Multi-category die casting processing intelligent adaptation method

Abstract

The application relates to the technical field of industrial control systems and discloses a multi-category die-casting processing intelligent adaptation method, which comprises the following steps: collecting and identifying categories to initialize, extracting a sudden resistance drop to generate a feature set, updating and mapping compensation for a ratio anomaly, searching a residual amount table to obtain a correction parameter, and ignoring compensation summation and amplitude limiting output; the method extracts the double features of pressure sudden drop and stroke end resistance by collecting the ejection stage signal, matches the former with the mold clamping force steady-state signal, compares the former with historical statistical values to determine the anomaly and dynamically update the category mapping threshold, realizes automatic tracking of mold state changes, compares the latter with the last period compensation amount, adopts ascending or attenuating iterative calculation of the current compensation amount, makes the pressure gradually adjust with the friction of the cavity to avoid jumping, introduces first production change ignoring compensation and single-step amplitude limiting, ensures smooth and safe switching, improves the identification accuracy and pressure adaptation stability in long-term multi-category production, and reduces the scrap rate and manual intervention.

Description

Technical Field

[0001] This invention relates to the field of industrial control system technology, specifically to an intelligent adaptation method for multi-category die casting processing. Background Technology

[0002] The field of industrial control systems technology includes automated control and monitoring systems. Its core content involves real-time perception, logical decision-making, and precise control of the operating status of various equipment in the production process. The overall technology field covers programmable logic controllers, distributed control systems, fieldbus communication protocols, and the collaborative work of sensors and actuators, aiming to achieve stable, efficient, and flexible operation of industrial production processes.

[0003] Among them, the intelligent adaptation method for multi-category die casting processing refers to the operation process of collecting real-time parameters such as die casting machine clamping force, injection rod displacement, molten metal temperature, and mold cavity pressure for die castings of different specifications and materials, combining them with a preset category feature database to identify the current workpiece category, and then calling the corresponding injection speed curve, pressure trigger point, holding time window, cooling water flow ratio and other specific process parameters to adjust the servo valve opening, frequency converter motor speed, heating power output and hydraulic system pressure setting value. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a multi-category die casting intelligent adaptation method, which has the function of dynamically updating category thresholds and adaptively compensating for pressure through a dual-feature cross-cycle closed loop in the ejection stage, thus solving the problems mentioned in the background technology.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A method for intelligent adaptation of multi-category die casting processing includes the following steps:

[0007] S1: Collect clamping force, ejection pressure, and ejector rod displacement signals to obtain steady-state signals, call the preset category mapping table to identify the category, call the preset clamping force rated value and preset basic pressure set value to calculate the clamping force margin, initialize the ratio queue and the compensation amount of the previous cycle to zero, and generate a basic parameter set;

[0008] S2: Based on the basic parameter set, extract the pressure drop amplitude and the resistance amplitude at the end of the stroke to generate a dual feature parameter set;

[0009] S3: Based on the basic parameter set and dual characteristic parameter set, calculate the ratio of pressure drop to steady-state signal. If the ratio deviates from the queue statistics, update the category mapping table. Obtain the current compensation amount by comparing the end-of-stroke resistance with the compensation amount of the previous cycle.

[0010] S4: Based on the clamping force allowance, look up the preset allowance-pressure correction table to obtain the allowance correction parameters;

[0011] S5: Based on the margin correction parameter, the current compensation amount and the preset basic pressure setting value, if it is the first production change, the current compensation amount is ignored; otherwise, the three are summed, and the target pressure value is output after the preset single-step limit is set. The execution unit is then sent to generate the pressure control command value.

[0012] Preferably, in step S1, the steady-state signal is obtained by continuously acquiring the output signal of the clamping force sensor at a preset fixed sampling interval. When the fluctuation amplitude between the preset number of continuously acquired sampling points is less than the preset fluctuation threshold, the average value of the data set is taken as the steady-state signal.

