensation degree.

[0037] The interface pressure deviation and pressure fluctuation deviation are calculated by standardizing the average interface pressure and the pressure fluctuation amplitude with the corresponding preset pressure threshold, respectively.

[0038] The interface pressure compensation deviation is obtained by weighted calculation based on the interface pressure deviation and the pressure fluctuation deviation.

[0039] The corresponding pressure compensation degree is determined based on the correspondence between the interface pressure compensation deviation and the compensation degree.

[0040] Furthermore, the process of determining and executing the corresponding compensation and control operation includes:

[0041] When the process state category is the abnormal state of pouring impact, the pouring speed is adjusted and the adjustment range is positively correlated with the temperature compensation degree difference. The temperature compensation degree difference is obtained by calculating the difference between the temperature compensation degree and the preset temperature compensation degree.

[0042] When the process state category is the shrinkage compensation mismatch state, the pre-demolding vibration is started and the vibration amplitude is positively correlated with the pressure compensation degree difference. The pressure compensation degree difference is obtained by calculating the difference between the pressure compensation degree and the preset pressure compensation degree.

[0043] Furthermore, the process of determining the regulation efficiency index includes:

[0044] Re-collect the process status parameters corresponding to the detection points of the outer frame of the pouring cup, the sprue, and the casting cavity after the compensation and control operation is performed;

[0045] Calculate the outer frame temperature compensation deviation after the compensation and control operation and record it as the outer frame temperature adjustment deviation;

[0046] The temperature control efficiency index is obtained by calculating the absolute value of the ratio between the outer frame temperature compensation deviation and the outer frame temperature adjustment deviation.

[0047] Alternatively, calculate the interface pressure compensation deviation after the compensation and control operation and record it as the interface pressure regulation deviation;

[0048] The pressure regulation efficiency index is obtained by calculating the absolute value of the ratio between the interface pressure compensation deviation and the interface pressure regulation deviation.

[0049] Furthermore, the process of determining whether to adjust the preset compensation degree includes:

[0050] Based on the comparison results of the temperature control efficiency index being less than or equal to the preset temperature efficiency index, or the pressure control efficiency index being less than or equal to the preset pressure efficiency index, the preset compensation degree is determined to be adjusted based on the temperature control index difference or the pressure control index difference.

[0051] Wherein, the temperature control index difference is the difference between the preset temperature efficiency index and the temperature control efficiency index, and the pressure control index difference is the difference between the preset pressure efficiency index and the pressure control efficiency index.

[0052] Furthermore, the pouring speed or the vibration amplitude is corrected based on the adjusted preset compensation degree to obtain a cooled casting.

[0053] Compared with the prior art, the beneficial effects of this invention are as follows: by acquiring the process state parameters corresponding to the detection points of the outer frame of the pouring cup, the sprue, and the casting cavity, multi-point real-time perception of the pouring process is achieved, providing precise input for subsequent dynamic control based on the actual process state by obtaining quantitative indicators characterizing the dynamic changes of the pouring process; by fusing several outer frame thermal field parameters and several interface pressure field parameters to generate several comprehensive pouring feature vectors, the information in the temperature and pressure dimensions complement each other, forming a complete description of the pouring process state; and by inputting several comprehensive pouring feature vectors into the process state identification... The model identifies several current process state categories and outputs corresponding state category confidence scores, enabling automatic identification of abnormal casting impact states and shrinkage compensation misalignment states, providing a basis for differentiated compensation. By matching state category confidence scores with preset state category confidence scores to determine a process category filter set, and then determining the accurate identification rate based on this filter set, reliable identification results are selected using a confidence threshold to avoid low-confidence identification results misleading subsequent adjustments. The comparison between the accurate identification rate and a preset accurate identification rate determines whether process state parameters need to be reacquired; automatic triggering occurs when the accurate identification rate fails to meet the target. Data re-sampling establishes a data quality self-inspection mechanism to ensure the validity of input data for control decisions. By selecting corresponding process state categories based on a defined process category filter set and calculating the outer frame temperature compensation deviation or interface pressure compensation deviation, targeted compensation deviation calculations are implemented for different abnormal states, ensuring that the compensation deviation reflects the deviation information of the current process state. The corresponding compensation degree is determined based on the outer frame temperature compensation deviation or interface pressure compensation deviation, and the corresponding compensation control operation is executed based on the difference between the compensation degree and the corresponding preset compensation degree, achieving dynamic matching between the control intensity and the actual deviation degree. Furthermore, based on the execution... After performing corresponding compensation and control operations, the process state parameters reacquired are used to determine the control efficiency index. By comparing the changes in process state parameters before and after control, the control effect is quantified, providing a quantitative indicator for evaluating the control effect. By comparing the control efficiency index with the preset efficiency index, it is determined whether to adjust the preset compensation degree, forming a self-optimizing closed loop. Ultimately, real-time perception, state recognition, precise compensation, and self-optimizing adjustment of the casting process are realized. This solves the problems of difficult demolding of the outer frame of the casting cup, isolated process parameters, and inability to adaptively adjust in the existing technology, effectively improving the demolding success rate, extending the service life of the outer frame, and reducing the casting defect rate.

[0054] Furthermore, this invention calculates the average inner wall temperature of the outer frame and the average inner wall temperature of the sprue based on several inner wall temperatures of the outer frame and the sprue obtained within a preset time period. The difference is then used to calculate the heat loss temperature difference, which characterizes the degree of temperature loss of the molten metal as it flows from the outer frame of the casting cup through the sprue. Simultaneously, the average interface pressure is calculated based on several inner wall pressures of the casting cavity obtained within a preset time period, and the variance is calculated based on several inner wall pressures of the outer frame of the casting cup to obtain the pressure fluctuation amplitude. The average interface pressure reflects the average compressive stress level at the interface between the sand mold and the outer frame, while the pressure fluctuation amplitude reflects the stability of the molten metal impact. Then, through normalization and weighted calculation, the outer frame thermal field parameters and interface pressure field parameters are obtained, achieving dimensionality reduction and standardization of the original data. Finally, several outer frame thermal field parameters and interface pressure field parameters within the adjustment period are aligned according to the time dimension to construct a comprehensive casting feature vector. This allows the feature vector to reflect the continuous changing trend of the process state, avoiding misjudgments caused by data fluctuations at a single time point, and providing high-quality input features for subsequent state identification and compensation calculations.

