Block machine mould automatic replacement control method based on state feedback

Abstract

The application provides a block machine mold automatic replacement control method based on state feedback, and relates to the technical field of mold control, which comprises the following steps: based on a plurality of pressure relief stages divided according to the adaptive target pressure drop trajectory, hierarchical pressure relief and reverse vibration control are cooperatively executed to generate a mold separable confirmation signal; the connection state of the mold connection part is judged, and a targeted cooperative loosening scheme is generated to remove the old mold and replace the to-be-replaced mold; by constructing a full-chain feedback mechanism covering the service state of the mold, the locking interface mechanical properties and the dynamic response of the actuator, the deficiencies of the existing automatic mold replacement technology in safety, adaptability and reliability are systematically solved.

Description

Technical Field

[0001] This application relates to the field of mold control technology, and in particular to an automatic mold changing control method for block making machines based on state feedback. Background Technology

[0002] Energy-saving building materials, due to their excellent thermal insulation performance, lightweight and high-strength characteristics, and resource recycling advantages, have become a key component of green building and prefabricated building systems. As an important category of energy-saving building materials, autoclaved aerated concrete blocks and lightweight aggregate concrete blocks are widely used in self-insulating wall systems, and their manufacturing quality is highly dependent on the precision and stability of the molding equipment. Against this backdrop, energy-saving building material manufacturing equipment is accelerating its development towards high-end, intelligent, and integrated directions. Among these, the manufacturing of industrial automatic control system devices serves as a core supporting link, providing the entire equipment with sensing, decision-making, and execution capabilities. Especially in intelligent heat treatment production lines, higher demands are placed on the rapid, accurate, and reliable replacement of tooling molds. This not only requires ensuring production continuity but also avoiding increased energy consumption, higher scrap rates, and equipment downtime losses due to abnormal mold changes.

[0003] As a key piece of equipment in the manufacture of energy-saving building materials, the molds of the block forming machine are used for a long time in high-frequency vibration, high-pressure locking and alternating heat field environment, which can easily cause wear, deformation or jamming. While current automated mold-changing systems have partially replaced manual operation, significant shortcomings remain in practical applications: Existing technologies generally lack a comprehensive judgment mechanism for mold-changing timing and operational safety, failing to effectively integrate multi-dimensional constraints such as aggregate particle size and mold cavity clearance matching, resonance avoidance between excitation frequency and mold natural frequency, and thermal stress shrinkage deformation; pressure relief control strategies mostly employ preset fixed curves, failing to dynamically generate appropriate graded pressure relief trajectories based on the actual contact state of the locking surface (such as pressure center offset and effective bearing area changes), making it difficult to cope with instability risks caused by non-uniform contact or local stress concentration; after pressure relief, there is a lack of quantitative assessment methods for mold posture stability, making it impossible to accurately identify potential hazards such as micro-oscillations, elastic rebound, or static deflection angles, easily leading to separation failure or mechanical impact; furthermore, when jamming occurs at mold connection points, there is a lack of an adaptive and coordinated loosening mechanism based on real-time feedback of lifting load (such as displacement change rate and current fluctuation rate), often relying on experience-set fixed parameters for impact operation, which is inefficient and may cause structural damage. The aforementioned problems severely restrict the operational reliability, energy utilization efficiency, and intelligent level improvement of energy-saving building material manufacturing equipment, making it difficult to meet the urgent needs of intelligent heat treatment production lines for high-speed, high-consistency, and unmanned mold-changing operations. Summary of the Invention

[0004] This application aims to at least partially address one of the technical problems in the related art.

[0005] To achieve the above objectives, this application proposes an automatic mold changing control method for block making machines based on state feedback, comprising the following steps: Step 1: Real-time acquisition of temperature field distribution data on the inner wall of the mold cavity and real-time pressure value of the hydraulic locking system during the operation of the block machine; Step 2: Based on the temperature field distribution data of the inner wall of the mold cavity, execute a progressive composite judgment logic to identify the safe boundary conditions for mold separation under the current working conditions; Step 3: Under the safety boundary conditions, multiple control indicators are combined with the real-time pressure value of the hydraulic locking system. Based on the entire process of pressure drop from the initial locking pressure to zero, multiple pressure relief stages are formed and together constitute the target pressure drop trajectory to generate corresponding reverse compensation commands. Step 4: Based on the multiple pressure relief stages divided by the adaptive target pressure drop trajectory and the corresponding reverse compensation commands, graded pressure relief and reverse vibration control are executed in a coordinated manner. After the pressure relief is completed, a mold separability confirmation signal is generated based on the stability index of the mold throughout the pressure relief process. Step 5: In response to the confirmation signal that the mold can be separated, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are collected to construct a two-dimensional phase plane to obtain the slope of the motion trajectory, determine the connection status of the mold connection part, and generate a targeted collaborative loosening scheme to remove the old mold and replace the mold to be replaced.