[0013] Preferably, in step S1, the preset category mapping table includes the center values ​​of the clamping force threshold intervals corresponding to multiple categories; the category identification specifically involves comparing the steady-state signal with the center value of the clamping force threshold interval for each category, and selecting the category corresponding to the closest center value as the current production category.

[0014] Preferably, in step S1, the clamping force margin is calculated by calculating the difference between the steady-state signal and the preset clamping force rating; the initialization of the ratio queue and the operation of setting the compensation amount of the previous cycle to zero are performed in response to the production change trigger signal.

[0015] Preferably, in step S2, the extraction process of the pressure drop amplitude is as follows: after the ejection action is triggered, delay for a preset number of sampling periods, locate the first sampling point in the ejection cylinder pressure signal where the cumulative drop of a preset number of consecutive sampling points reaches a preset drop threshold, calculate the average pressure of a preset number of sampling points before this sampling point and the average pressure of a preset number of sampling points after this sampling point, and use the difference between the average pressure of the preceding and following points as the pressure drop amplitude; the extraction process of the stroke end resistance amplitude is as follows: locate the first moment when the ejector rod displacement signal reaches a first preset percentage of the total stroke and the second moment when it reaches a second preset percentage of the total stroke, extract the ejection cylinder pressure signal between the first moment and the second moment, calculate the average pressure within this interval, and then calculate the difference between the average pressure within this interval and the average pressure of the subsequent preset number of sampling points as the stroke end resistance amplitude.

[0016] Preferably, in step S3, the method for judging the deviation of the ratio from the queue statistics is as follows: calculate the ratio of the current pressure drop amplitude to the steady-state signal; if the previous batch of ratio queues is empty, then based on the preset mold projection area parameter and preset estimated demolding resistance parameter of the current production type, query the built-in mapping table to obtain the default ratio mean, and call the preset fixed ratio standard deviation parameter; if the previous batch of ratio queues is not empty, then calculate the arithmetic mean and standard deviation of the queue; when the standard deviation is zero, replace it with the minimum threshold of the preset multiple, and then calculate the absolute value of the difference between the current ratio and the arithmetic mean. For any given value, if the absolute value is greater than the product of a preset multiple and the standard deviation, it is considered a deviation. If a deviation is identified, the center value of the clamping force threshold interval corresponding to the current production category in the category mapping table is updated. The current production part is then assessed for conformity. This conformity assessment is based on a comparison between the appearance inspection signal and / or dimensional measurement signal of the die casting and a preset conformity threshold. If the part is determined to be conforming, the current ratio is appended to the end of the ratio queue of the previous batch. If the queue length is greater than a preset length threshold, the earliest data at the head of the queue is removed. If the part is determined to be non-conforming, the current ratio is not appended.

[0017] Preferably, in step S3, the process of obtaining the current compensation amount based on the comparison between the end-of-stroke resistance and the compensation amount of the previous cycle is as follows: when the current production part is determined to be a qualified part, the magnitude of the end-of-stroke resistance is compared with the magnitude of the compensation amount of the previous cycle; if the end-of-stroke resistance is greater than the compensation amount of the previous cycle, the sum of the compensation amount of the previous cycle and the difference between the two multiplied by a preset increase coefficient is calculated; if the end-of-stroke resistance is less than or equal to the compensation amount of the previous cycle, the product of the compensation amount of the previous cycle multiplied by a preset attenuation coefficient is calculated; after relative upper and lower limits are applied to the calculated sum or product relative to the preset basic pressure setting value, it is used as the current compensation amount; when the current production part is determined to be a defective part, the compensation amount of the previous cycle remains unchanged.

[0018] Preferably, in step S4, the preset allowance-pressure correction table is a table pre-calibrated experimentally based on the sign and magnitude of the clamping force allowance; the allowance correction parameter is obtained by looking up the entry in the table that corresponds to the current clamping force allowance.