[0055] Furthermore, this invention also compiles a process category screening set by statistically analyzing process state categories whose confidence levels are greater than a preset confidence level. Low-confidence identification results are then filtered out using a confidence threshold, ensuring high reliability of the state categories participating in subsequent control. The accurate identification rate is then determined based on the ratio of the number of state categories in the process category screening set to the total number of outputs from the identification model. When the accurate identification rate fails to meet the standard, the process state parameters are automatically reacquired, forming a self-checking and re-acquisition mechanism for the quality of the identification results. Based on this, the process state category corresponding to the highest confidence level is selected from the process category screening set, and the outer frame temperature compensation deviation or interface pressure compensation deviation is calculated accordingly, achieving targeted deviation calculation for different abnormal states.

[0056] Furthermore, this invention calculates the temperature difference of the inner wall of the outer frame by standardizing the deviation between the average temperature of the inner wall and the preset ideal value, and calculates the temperature difference deviation of the heat loss by standardizing the deviation between the heat loss temperature difference and the preset ideal value. The standardization deviation includes both directional and amplitude information. Then, the temperature compensation deviation of the outer frame is obtained by weighted summing of the temperature difference of the inner wall and the temperature difference deviation of the heat loss temperature difference, so that the deviation comprehensively reflects the deviation information of both the average temperature and the temperature difference of the heat loss temperature difference. Similarly, the interface pressure deviation is calculated by standardizing the deviation between the average interface pressure and the preset ideal value, and the pressure fluctuation deviation is calculated by standardizing the deviation between the pressure fluctuation amplitude and the preset ideal value. The interface pressure compensation deviation is obtained by weighted summing. Finally, the temperature compensation degree or pressure compensation degree is determined based on the correspondence between the compensation deviation and the compensation degree, so that the compensation degree can quantitatively reflect the degree of deviation and provide a precise basis for subsequent control.

[0057] Furthermore, this invention also involves adjusting the pouring speed when the process state is an abnormal pouring impact state, with the adjustment range being positively correlated with the temperature compensation degree; and initiating a pre-demolding vibration operation when the process state is a shrinkage compensation mismatch state, with the vibration amplitude being positively correlated with the pressure compensation degree, thus matching the adjustment intensity with the compensation degree. After performing the compensation adjustment operation, the process state parameters are re-acquired, the adjusted outer frame temperature compensation deviation or interface pressure compensation deviation is calculated, and the absolute value of the ratio is calculated with the compensation deviation before adjustment to obtain the temperature control efficiency index or pressure control efficiency index. This quantifies the degree of attenuation of the deviation after adjustment relative to the deviation before adjustment, and achieves a quantitative evaluation of the control effect.

[0058] Furthermore, this invention also determines the need to adjust the preset compensation degree based on the comparison results of the temperature control efficiency index being less than or equal to the preset temperature efficiency index or the pressure control efficiency index being less than or equal to the preset pressure efficiency index. That is, when the control effect is not up to standard, the compensation degree adjustment is triggered. Then, the speed adjustment amount or vibration adjustment amount is determined based on the difference between the preset efficiency index and the control efficiency index, so that the adjustment amount matches the difference between the control effect. Finally, the adjusted pouring speed is determined based on the speed adjustment amount, and the adjusted vibration amplitude is determined based on the vibration adjustment amount, and the cooled casting is obtained in this way, thus completing a complete closed-loop control process from detection, identification, compensation, evaluation to self-optimization. Attached Figure Description

[0059] Figure 1 This is a flowchart of the efficient casting process based on a high-strength cast pouring cup outer frame in an embodiment of the present invention;

[0060] Figure 2 This is a logic decision diagram for determining the corresponding accurate recognition rate in an embodiment of the present invention;

[0061] Figure 3 This is a logic decision diagram for determining the corresponding compensation degree in an embodiment of the present invention;

[0062] Figure 4 This is a logic diagram for determining whether to adjust the preset compensation degree in an embodiment of the present invention. Detailed Implementation

[0063] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0064] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0065] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0066] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0067] Please see Figure 1 The diagram shows a flowchart of an efficient casting process for a high-strength cast pouring cup frame according to an embodiment of the present invention. The process includes at least the following steps:

[0068] S1: Obtain the process state parameters corresponding to the detection points of the outer frame of the pouring cup, the direct sprue, and the casting cavity during the pouring process, and extract features to obtain the outer frame thermal field parameters and interface pressure field parameters. The outer frame thermal field parameters include the average temperature of the inner wall of the outer frame and the temperature difference due to heat loss, and the interface pressure field parameters include the average interface pressure and the pressure fluctuation amplitude.

[0069] S2: Determine several corresponding comprehensive feature vectors for casting based on several outer frame thermal field parameters and several interface pressure field parameters;

[0070] S3: Input several casting comprehensive feature vectors into the process state recognition model to output several current process state categories and determine the corresponding state category confidence. The process state categories include normal state, casting impact abnormal state, and shrinkage compensation mismatch state.

[0071] S4: Match the state category confidence score with the preset state category confidence score to output the corresponding process category screening set and determine the corresponding accurate recognition rate;

[0072] S5: Determine whether it is necessary to reacquire process status parameters based on the comparison results between the accurate recognition rate and the preset accurate recognition rate;

[0073] S6: Select the corresponding process state category based on the process category filter set, and calculate the outer frame temperature compensation deviation or interface pressure compensation deviation.

[0074] S7: Determine the corresponding compensation degree based on the outer frame temperature compensation deviation or interface pressure compensation deviation;

[0075] S8: Based on the difference between the compensation degree and the corresponding preset compensation degree, determine the corresponding compensation control operation to be executed. The compensation control operation includes adjusting the pouring speed and starting the pre-demolding vibration.

[0076] S9: Determine the control efficiency index based on the process state parameters re-acquired after performing the corresponding compensation control operation, and compare it with the preset efficiency index to determine whether to adjust the preset compensation degree.

[0077] In the efficient casting process based on a high-strength cast pouring cup frame provided in this embodiment, the high-strength cast pouring cup frame is used as a sensing platform. The process state parameters of the pouring process are obtained through multi-point real-time detection. Based on state recognition and closed-loop compensation, adaptive control is achieved, which solves the problem of isolated process parameters and inability to adaptively adjust them in the prior art. This effectively improves the demolding success rate and reduces the casting defect rate.