[0006] Furthermore, based on the three-dimensional working condition safety verification matrix, a progressive composite judgment logic is executed to identify the safety boundary conditions for mold separation under the current working condition, including the following steps: Step 21: Obtain the maximum aggregate particle size parameter corresponding to the current production task and the mold cavity design clearance of the mold to be replaced. If the mold cavity design clearance is greater than 1.5 times the maximum aggregate particle size parameter, proceed to the next step; otherwise, determine that the safety boundary conditions are not met and terminate the replacement process. Step 22: Obtain the current excitation frequency and the natural frequency of the mold to be replaced. Determine the frequency difference based on the excitation frequency and the natural frequency. If the frequency difference is within the preset resonance avoidance range, it is determined that the safety boundary conditions are not met and the replacement process is terminated; otherwise, proceed to the next step. Step 23: Based on the temperature field distribution data of the inner wall of the mold cavity, obtain the stress amplitude of the maximum thermal stress concentration area, as well as the mold length and the difference between the average temperature of the mold cavity and the room temperature. Determine the shrinkage deformation based on the stress amplitude, mold length, and the difference between the average temperature of the mold cavity and the room temperature. Step 24: If the shrinkage deformation is less than the preset tolerance threshold, the safety boundary condition is satisfied; otherwise, the safety boundary condition is not satisfied and the replacement process is terminated.

[0007] Furthermore, to meet multiple control indicators of the mold locking surface under the aforementioned safety boundary conditions, and to divide the entire pressure drop process from the initial locking pressure to zero into multiple pressure relief stages that together constitute the target pressure drop trajectory, a corresponding reverse compensation command is generated, including the following steps: Step 31: Obtain contact stress distribution data at each coordinate position of the mold locking surface based on the shrinkage deformation; determine the pressure center coordinates based on all coordinate positions of the mold locking surface; determine the effective bearing area of ​​the locking surface based on the contact stress distribution data. Step 32: Determine the effective bearing area ratio and pressure center offset distance based on the effective bearing area and pressure center coordinates; Step 33: If the effective bearing area ratio is greater than or equal to a preset area ratio threshold and the pressure center offset distance is less than or equal to a preset distance threshold, then enter the uniform contact pressure relief mode. The entire pressure drop process from the initial locking pressure to zero is divided into three pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: rapid pressure relief stage, pressure holding and balancing stage, and slow pressure relief stage; the first target pressure drop trajectory is formed based on the continuous pressure intervals corresponding to the three pressure relief stages. Step 34, otherwise, enter the non-uniform pressure relief mode: divide the entire pressure drop process from the initial locking pressure to zero into two pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: local stress equalization pressure relief stage and global rapid pressure relief stage; based on the continuous pressure intervals corresponding to the two pressure relief stages, a second target pressure drop trajectory is formed; Step 35: Based on the global rapid depressurization stage of the second target pressure drop trajectory and the three depressurization stages of the first target pressure drop trajectory, a basic reverse compensation command is generated; a dynamic reverse compensation command is generated based on the local stress equilibrium depressurization stage; both the dynamic reverse compensation command and the basic reverse compensation command consist of phase and amplitude.

[0008] Furthermore, [initial locking pressure, 0.6 times the initial locking pressure] is defined as the rapid pressure relief stage, [0.6 times the initial locking pressure, 0.6 times the initial locking pressure] is held for 3-5 seconds and defined as the pressure holding balance stage, and [0.6 times the initial locking pressure, 0] is defined as the slow pressure relief stage.

[0009] Furthermore, [initial locking pressure, x times the initial locking pressure] is defined as the local stress equalization and pressure relief stage, and [x times the initial locking pressure, 0] is defined as the pressure holding and balance stage, with x between 0.3 and 0.5.

[0010] Furthermore, based on the multiple depressurization stages defined by the adaptive target pressure drop trajectory, and the corresponding reverse compensation commands, graded depressurization and reverse vibration control are executed in a coordinated manner. After depressurization is completed, a mold separability confirmation signal is generated based on the mold's stability index throughout the depressurization process, including the following steps: Step 41: After entering each pressure relief stage, execute the pressure relief action and reverse compensation command, and monitor the real-time pressure value of the hydraulic locking system. If the real-time pressure value of the hydraulic locking system exceeds the continuous pressure range of the current pressure relief stage, pause the current pressure relief action, maintain the current real-time pressure value of the hydraulic locking system until it returns to the allowable range, and then continue the pressure relief and reverse compensation command operation of the next stage. Step 42: The angular displacement and angular velocity of the mold are synchronously sampled at a fixed control cycle to obtain the angular displacement data sequence and angular velocity data sequence arranged in time sequence; Step 43: After depressurization is complete and the hydraulic pressure drops to zero, execute the following composite judgment logic. If none of the composite judgment logics are satisfied, proceed to step 54; otherwise, determine that the mold posture is unstable and do not generate a mold separability confirmation signal. The composite judgment logic includes three sub-conditions: a. Determine the range of continuous sampling times corresponding to each pressure relief stage. If the absolute value of the angular displacement corresponding to any sampling time within the range of continuous sampling times is greater than the first safety threshold corresponding to the current pressure relief stage; b. If the absolute value of the angular velocity measured at each of at least three consecutive sampling times is greater than the preset second safety threshold. c. Within the sampling time range corresponding to the pressure holding balance section or the slow pressure relief section, identify local extreme points based on the angular displacement data sequence. Starting from the sampling time corresponding to the local extreme point, if there is another sampling time with an angular displacement value that has the opposite sign to the angular displacement value of the local extreme point, and the angular displacement change amplitude between the two is greater than the preset third safety threshold. Step 44: Calculate the energy entropy value of the angular displacement data sequence during the entire depressurization process. If the energy entropy value is less than the preset entropy threshold, generate the mold separability confirmation signal; otherwise, do not generate it.