[0019] Preferably, in step S5, the criterion for ignoring the current compensation amount if it is the first production change is: the current cycle is the first die casting cycle after the production change trigger signal; the summation of the three refers to adding the preset basic pressure setting value, the margin correction parameter, and the current compensation amount; the process of outputting the target pressure value through the preset single-step limit is as follows: calculate the absolute value of the difference between the result obtained by the summation of the three and the pressure control command value generated in the previous cycle. If the absolute value is greater than the preset single-step limit, then the pressure control command value generated in the previous cycle is added to or subtracted from the preset single-step limit as the target pressure value; otherwise, the result obtained by the summation of the three is directly used as the target pressure value; the target pressure value is sent to the execution unit to generate the pressure control command value, and the pressure feedback signal of the hydraulic execution unit is collected. If the absolute value of the deviation between the feedback signal and the pressure control command value exceeds the preset deviation threshold, an alarm is triggered and the current parameters are maintained; otherwise, the pressure control command value is stored as the pressure control command value generated in the previous cycle for the next cycle.

[0020] Preferably, the qualification determination also includes comparing the internal defect detection signal of the die casting with a preset internal defect threshold.

[0021] Compared with existing technologies, this invention provides a smart adaptation method for multi-category die casting processing, which has the following beneficial effects:

[0022] This solution collects pressure and displacement signals during the ejection stage, extracting two independent features: the magnitude of pressure drop and the magnitude of resistance at the end of the stroke. The former is compared with the steady-state clamping force signal to calculate a ratio. Anomalies are identified by comparing this ratio with historical queue statistics, and the category mapping table threshold is dynamically updated, enabling automatic tracking of mold state changes for category identification. The latter is compared with the compensation amount from the previous cycle, using an increasing or decreasing method to iteratively calculate the current compensation amount. This allows the pressure setpoint to gradually adjust with the cavity friction trend, avoiding single-cycle jumps. Simultaneously, a mechanism is introduced to ignore compensation amounts during the first production changeover and a single-step limiting mechanism to ensure smooth switching and safe execution. This improves the accuracy of category identification and the stability of pressure adaptation in long-term multi-category production, reducing scrap rates and the frequency of manual intervention. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the intelligent adaptation method for multi-category die casting processing according to the present invention. Detailed Implementation

[0024] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0025] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings. Figure 1The flowchart of the intelligent adaptation method for multi-category die casting processing provided in the embodiments of the present invention illustrates the complete data flow and logical judgment path from signal acquisition to pressure control command output.

[0026] Example 1

[0027] This embodiment applies to a die-casting machine system equipped with a digital controller, clamping force sensor, ejector cylinder pressure sensor, ejector rod displacement sensor, and hydraulic actuator. The digital controller has a built-in non-volatile memory unit for storing product category mapping tables, preset parameters, and historical operating data.

[0028] Step S1: The digital controller acquires the output signal of the clamping force sensor continuously at a preset fixed sampling interval. When the fluctuation amplitude between a preset number of continuously acquired sampling points is less than a preset fluctuation threshold, the average value of this set of data is taken as the steady-state signal. Simultaneously, the ejection pressure signal output by the ejector cylinder pressure sensor and the ejector rod displacement signal output by the ejector rod displacement sensor are acquired. The digital controller calls a preset category mapping table, which contains the center values ​​of the clamping force threshold intervals corresponding to multiple categories. The steady-state signal is compared with the center value of the clamping force threshold interval for each category, and the category corresponding to the closest center value is selected as the current production category. The preset clamping force rating and preset basic pressure setting are called, and the difference between the steady-state signal and the preset clamping force rating is calculated to obtain the clamping force margin. In response to the production change trigger signal, the ratio queue and the compensation amount of the previous cycle are initialized to zero, and a basic parameter set is generated.

[0029] Step S2: The digital controller performs feature extraction based on the basic parameter set. After a preset number of sampling periods following the triggering of the ejection action, the first sampling point in the ejection cylinder pressure signal where the cumulative decrease of a preset number of consecutive sampling points reaches a preset decrease threshold is located. The average pressure of the preset number of sampling points before and after this sampling point is calculated, and the difference between the two average pressure values ​​is used as the pressure drop amplitude. The first moment when the ejector rod displacement signal reaches a first preset percentage of the total stroke and the second moment when it reaches a second preset percentage of the total stroke are located. The ejection cylinder pressure signal within the interval between these two moments is extracted, and the difference between the average pressure of this interval and the average pressure of the subsequent preset number of sampling points is calculated as the end-stroke resistance amplitude, generating a dual feature parameter set.