[0078] In this embodiment, the outer frame of the high-strength casting cup is made of a 40mm thick steel plate with a 7° draft angle from top to bottom. The outer frame of the casting cup has an octagonal structure without sharp corners, and M22 nuts are welded to the side walls and equipped with M22×100 bolts. The bolts are fully screwed in during the molding stage, with their heads extending into the inner wall of the outer frame to anchor the resin sand mold and prevent it from turning over and falling off. During the demolding stage, the bolts can be controlled to exit, assisting in the separation of the sand mold from the outer frame.

[0079] Understandably, the 40mm thick steel plate, with a mass much greater than that of a conventional outer frame (usually 10-20mm), significantly reduces the heating rate and makes the temperature change more gradual during the molten metal pouring process. At the same time, the fluctuation range of the heat loss temperature difference between the outer frame and the sprue is also reduced accordingly. This makes the temperature change of the inner wall of the outer frame exhibit a predictable, slow evolution pattern due to thermal inertia, rather than an instantaneous change. This provides a physical basis for the stable extraction of subsequent temperature field characteristics (average temperature, heat loss temperature difference). The more gradual the temperature change, the lower the noise of the data collected by the sensor and the more stable the characteristic values, thus ensuring the signal-to-noise ratio of the thermal field characteristics.

[0080] In this embodiment, a 7° slope is used to create a tapered structure on the inner wall of the outer frame, narrower at the top and wider at the bottom. When the sand mold shrinks after casting, this slope transforms the compressive stress perpendicular to the inner wall into a component force along the slope direction, guiding the sand mold to displace towards the center of the outer frame, thereby reducing the initial resistance during demolding. The octagonal cross-section design without sharp corners allows for a more uniform circumferential stress distribution compared to the traditional rectangular cross-section, preventing stress concentration at the corners. Furthermore, the stress-directing characteristics ensure a repeatable regularity in the interface pressure distribution, providing structural assurance for the accurate measurement of interface pressure field parameters (average pressure, pressure fluctuation amplitude). The more uniform the stress distribution, the more representative the data collected by the pressure sensor is of the overall interface state, rather than local anomalies.

[0081] In this embodiment, four K-type thermocouples are evenly distributed along the circumference of the inner wall of the outer frame of the casting cup. The measuring ends of the thermocouples are embedded in shallow grooves in the inner wall of the outer frame and are flush with the inner wall surface. The inner wall of the outer frame is in direct contact with the resin sand mold. When the molten iron is poured, the heat is conducted to the inner wall of the outer frame through the sand mold. The thermocouples measure the temperature of the inner wall of the outer frame, thereby indirectly reflecting the thermal state of the molten metal. Eight thin-film pressure sensors are pre-embedded at the center of each face of the octagon on the inner wall of the outer frame. The sensitive surface of the sensors is flush with the inner wall of the outer frame. During molding, the sand mold directly covers the surface of the sensors. The pressure generated by the molten iron impacting the sand mold is transmitted to the sensors through the sand mold, thereby measuring the interface pressure. A K-type thermocouple is pre-embedded in the upper part of the sprue. The thermocouple is encapsulated in a high-temperature resistant glass protective tube. The protective tube is pre-embedded in the sand mold wall of the sprue, and its end is flush with the inner wall of the sprue. When the molten iron flows through the sprue, the heat is conducted to the thermocouple through the sand mold and the glass protective tube, thereby indirectly measuring the temperature of the inner wall of the sprue.

[0082] Understandably, as the molten metal flows from the outer frame of the pouring cup into the sprue, heat is conducted to the sand mold and the outer frame, causing the temperature to gradually decrease. Therefore, it is necessary to calculate the thermal field parameters of the outer frame to obtain a quantitative indicator characterizing the dynamic changes in heat during the pouring process, providing accurate input for subsequent dynamic control based on the actual process conditions.

[0083] In one specific embodiment, the system acquires data from each sensor during the pouring process and performs feature extraction within a preset time window Q, with Q exemplarily set to 10 seconds.

[0084] Specifically, the average value of the temperature data of the inner wall of the outer frame collected by the four thermocouples within the current preset time window Q is calculated to obtain the average temperature of the inner wall of the outer frame Tf; at the same time, the average value of the temperature data collected by the thermocouples above the sprue within the current preset time window Q is calculated to obtain the average temperature of the inner wall of the sprue Tg.

[0085] Then, the heat loss temperature difference R = Tf - Tg is calculated, where the heat loss temperature difference R characterizes the degree of temperature loss of the molten metal in the inlet section of the casting system. The larger the R value, the greater the heat loss and the worse the fluidity of the molten metal.

[0086] In this embodiment, the average temperature Tf of the inner wall of the outer frame and the temperature difference R of heat loss are normalized to eliminate the difference in dimensions, and the thermal field parameter Hf of the outer frame is obtained by further weighted calculation.

[0087] The specific formula for normalizing the average temperature Tf of the inner wall of the outer frame is: Tfno=(Tf-Tmin) / (Tmax-Tmin); the specific formula for normalizing the temperature difference R of heat loss is: Rno=(R-Rmin) / (Rmax-Rmin).

[0088] Wherein, Tmin, Tmax, Rmin, and Rmax represent the minimum, maximum, minimum, and maximum temperature extremes, respectively, for historical normal batches. Specifically, they are determined as follows: From the historical database of the production management system, all production records showing that the castings meet quality standards (i.e., subsequent inspections show that dimensional deviations, internal defects, and surface quality all meet process standards) are selected. The average temperature Tf of the inner wall of the outer frame and the heat loss temperature difference R within the corresponding time window for each batch are extracted. The numerical distributions of Tf and R are then statistically analyzed, with the lower boundary of the distribution taken as Tmin and Rmin, and the upper boundary taken as Tmax and Rmax. For example, Tmin = 1250℃, Tmax = 1550℃, Rmin = 30℃, and Rmax = 70℃ can be used.

[0089] The normalized Tfno and Rno are weighted and summed to obtain the outer frame thermal field parameter Hf: Hf = w1·Tfno + w2·Rno, where the weights w1 and w2 reflect the influence of the mean temperature and the heat loss temperature difference on subsequent state identification, respectively. They are determined as follows: qualified batches are selected from the historical database, and the correlation coefficients between the mean temperature and the heat loss temperature difference of each batch and the final casting quality indicators (such as dimensional deviations and surface defects) are calculated. The normalized value of the correlation coefficient is taken as the weight. Since the mean temperature of the inner wall of the outer frame directly reflects the heat input intensity of the molten metal and is the dominant factor determining the thermal expansion of the outer frame and the heating state of the sand mold, its impact on casting quality is more direct; while the heat loss temperature difference reflects the additional heat loss of the molten metal during the flow process, its impact on casting quality is relatively indirect. Therefore, the weight coefficient of the mean temperature is greater than the weight coefficient of the heat loss temperature difference. For example, w1 = 0.6 and w2 = 0.4 can be taken.

[0090] Understandably, during the casting process, the molten metal flowing within the pouring cup exerts dynamic pressure on the interface between the inner wall of the pouring cup frame and the sand mold. This interface pressure includes the average compressive stress level Pf, reflecting the overall fit between the sand mold and the outer frame, and the pressure fluctuation amplitude σP, reflecting the stability (laminar or turbulent) of the molten metal flow. Therefore, it is necessary to calculate the interface pressure field parameters to obtain quantitative indicators characterizing the dynamic changes in pressure during the casting process, providing accurate input for subsequent dynamic control based on actual process conditions.

[0091] In this embodiment, the average value of the interface pressure data collected by eight thin-film pressure sensors within a preset time window Q is calculated to obtain the average interface pressure Pf, which reflects the average compressive stress level at the interface between the sand mold and the outer frame; and the variance of the eight interface pressure data within the preset time window Q is calculated to obtain the pressure fluctuation amplitude σP. σP characterizes the stability of the molten metal impact. The larger the σP, the more unstable the impact and the higher the risk of turbulence.

[0092] Furthermore, the mean interfacial pressure Pf and the pressure fluctuation amplitude σP are normalized and then weighted and summed to obtain the interfacial pressure field parameter Hp.

[0093] The specific formula for normalizing the mean interface pressure Pfno is: Pfno = (Pf - Pmin) / (Pmax - Pmin); the specific formula for normalizing the pressure fluctuation amplitude σP is: σPno = (σP - σmin) / (σmax - σmin).

[0094] Wherein, Pmin and Pmax represent the minimum and maximum extreme values ​​of interface pressure in historical normal batches, respectively, and σmin and σmax represent the minimum and maximum extreme values ​​of pressure fluctuation amplitude in historical normal batches, respectively. Specifically, the determination method is as follows: From the historical database of the production management system, all production records of castings that meet quality standards are selected. The mean interface pressure Pf and pressure fluctuation amplitude σP within the corresponding time window of each batch are extracted. The numerical distributions of Pf and σP are statistically analyzed, and the lower boundary of the distribution is taken as Pmin and σmin, and the upper boundary of the distribution is taken as Pmax and σmax. For example, Pmin = 0.3 MPa, Pmax = 0.8 MPa, and σmin = 0.02 MPa can be chosen. 2 σmax = 0.08 MPa 2 .

[0095] The normalized Pfno and σPno are weighted and summed to obtain the interface pressure field parameter Hp: Hp = w3·Pfno + w4·σPno, where the weights w3 and w4 are determined based on historical production data statistics. Specifically, qualified batches are selected from the historical database, and the correlation coefficients between Pfno and σPno in each batch and the final casting quality indicators (such as surface defect score and demolding success rate) are calculated. The normalized value of the correlation coefficient is used as the weight, giving higher weight to features with a greater impact on quality during the fusion process. For example, if historical data shows that the average interface pressure and the pressure fluctuation amplitude have roughly the same impact on demolding adhesion, w3 = 0.5 and w4 = 0.5 can be set accordingly.

[0096] Understandably, the casting process is a typical thermo-mechanical coupling process. The intensity of the heat input from the molten metal directly affects the interfacial pressure distribution, and fluctuations in the interfacial pressure affect the heat transfer efficiency. At the same time, the casting process has temporal continuity. Characteristic values ​​at a single point in time (such as the outer frame thermal field parameters at a certain moment) may be distorted due to instantaneous disturbances (such as molten metal splashing or sensor noise), and cannot represent the true level of the process state.

[0097] The evolution of the process state (such as whether the heat input intensity continues to rise or tends to stabilize) is crucial for risk prediction. Even if the current thermal field characteristic value does not exceed the limit, if it shows an upward trend for multiple consecutive time windows, it indicates that the risk of thermal shock is accumulating and early intervention is required.

[0098] Therefore, it is necessary to calculate the casting composite eigenvector to obtain a quantitative value that characterizes the intrinsic physical coupling relationship between the thermal field characteristics and the pressure field characteristics.

[0099] In this embodiment, the adjustment period t is exemplarily set to 30 seconds, which includes 3 preset time windows Q. Several consecutive outer frame thermal field parameters and interface pressure field parameters within the adjustment period t are aligned according to the time dimension to construct a casting comprehensive feature vector V: V=[Hf(t1),Hf(t2),Hf(t3),Hp(t1),Hp(t2),Hp(t3)], where V reflects the continuous change trend of the process state within the adjustment period.

[0100] In this embodiment, the comprehensive feature vector V of casting is input into the pre-constructed process state identification model. The process state identification model is constructed using the K-means clustering algorithm. The specific construction method is as follows: extract the comprehensive feature vector of casting of qualified batches from the historical production database, set the number of clusters K=3, corresponding to three typical working conditions: normal state, abnormal casting impact state, and shrinkage compensation imbalance state.

[0101] The distance from each sample to the cluster center is calculated iteratively using the K-means clustering algorithm, and finally converges to obtain three cluster centers Ck, which are denoted as C1, C2 and C3 respectively.

[0102] After obtaining three cluster centers C1, C2, and C3 through K-means clustering, each cluster center needs to be assigned a specific process status category label. The specific method is as follows: extract several typical sample batches belonging to each cluster center from the historical production database, and analyze the operation logs, sensor trends, and final casting quality reports recorded during the actual production process for these batches.

[0103] If, in a sample batch corresponding to a certain cluster center, the average temperature of the inner wall of the outer frame is significantly higher than the normal range and the heat loss temperature difference is slower, and the casting has hot cracks or sand inclusion defects, then the cluster center is marked as "pouring impact abnormal state".