[0030] Step S3: The digital controller calculates the ratio of the pressure drop amplitude to the steady-state signal based on the basic parameter set and the dual characteristic parameter set. It then determines whether the previous batch ratio queue is empty: if empty, it queries the built-in mapping table to obtain the default ratio mean based on the preset mold projection area parameter and preset estimated demolding resistance parameter for the current production category, and calls the preset fixed ratio standard deviation parameter; if not empty, it calculates the arithmetic mean and standard deviation of the queue. When the standard deviation is zero, it replaces it with the minimum threshold of the preset multiple, and then calculates the absolute value of the difference between the current ratio and the arithmetic mean. If this absolute value is greater than the product of the preset multiple and the standard deviation, it is determined to be a deviation. If a deviation is determined, the center value of the clamping force threshold interval corresponding to the current production category in the category mapping table is updated. The current production part is then assessed for conformity. This conformity assessment is based on the comparison between the appearance inspection signal and / or dimensional measurement signal of the die casting and the preset conformity threshold, as well as the comparison between the internal defect detection signal and the preset internal defect threshold. Internal defect detection signals include the internal porosity index of the casting obtained through online X-ray or ultrasonic testing. When the internal defect detection signal exceeds the preset internal defect threshold, the current production part is determined to be a defective part, and the current ratio is not added to the ratio queue, nor is the compensation amount updated. If it is determined to be a qualified part, the current ratio is added to the end of the previous batch ratio queue. If the queue length is greater than the preset length threshold, the earliest data at the head of the queue is removed. If it is determined to be a defective part, the current ratio is not added. When the current production part is determined to be a qualified part, the magnitude of the resistance amplitude at the end of the stroke is compared with the compensation amount of the previous cycle: if the resistance amplitude at the end of the stroke is greater than the compensation amount of the previous cycle, the sum of the previous cycle compensation amount and the difference between the two multiplied by a preset rise coefficient is calculated; if the resistance amplitude at the end of the stroke is less than or equal to the compensation amount of the previous cycle, the product of the previous cycle compensation amount multiplied by a preset attenuation coefficient is calculated. The calculated sum or product is then subject to upper and lower limits relative to the preset base pressure setting value and used as the current compensation amount. When the current production part is determined to be a defective part, the compensation amount of the previous cycle remains unchanged.

[0031] Step S4: The digital controller looks up the preset margin-pressure correction table based on the clamping force margin. This table is a table that has been experimentally calibrated in advance based on the sign and magnitude of the clamping force margin. The margin correction parameter is obtained by looking up the entry in the table that corresponds to the current clamping force margin.

[0032] Step S5: The digital controller generates a pressure command based on the margin correction parameter, the current compensation amount, and the preset basic pressure setting. If the current cycle is the first die-casting cycle after the production change trigger signal, the current compensation amount is ignored; otherwise, the preset basic pressure setting, the margin correction parameter, and the current compensation amount are summed. The absolute value of the difference between the summed result and the pressure control command value generated in the previous cycle is calculated. If this absolute value is greater than the preset single-step limit, the pressure control command value generated in the previous cycle is added to or subtracted from the preset single-step limit to obtain the target pressure value; otherwise, the summed result is directly used as the target pressure value. The target pressure value is sent to the execution unit to generate the pressure control command value, and the pressure feedback signal from the hydraulic execution unit is collected. If the absolute value of the deviation between the feedback signal and the pressure control command value exceeds the preset deviation threshold, an alarm is triggered and the current parameters are maintained; otherwise, the pressure control command value is stored as the pressure control command value generated in the previous cycle for the next cycle.