[0104] If the average interface pressure is high and the pressure fluctuation amplitude is large in the sample batch corresponding to a certain cluster center, and obvious jamming or adhesion occurs during demolding, then the cluster center is marked as "shrinkage compensation imbalance state".

[0105] If the process parameters are stable and the casting quality is qualified in the sample batch corresponding to a certain cluster center, then the cluster center is marked as "normal state".

[0106] In one specific embodiment, based on the established mapping relationship between cluster centers and process state categories, the Euclidean distance from the current casting comprehensive feature vector V to the three cluster centers is calculated, and the cluster center corresponding to the minimum distance is taken. Its pre-labeled process state category label is the current process state category. The specific formula for calculating the Euclidean distance is: dk=‖V-Ck‖, where k∈{1,2,3}.

[0107] The state category corresponding to the minimum distance is taken as the current process state category, and the reciprocal of the minimum distance is normalized and used as the state category confidence D (the smaller the distance, the higher the confidence). The closer the D value is to 1, the more reliable the recognition result is.

[0108] In one specific embodiment, the process state identification model takes the constructed casting comprehensive feature vector V as the unit of adjustment period t, inputs it into the process state identification model, takes the process state category marked by the cluster center Ck corresponding to the minimum distance as the process state category of the current adjustment period t; at the same time, it outputs the confidence level D of the state category.

[0109] Please see Figure 2As shown, this is a logical decision diagram for determining the corresponding accurate recognition rate in an embodiment of the present invention. The recognition results for all state categories with a confidence level D greater than Dth are statistically analyzed to obtain a process category screening set S, which includes several process state categories. The accurate recognition rate A is then calculated: A = S / N, where N is the total number of state categories output by the process state recognition model. The preset state category confidence level Dth is obtained based on historical production data. Specifically, qualified batches are selected from the historical production database, and the distribution of their state category confidence levels is statistically analyzed. The 95th percentile is used as the threshold; for example, Dth = 0.8 can be set.

[0110] Understandably, the accuracy rate A reflects the proportion of high-confidence identification results within the current adjustment cycle and is an indicator of the overall reliability of the identification results. If A is too low, it indicates that most identification results lack confidence. In this case, directly entering compensation control may lead to erroneous compensation control operations due to misjudgment of the state. Therefore, it is necessary to compare A with the preset accuracy rate Ath.

[0111] When A≤Ath, it indicates that the reliability of the current identification result is insufficient and the process status parameters need to be re-acquired. The preset accurate identification rate Ath is determined based on the statistics of historical production data. The specific method is as follows: select qualified batches from the historical production database, count their accurate identification rate distribution, and take the 90th percentile as the threshold. For example, Ath=0.9 can be set.

[0112] When A > Ath, it indicates that the current process state category output by the process state identification model is sufficiently reliable. Therefore, it is determined that the compensation and control operation stage should be entered.

[0113] Understandably, when the process condition identification model detects abnormal operating conditions, in order to clarify the direction and extent of defect compensation, it is necessary to quantify the degree to which the current operating condition deviates from the ideal state in order to achieve accurate compensation for the defects.

[0114] Please see Figure 3 As shown, this is a logic decision diagram for determining the corresponding compensation degree in an embodiment of the present invention. The process state category with the highest state category confidence degree D is selected from the process category filter set S as the most reliable process state. Specifically, if the process state category is "abnormal casting impact state," then the outer frame temperature compensation deviation is calculated; if the process state category is "shrinkage compensation misalignment state," then the interface pressure compensation deviation is calculated; if the process state category is "normal state," then no compensation is required, and the corresponding compensation degree is zero.

[0115] In this embodiment, the corresponding preset temperature threshold includes the preset average temperature of the inner wall of the outer frame, Tfth, and the preset heat loss temperature difference, Rth. The standardized deviation ZT between the average temperature of the inner wall of the outer frame, Tf, and the preset average temperature of the inner wall of the outer frame, Tfth, is calculated: ZT = (Tf - Tfth) / σT. Wherein, Tfth is determined based on the process design value, and is set to 1400℃ according to the optimal pouring temperature of the molten metal and the characteristics of the outer frame material. σT is the standard deviation of the average temperature of the inner wall of the outer frame in historical normal batches. For example, σT can be set to 30℃.

[0116] In this embodiment, the standardized deviation ZR between the heat loss temperature difference R and the preset heat loss temperature difference Rth is calculated as follows: ZR = (R - Rth) / σR; where Rth is obtained based on historical production data statistics. Specifically, from the historical database of the production management system, all production records of castings that meet quality standards are selected, the heat loss temperature difference R within the corresponding time window of each batch is extracted, the distribution of these R values ​​is statistically analyzed, and the median is taken as Rth. For example, Rth can be set to 50℃. σR is the standard deviation of the heat loss temperature difference in historical normal batches. For example, σR can be set to 10℃.

[0117] The weighted sum of ZT and ZR yields the outer frame temperature compensation deviation ΔT: ΔT = α·ZT + β·ZR, where the weights α and β are determined based on the influence of temperature deviation and heat loss temperature difference deviation on demolding difficulty in historical data, respectively. By analyzing the correlation coefficients between ZT, ZR, and demolding success rate in the product, normalized values ​​are used as weights. Since temperature deviation reflects the heat input intensity of the molten metal, it directly determines the degree of thermal expansion of the outer frame and the heating state of the sand mold, thus having a more direct impact on demolding resistance; heat loss temperature difference deviation reflects the fluidity of the molten metal, and its impact on demolding success rate is relatively indirect. For example, α = 0.6 and β = 0.4 are set.

[0118] In this embodiment, the preset pressure threshold includes a preset average interface pressure Pfth and a preset pressure fluctuation amplitude σPth. The standardized deviation ZP between the average interface pressure Pf and the preset average interface pressure Pfth is calculated as follows: ZP = (Pf - Pfth) / σPf. Pfth is determined based on historical production data. Specifically, from the historical database of the production management system, all production records showing qualified castings are selected. The average interface pressure Pf within the corresponding time window of each batch is extracted, and the distribution of these Pf values ​​is statistically analyzed. The median is taken as Pfth; for example, Pfth = 0.5 MPa. σPf is the standard deviation of the average interface pressure in historical normal batches; for example, σPf = 0.1 MPa.

[0119] In this embodiment, the standardized deviation Zσ between the pressure fluctuation amplitude σP and the preset pressure fluctuation amplitude σPth is calculated: Zσ = (σP - σPth) / σz; where σPth is determined based on historical production data statistics. Specifically, from the historical database of the production management system, all production records of castings that meet quality standards are selected, the pressure fluctuation amplitude σP within the corresponding time window of each batch is extracted, the distribution of these σP values ​​is statistically analyzed, and the median is taken as σPth. For example, σPth can be set to 0.05 MPa. 2 σz represents the standard deviation of pressure fluctuation amplitude in historical normal batches; for example, σz can be set to 0.015 MPa. 2 .

[0120] The weighted sum of ZP and Zσ yields the interface pressure compensation deviation ΔP: ΔP = γ·ZP + δ·Zσ, where the weights γ and δ are determined based on the influence of the average pressure deviation and pressure fluctuation deviation on interface adhesion in historical data, respectively. Specifically, qualified batches are selected from the historical production database, and the ZP, Zσ, and interface adhesion risk score (or demolding success rate) of each batch are extracted for multiple linear regression analysis. The absolute values ​​of the regression coefficients are normalized and used as weights. Since the average pressure deviation ZP reflects the overall compressive stress level of the interface and is directly related to the tightness of the fit between the sand mold and the outer frame, its impact on demolding adhesion is more direct; the pressure fluctuation deviation Zσ reflects the stability of the molten metal impact and its impact on interface adhesion is relatively indirect. For example, γ = 0.6 and δ = 0.4 are set.

[0121] Understandably, reducing the pouring speed decreases the heat input of the molten metal per unit time, thereby lowering the outer frame temperature and mitigating the temperature difference caused by heat loss; conversely, increasing the pouring speed increases the heat input, raises the temperature, and accelerates the temperature difference caused by heat loss. Speed ​​adjustment allows the heat input to be matched with the target value.

[0122] Specifically, the temperature compensation degree δT is determined based on the outer frame temperature compensation deviation ΔT, using the formula: δT = kT·ΔT, where kT is the temperature compensation ratio coefficient. kT is obtained based on historical production data. The specific method is as follows: batches with good compensation effects (i.e., significantly reduced deviation after adjustment) are selected from the historical production database. The |ΔT| of these batches is extracted along with the corresponding optimal speed adjustment range (the adjustment ratio based on the current pouring speed). The distribution of the ratio between the two is calculated, and the median is taken as kT. For example, kT = 0.1 can be used.

[0123] In this embodiment, when the process state category is abnormal casting impact state, the temperature compensation degree difference ΔδT = δT - δTth is calculated, where δTth is the preset temperature compensation degree corresponding to the temperature compensation degree. δTth is determined based on historical data statistics. Specifically, batches with good compensation effects are selected from the historical production database, the temperature compensation degrees of these batches are extracted, their distribution is statistically analyzed, and the median is taken as δTth. For example, δTth = 0.5 is taken.

[0124] When ΔδT>0, it indicates that the current temperature is too high or the heat loss temperature difference is too large. The pouring speed is reduced. The reduction is positively correlated with |ΔδT|. The new pouring speed is: vnew=v0·(1-|ΔδT|).

[0125] When ΔδT < 0, it indicates that the current temperature is too low or the heat loss temperature difference is too fast. The operation of increasing the pouring speed is executed. The increase is positively correlated with |ΔδT|. The new pouring speed is: vnew = v0·(1+|ΔδT|).

[0126] When ΔδT=0, it means that the current temperature compensation degree is consistent with the preset temperature compensation degree, and there is no need to adjust the pouring speed;

[0127] Wherein, v0 is the initial pouring speed, which is determined based on the process design value. Specifically, the reference pouring speed is determined by simulating the casting process using casting process simulation software based on process parameters such as casting material, pouring temperature, and mold conditions. This reference pouring speed is then used as the initial value of v0. For example, v0 = 1.5 m / s is taken.

[0128] Understandably, pre-demolding vibration breaks the adhesion and local bonding between the sand mold and the outer frame interface through high-frequency excitation. When the interface pressure is too high or fluctuates too much, vibration can effectively reduce demolding resistance. When the interface pressure is normal, vibration is stopped to avoid unnecessary energy consumption and sand mold disturbance.

[0129] Specifically, the pressure compensation degree δP is determined based on the interface pressure compensation deviation ΔP, using the formula: δP = kP·ΔP, where kP is the pressure compensation ratio coefficient. kP is obtained based on historical production data. The specific method involves selecting batches with good compensation effects (i.e., significantly reduced deviation after adjustment) from the historical production database, extracting |ΔP| from these batches and their corresponding optimal vibration amplitude (the adjustment ratio based on the maximum vibration amplitude), calculating the distribution of their ratios, and taking the median as kP. For example, kP = 0.15.

[0130] In this embodiment, when the process state is a shrinkage compensation mismatch state, the pressure compensation difference ΔδP = δP - δPth is calculated, where δPth is the preset pressure compensation degree corresponding to the pressure compensation degree. δPth is determined based on historical production data statistics. Specifically, batches with good compensation effects are selected from the historical production database, the pressure compensation degree δP of these batches is extracted, their distribution is statistically analyzed, and the median is taken as δPth. For example, δPth = 0.4 is used.

[0131] When ΔδP > 0, it indicates that the interface pressure is too high or the fluctuation is too large. The vibration device is then activated. The vibration amplitude is positively correlated with |ΔδP|, and the vibration amplitude is: Anew = A0·(1 + |ΔδP|), where A0 is the initial vibration amplitude of the vibration device. A0 is determined based on equipment parameters and process experience. Specifically, based on the maximum output capacity (maximum vibration amplitude) of the vibration device and the typical excitation intensity required for demolding, 30% of the maximum vibration amplitude is taken as the initial vibration amplitude. For example, A0 = 3mm is taken.

[0132] When ΔδP < 0, it indicates that the interface pressure is too low or the fluctuation is too small, and the vibration stops (i.e. the vibration amplitude is zero).

[0133] When ΔδP=0, it means that the current pressure compensation degree is consistent with the preset pressure compensation degree, and there is no need to change the vibration state.

[0134] In one specific embodiment, after re-acquiring the compensation control operation, the average inner wall temperature of the outer frame (Tfnew), the average inner wall temperature of the sprue (Tgnew), the average interface pressure (Pfnew), and the pressure fluctuation amplitude (σPnew) within the current adjustment cycle are re-acquired. Following the same calculation method as before the compensation control operation, the process state parameters after the compensation control operation are re-acquired, and the outer frame temperature adjustment deviation (ΔTnew) and interface pressure adjustment deviation (ΔPnew) after the compensation control operation are calculated.