[0033] Through the above steps, this embodiment achieves the technical effect of cross-cycle closed-loop control using dual features during the ejection stage. Two independent features are extracted from the pressure and displacement signals during the ejection stage. The former is paired with the steady-state clamping force signal to calculate the ratio, and anomalies are identified by comparing it with historical queue statistics, dynamically updating the category mapping table threshold so that category identification can automatically track changes in mold state. The latter is compared with the compensation amount of the previous cycle, and the current compensation amount is iteratively calculated using an increasing or decreasing method, allowing the pressure setpoint to gradually adjust with the cavity friction trend, avoiding single-cycle jumps. The introduction of ignoring the compensation amount during the first production changeover and the single-step limiting mechanism ensures smooth switching and safe execution, thereby improving the accuracy of category identification and the stability of pressure adaptation in long-term multi-category production, reducing scrap rate and the frequency of manual intervention.

[0034] Example 2

[0035] This embodiment, based on embodiment 1, further refines the processing logic in the initialization scenario and demonstrates the specific evolution process of cross-cycle iteration.

[0036] When introducing a new product category for the first time during a production changeover, the initialization operations performed in step S1 in response to the production changeover trigger signal specifically include: resetting the previous batch ratio queue to empty and initializing the previous cycle compensation amount to zero. At this time, due to the lack of historical ratio data, the branch in step S3 where the previous batch ratio queue is empty is triggered. The digital controller, based on the preset mold projection area parameters and preset estimated demolding resistance parameters for the current product category, queries the built-in mapping table to obtain the default ratio mean and calls the preset fixed ratio standard deviation parameter. This default ratio mean and standard deviation are only used for anomaly detection in the current cycle and are not stored in the historical ratio queue.

[0037] In the first die-casting cycle, after calculating the current ratio in step S3, the default ratio mean is used as the queue statistical value for deviation judgment. The absolute value of the difference between the current ratio and the default ratio mean is calculated. If the absolute value of this difference is greater than three times the preset fixed ratio standard deviation, it is judged as a deviation, and the center value of the clamping force threshold interval corresponding to the current production category in the category mapping table is updated to the difference between the clamping force steady-state signal and the pressure drop amplitude multiplied by the preset compensation coefficient. If it is judged to be a qualified part, the current ratio is added to the ratio queue, so that the queue has real historical data from the second cycle onwards.

[0038] In step S5, since the current cycle is the first die-casting cycle after the production change trigger signal, the digital controller ignores the current compensation amount and directly calculates the sum of the preset basic pressure setpoint and the allowance correction parameter as the target pressure value, which is then output after a preset single-step limit. The target pressure value is sent to the execution unit to generate a pressure control command value. At this time, the pressure control command value is obtained by adding the preset basic pressure setpoint and the allowance correction parameter obtained by looking up the table based on the clamping force allowance.

[0039] Starting from the second die-casting cycle, the ratio queue already contains the actual ratio data from the first cycle. Step S3 proceeds to the branch where the previous batch ratio queue is not empty. The digital controller calculates the arithmetic mean and standard deviation of the queue. If the standard deviation is zero, it is replaced by the minimum threshold of a preset multiple. The absolute value of the difference between the current ratio and the arithmetic mean is compared with three times the standard deviation, and the normal deviation judgment and threshold update logic is executed. At the same time, the first production change condition in step S5 is no longer met, and the digital controller performs a summation of the three factors, that is, the target pressure value is obtained by adding the preset basic pressure setpoint, the margin correction parameter, and the current compensation amount, and then issuing the generation pressure control command value after limiting.

[0040] As production continues, mold wear gradually causes the steady-state clamping force signal to drift. Assuming that in a certain cycle, mold wear causes the steady-state clamping force signal to decrease by a certain margin from its initial value, the pressure drop amplitude also changes accordingly, causing the ratio to deviate from the historical queue statistics. Step S3 detects that the absolute value of the difference between the current ratio and the arithmetic mean is greater than three times the standard deviation, classifying it as a deviation. The center value of the clamping force threshold interval corresponding to the current production category in the category mapping table is updated to the difference between the steady-state clamping force signal and the pressure drop amplitude multiplied by a preset compensation coefficient. This allows the category identification threshold to automatically track the clamping force drift, avoiding misclassification of categories in subsequent cycles.