[0135] The specific formula for the temperature control efficiency index ET is: ET=∣ΔT0∣ / ∣ΔTnew∣, where ΔT0 is the temperature compensation deviation of the outer frame before the compensation control operation; ET>1 indicates that the deviation decreases after control and the control is effective; ET≤1 indicates that the deviation increases and the control is ineffective.

[0136] The specific formula for the pressure regulation efficiency index EP is: EP=|ΔP0| / |ΔPnew|, where ΔP0 is the interface pressure compensation deviation before the compensation regulation operation; EP>1 indicates that the deviation decreases after regulation and the regulation is effective; EP≤1 indicates that the deviation increases and the regulation is ineffective.

[0137] Please see Figure 4The diagram shown illustrates the logic for determining whether to adjust the preset compensation level in an embodiment of the present invention. It is understood that the control efficiency index directly reflects the effectiveness of the compensation control operation. When ET > 1 or EP > 1, it indicates that the control direction is correct and the magnitude is appropriate, and the deviation is effectively suppressed; when ET ≤ 1 or EP ≤ 1, it indicates that the control strength is insufficient (the deviation has not decreased) or the control direction is incorrect (the deviation has increased), and the preset compensation level needs to be adjusted to enhance the next control strength.

[0138] Furthermore, the temperature control efficiency index ET is compared with the preset temperature efficiency index ETth, or the pressure control efficiency index EP is compared with the preset pressure efficiency index EPth, and the comparison result is used as the benchmark for judging whether the control effect meets the standard. The preset temperature efficiency index ETth and the preset pressure efficiency index EPth are both determined based on historical production data. Specifically, qualified batches are selected from the historical production database, and the distribution of their efficiency indices after control is statistically analyzed. The median is then taken as the preset value. For example, ETth = 1.2 and EPth = 1.15.

[0139] When ET > ETth, it indicates that the control effect has reached or exceeded the preset target, and there is no need to adjust the preset temperature compensation degree.

[0140] When ETth > ET > 1, it indicates that the regulation is effective but has not achieved the preset target, and it is necessary to appropriately reduce the preset temperature compensation degree to enhance the next regulation.

[0141] When ET≤1, it indicates that the control is ineffective and the preset temperature compensation degree needs to be reduced.

[0142] When ET < ETth, it indicates that the current temperature compensation control effect has not reached the preset target, and the corresponding preset temperature compensation adjustment amount needs to be determined by the temperature control index difference. The specific formula for the temperature control index difference ΔET is: ΔET = ETth - ET, where ΔET > 0 indicates insufficient control effect, and the larger the difference, the more severe the deficiency.

[0143] Then, based on the temperature control index difference ΔET, the adjustment amount ΔδTth of the preset temperature compensation degree is determined. The specific formula is: ΔδTth=λT·ΔET, where λT is the temperature proportional coefficient. λT is determined based on historical production data statistics. The specific method is: it is obtained through regression analysis of the control effect and the change in compensation degree in historical data. For example, λT=0.05 is taken.

[0144] When ET < ETth, the preset temperature compensation degree δTth corresponding to the temperature compensation degree is updated to: δTthnew = δTth - ΔδTth, in order to enhance the next regulation.

[0145] Similarly, when EP > EPth, it indicates that the control effect has reached or exceeded the preset target, and there is no need to adjust the preset pressure compensation degree.

[0146] When EPth > EP > 1, it indicates that the regulation is effective but has not achieved the preset target, and it is necessary to appropriately reduce the preset pressure compensation degree to enhance the next regulation intensity.

[0147] When EP≤1, it indicates that the regulation is ineffective and the preset pressure compensation degree needs to be reduced.

[0148] When EP < EPth, it indicates that the current pressure compensation control effect has not reached the preset target, and the corresponding preset pressure compensation adjustment amount needs to be determined by the pressure control index difference. The specific formula for the pressure control index difference ΔEP is: ΔEP = EPth - EP.

[0149] The adjustment amount ΔδPth of the preset pressure compensation degree is determined based on the pressure regulation index difference ΔEP. The specific formula is: ΔδPth=λP·ΔEP, where λP is the pressure proportional coefficient. λP is determined based on historical production data statistics. The specific method is: it is obtained through regression analysis of the regulation effect and the change in compensation degree in historical data. For example, λP=0.08 is taken.

[0150] The revised preset pressure compensation degree δPth is then updated to: δPthnew=δPth-ΔδPth.

[0151] In one specific embodiment, based on the updated preset temperature compensation degree δTthnew, the temperature compensation degree difference ΔδTnew=δT-δTthnew is recalculated to correct the pouring speed;

[0152] Alternatively, based on the updated preset pressure compensation degree δPthnew, the pressure compensation degree difference ΔδPnew=δP-δPthnew can be recalculated to correct the vibration amplitude.

[0153] Repeat the iteration until ET≥ETth and EP≥EPth. The system determines that the control effect has met the standard, outputs the cooled casting, and completes the casting of this batch.

[0154] If the maximum number of iterations (e.g., 5) is reached and the target is still not met, a manual review alert will be issued.

[0155] All technologies not mentioned in the above embodiments are applicable to existing technologies. It is understood that no specific limitation is made to any preset parameter or critical parameter in the embodiments of the present invention, and the above values ​​are not limited thereto. Those skilled in the art can adjust the preset parameters or critical parameters accordingly based on actual needs, analysis of historical data, or equipment usage.

[0156] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A high-efficiency casting process based on a high-strength cast pouring cup outer frame, characterized in that, include: The process state parameters corresponding to the detection points of the outer frame of the pouring cup, the detection point of the sprue, and the detection point of the casting cavity during the pouring process are obtained and feature extraction is performed to obtain the thermal field parameters of the outer frame and the interface pressure field parameters. Based on several outer frame thermal field parameters and several interface pressure field parameters, several corresponding comprehensive casting feature vectors are determined. Several casting comprehensive feature vectors are input into the process state recognition model to output several current process state categories and determine the corresponding state category confidence. The process state categories include normal state, casting impact abnormal state, and shrinkage compensation mismatch state. The confidence level of the stated state category is matched with the confidence level of a preset state category to output the corresponding process category screening set and determine the corresponding accurate recognition rate; Based on the comparison result between the accurate recognition rate and the preset accurate recognition rate, it is determined whether it is necessary to reacquire the process state parameters; Based on the process category filter set, the corresponding process state category is selected, and the outer frame temperature compensation deviation or interface pressure compensation deviation is calculated. The corresponding compensation degree is determined based on the outer frame temperature compensation deviation or the interface pressure compensation deviation. Based on the difference between the compensation degree and the corresponding preset compensation degree, the corresponding compensation control operation is determined and executed. The compensation control operation includes adjusting the pouring speed and starting the pre-demolding vibration. The control efficiency index is determined based on the process state parameters re-acquired after performing the corresponding compensation control operation, and compared with the preset efficiency index to determine whether to adjust the preset compensation degree.

2. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 1, characterized in that, The process of performing feature extraction to obtain the outer frame thermal field parameters and interface pressure field parameters includes: The average values ​​of the inner wall temperatures of the outer frame of several pouring cups and the inner wall temperatures of several straight pouring channels are calculated based on the data obtained within a preset time period to obtain the average inner wall temperature of the outer frame and the average inner wall temperature of the straight pouring channel. The difference is then calculated to obtain the temperature difference of heat loss. The average temperature of the inner wall of the outer frame and the temperature difference due to heat loss are processed and weighted to obtain the corresponding thermal field parameters of the outer frame. The average value of the interface pressure is obtained by averaging the pressures on the inner walls of several casting cavities acquired within a preset time period. The variance of the pressure fluctuation amplitude is calculated based on the pressure on the inner wall of the outer frame of several pouring cups obtained within a preset time. The average interface pressure and the amplitude of the pressure fluctuation are processed separately and weighted to obtain the corresponding interface pressure field parameters. Among them, the outer frame thermal field parameters include the average temperature of the inner wall of the outer frame and the temperature difference due to heat loss, and the interface pressure field parameters include the average interface pressure and the pressure fluctuation amplitude.

3. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 2, characterized in that, The process of determining the corresponding several casting comprehensive feature vectors includes: The outer frame thermal field parameters and the interface pressure field parameters within the adjustment period are aligned according to the time dimension to construct several corresponding casting comprehensive feature vectors, wherein the adjustment period includes several preset times.

4. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 2, characterized in that, The process of determining the corresponding accurate recognition rate includes: The process state categories whose confidence scores are greater than the preset confidence scores are statistically analyzed to obtain the process category filter set. The accuracy rate is determined by calculating the ratio between the number of process state categories in the process category filter set and the total number of process state categories output by the process state recognition model. Based on the comparison results where the accuracy recognition rate is less than or equal to the preset accuracy recognition rate, the process state parameters are determined to be reacquired.

5. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 4, characterized in that, The process of determining the corresponding compensation degree based on the outer frame temperature compensation deviation or the interface pressure compensation deviation includes: The process state category with the highest confidence level is selected from the process category filter set, and the outer frame temperature compensation deviation or the interface pressure compensation deviation is calculated based on the selected process state category. When the process state category is the abnormal state of casting impact, the outer frame temperature compensation deviation is calculated. When the process state category is the shrinkage compensation mismatch state, the interface pressure compensation deviation is calculated.

6. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 5, characterized in that, Based on the average temperature of the inner wall of the outer frame and the temperature difference of heat loss, respectively, the standardization deviation is calculated with respect to the corresponding preset temperature threshold to obtain the temperature difference of the inner wall of the outer frame and the temperature difference deviation of heat loss. The outer frame temperature compensation deviation is obtained by weighted calculation based on the temperature difference between the inner wall of the outer frame and the temperature difference deviation due to heat loss. The corresponding temperature compensation degree is determined based on the correspondence between the outer frame temperature compensation deviation and the compensation degree. The interface pressure deviation and pressure fluctuation deviation are calculated by standardizing the average interface pressure and the pressure fluctuation amplitude with the corresponding preset pressure threshold, respectively. The interface pressure compensation deviation is obtained by weighted calculation based on the interface pressure deviation and the pressure fluctuation deviation. The corresponding pressure compensation degree is determined based on the correspondence between the interface pressure compensation deviation and the compensation degree.

7. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 6, characterized in that, The process of determining the corresponding compensation and control operation includes: When the process state category is the abnormal state of pouring impact, the pouring speed is adjusted and the adjustment range is positively correlated with the temperature compensation degree difference. The temperature compensation degree difference is obtained by calculating the difference between the temperature compensation degree and the preset temperature compensation degree. When the process state category is the shrinkage compensation mismatch state, the pre-demolding vibration is started and the vibration amplitude is positively correlated with the pressure compensation degree difference. The pressure compensation degree difference is obtained by calculating the difference between the pressure compensation degree and the preset pressure compensation degree.

8. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 7, characterized in that, The process of determining the regulation efficiency index includes: Re-collect the process status parameters corresponding to the detection points of the outer frame of the pouring cup, the sprue, and the casting cavity after the compensation and control operation is performed; Calculate the outer frame temperature compensation deviation after the compensation and control operation and record it as the outer frame temperature adjustment deviation; The temperature control efficiency index is obtained by calculating the absolute value of the ratio between the outer frame temperature compensation deviation and the outer frame temperature adjustment deviation. Alternatively, calculate the interface pressure compensation deviation after the compensation and control operation and record it as the interface pressure regulation deviation; The pressure regulation efficiency index is obtained by calculating the absolute value of the ratio between the interface pressure compensation deviation and the interface pressure regulation deviation.

9. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 8, characterized in that, The process of determining whether to adjust the preset compensation degree includes: Based on the comparison results of the temperature control efficiency index being less than or equal to the preset temperature efficiency index, or the pressure control efficiency index being less than or equal to the preset pressure efficiency index, the preset compensation degree is determined to be adjusted based on the temperature control index difference or the pressure control index difference. Wherein, the temperature control index difference is the difference between the preset temperature efficiency index and the temperature control efficiency index, and the pressure control index difference is the difference between the preset pressure efficiency index and the pressure control efficiency index.

10. The efficient casting process based on a high-strength cast pouring cup outer frame according to claim 9, characterized in that, The pouring speed or the vibration amplitude is adjusted based on the preset compensation degree to obtain a cooled casting.

Images