[0041] Meanwhile, the resistance at the end of the stroke varies with the friction state of the cavity. Assuming the resistance at the end of a certain cycle is greater than the compensation amount of the previous cycle, step S3 executes an upward branch, calculating the current compensation amount as the compensation amount of the previous cycle plus the difference between the two multiplied by a preset upward coefficient. This compensation amount causes the pressure setpoint to gradually increase as cavity friction increases, preventing excessive ejection resistance due to increased friction. If the resistance at the end of the stroke in subsequent cycles decreases to below the compensation amount of the previous cycle, a decay branch is executed, calculating the current compensation amount as the product of the compensation amount of the previous cycle multiplied by a preset decay coefficient, causing the compensation amount to smoothly decrease.

[0042] Through the aforementioned cross-cycle iteration mechanism, in long-term multi-category switching production, this embodiment can automatically track the drift of clamping force reading caused by mold wear and adhesion, and the pressure setting value can be gradually compensated according to the trend change of the actual friction state of the cavity, which significantly reduces the scrap rate and the frequency of manual readjustment caused by the gradual change of mold state.

[0043] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended technical solutions and their equivalents.

Claims

1. A method for intelligent adaptation of multi-category die casting processing, characterized in that, Includes the following steps: S1: Collect clamping force, ejection pressure, and ejector rod displacement signals to obtain steady-state signals, call the preset category mapping table to identify the category, call the preset clamping force rated value and preset basic pressure set value to calculate the clamping force margin, initialize the ratio queue and the compensation amount of the previous cycle to zero, and generate a basic parameter set; S2: Based on the basic parameter set, extract the pressure drop amplitude and the resistance amplitude at the end of the stroke to generate a dual feature parameter set; S3: Based on the basic parameter set and dual characteristic parameter set, calculate the ratio of pressure drop to steady-state signal. If the ratio deviates from the queue statistics, update the category mapping table. Obtain the current compensation amount by comparing the end-of-stroke resistance with the compensation amount of the previous cycle. S4: Based on the clamping force allowance, look up the preset allowance-pressure correction table to obtain the allowance correction parameters; S5: Based on the margin correction parameter, the current compensation amount and the preset basic pressure setting value, if it is the first production change, the current compensation amount is ignored; otherwise, the three are summed, and the target pressure value is output after the preset single-step limit is set. The execution unit is then sent to generate the pressure control command value.

2. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S1, the steady-state signal is obtained by continuously acquiring the output signal of the clamping force sensor at a preset fixed sampling interval. When the fluctuation amplitude between the preset number of sampling points is less than the preset fluctuation threshold, the average value of the data set is taken as the steady-state signal.

3. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S1, the preset category mapping table includes the center value of the clamping force threshold interval corresponding to each of the multiple categories; the category identification specifically involves comparing the steady-state signal with the center value of the clamping force threshold interval of each category, and selecting the category corresponding to the closest center value as the current production category.

4. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S1, the clamping force margin is calculated by calculating the difference between the steady-state signal and the preset clamping force rating; the initialization of the ratio queue and the operation of setting the compensation amount of the previous cycle to zero are performed in response to the production change trigger signal.

5. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S2, the extraction process for the pressure drop amplitude is as follows: after the ejection action is triggered, delay for a preset number of sampling periods, locate the first sampling point in the ejection cylinder pressure signal where the cumulative drop of a preset number of consecutive sampling points reaches a preset drop threshold, calculate the average pressure of a preset number of sampling points before this sampling point and the average pressure of a preset number of sampling points after this sampling point, and use the difference between the average pressure of the preceding and following points as the pressure drop amplitude; the extraction process for the stroke end resistance amplitude is as follows: locate the first moment when the ejector rod displacement signal reaches a first preset percentage of the total stroke and the second moment when it reaches a second preset percentage of the total stroke, extract the ejection cylinder pressure signal between the first moment and the second moment, calculate the average pressure within this interval, and then calculate the difference between the average pressure within this interval and the average pressure of the subsequent preset number of sampling points as the stroke end resistance amplitude.

6. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S3, the method for judging the deviation of the ratio from the queue statistics is as follows: calculate the ratio of the current pressure drop amplitude to the steady-state signal. If the previous batch of ratio queues is empty, then based on the preset mold projection area parameters and preset estimated demolding resistance parameters of the current production type, query the built-in mapping table to obtain the default ratio mean, and call the preset fixed ratio standard deviation parameter. If the previous batch of ratio queue is not empty, calculate the arithmetic mean and standard deviation of the queue. When the standard deviation is zero, replace it with the minimum threshold of the preset multiple. Then calculate the absolute value of the difference between the current ratio and the arithmetic mean. If the absolute value is greater than the product of the preset multiple and the standard deviation, it is judged as a deviation. If it is determined to be a deviation, then update the center value of the clamping force threshold interval corresponding to the current product category in the category mapping table; The current production part is judged to be qualified. The qualification judgment is based on the comparison between the appearance inspection signal and / or dimensional measurement signal of the die casting part and the preset qualification threshold. If it is judged to be qualified, the current ratio is added to the end of the ratio queue of the previous batch. If the queue length is greater than the preset length threshold, the earliest data at the head of the queue is removed. If it is judged to be unqualified, the current ratio is not added.

7. The intelligent adaptation method for multi-category die casting processing according to claim 6, characterized in that: In step S3, the process of obtaining the current compensation amount based on the comparison between the end-of-stroke resistance and the compensation amount of the previous cycle is as follows: when the current production part is determined to be a qualified part, the magnitude of the end-of-stroke resistance is compared with the magnitude of the compensation amount of the previous cycle. If the resistance at the end of the stroke is greater than the compensation amount of the previous cycle, then the sum of the compensation amount of the previous cycle and the difference between the two multiplied by the preset rise coefficient is calculated; if the resistance at the end of the stroke is less than or equal to the compensation amount of the previous cycle, then the product of the compensation amount of the previous cycle and the preset attenuation coefficient is calculated. After applying upper and lower limits to the calculated sum or product relative to the preset base pressure setting, the current compensation amount is used; when the current production part is determined to be a defective part, the compensation amount of the previous cycle remains unchanged.

8. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S4, the preset allowance-pressure correction table is a table that has been experimentally calibrated in advance based on the sign and magnitude of the clamping force allowance; the allowance correction parameter is obtained by looking up the entry in the table that corresponds to the current clamping force allowance.

9. The intelligent adaptation method for multi-category die casting processing according to claim 1, characterized in that: In step S5, the criterion for ignoring the current compensation amount if it is the first production change is: the current cycle is the first die casting cycle after the production change trigger signal; the summation of the three refers to adding the preset basic pressure setting value, the margin correction parameter, and the current compensation amount; the process of outputting the target pressure value through the preset single-step limit is as follows: calculate the absolute value of the difference between the result obtained by the summation of the three and the pressure control command value generated in the previous cycle. If the absolute value is greater than the preset single-step limit, then the pressure control command value generated in the previous cycle is added to or subtracted from the preset single-step limit as the target pressure value. Otherwise, the result obtained by the summation of the three is directly used as the target pressure value. The target pressure value is sent to the execution unit to generate the pressure control command value, and the pressure feedback signal of the hydraulic execution unit is collected. If the absolute value of the deviation between the feedback signal and the pressure control command value exceeds the preset deviation threshold, an alarm is triggered and the current parameters are maintained. Otherwise, the pressure control command value is stored as the pressure control command value generated in the previous cycle for the next cycle.

10. The intelligent adaptation method for multi-category die casting processing according to claim 6, characterized in that: The qualification determination also includes comparing the internal defect detection signal of the die casting with a preset internal defect threshold.

Images