[0011] Furthermore, in response to the mold separability confirmation signal, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are collected to construct a two-dimensional phase plane to obtain the slope of the motion trajectory. The connection status of the mold connection parts is determined, and a targeted collaborative loosening scheme is generated to remove the old mold and replace the mold to be replaced. This includes the following steps: Step 51: In response to the mold separability confirmation signal, the real-time displacement value of the lifting mechanism and the real-time current value of the drive motor are collected in real time. Step 52: Perform central difference calculation on the displacement values ​​at three consecutive sampling times to obtain the displacement change rate at the current time. Calculate the relative change rate between the current value at the current sampling time and the current value at the previous time, and mark it as the current fluctuation rate. Step 53: Construct a two-dimensional phase plane based on the displacement change rate and current fluctuation rate; construct a data set based on the displacement change rate and current fluctuation rate and mark it as the working point; fit the working points of the five most recent consecutive sampling times into a motion trajectory and calculate the slope of the motion trajectory; Step 54: If the absolute difference between the current slope and the slope of the previous window is greater than the preset sudden change threshold and the current fluctuation rate is greater than the empirical threshold, then it is determined that there is mechanical jamming at the mold connection part; otherwise, it is determined that there is normal free separation at the mold connection part. Step 55: Based on the mechanical jamming, generate reciprocating motion commands for the lifting mechanism and hydraulic pulse pressure commands, and start and stop them synchronously for a duration of 1 to 3 seconds; if during the execution process, the displacement change rate returns to the normal range and the current fluctuation rate is less than or equal to the empirical threshold, then terminate the current loosening operation in advance. Step 56: Based on the normal free separation state, control the lifting mechanism to rise according to the preset constant speed curve until the maximum separation stroke is reached.

[0012] Furthermore, the process of obtaining the phase and amplitude of the local stress equalization and pressure relief stage is as follows: the phase of the local stress equalization and pressure relief stage is obtained through the pressure center coordinates, the current locking pressure is obtained, and the amplitude of the local stress equalization and pressure relief stage is obtained based on the pressure center offset distance and the current locking pressure.

[0013] Furthermore, the reciprocating motion command of the lifting mechanism is as follows: the servo driver controlling the lifting mechanism outputs a sine or triangular waveform position command. When the current fluctuation rate is between 0.1 and 0.3, the frequency of the sine or triangular waveform is set to 5 Hz; when it is between 0.3 and 0.6, the frequency of the sine or triangular waveform is set to 10 Hz; when it is greater than 0.6, the frequency of the sine or triangular waveform is set to 20 Hz; and the amplitude of the sine or triangular waveform is fixed at 0.3 mm.

[0014] Furthermore, the hydraulic pulse pressure command: controls the hydraulic proportional valve to output a pulse pressure signal with the same frequency and phase as the reciprocating motion command of the lifting mechanism; obtains the amplitude reference, generates an incremental term based on the current fluctuation rate, and obtains the amplitude of the pulse pressure based on the sum of the amplitude reference and the incremental term.

[0015] Compared with existing technologies, the automatic mold replacement control method for block making machines based on state feedback provided in this application, based on the control concept of state feedback, systematically solves the shortcomings of existing automatic mold replacement technologies in terms of safety, adaptability, and reliability by constructing a full-chain feedback mechanism covering the mold's service status, the mechanical properties of the locking interface, and the dynamic response of the actuator. Specifically, in the mold replacement decision stage, multi-source state parameters such as the cumulative number of working cycles of the mold, the temperature field distribution of the mold cavity wall, the pressure of the hydraulic locking system, and the vibration table spectrum are collected in real time. Based on these parameters, a three-dimensional working condition safety verification matrix is ​​constructed to achieve progressive judgment on key safety boundary conditions such as aggregate particle size matching, resonance risk, and thermal shrinkage deformation, ensuring that subsequent processes are triggered only when the comprehensive state meets the safety threshold. In the pressure relief execution stage, based on the shrinkage deformation determined by the aforementioned state feedback, the contact stress distribution of the locking surface is further analyzed, the pressure center coordinates and effective bearing area are calculated, and a uniform or non-uniform pressure relief mode is dynamically selected accordingly. A multi-stage target pressure drop trajectory adapted to the current contact state is also defined. Simultaneously, reverse compensation commands containing phase and amplitude are generated, forming a technical path of state perception—trajectory planning—command generation; in the separation confirmation stage, the mold angular displacement and angular velocity sequence are continuously fed back, and the attitude stability of the entire pressure relief process is quantified through composite logic criteria and energy entropy analysis. Only when the state indicators meet the standards is a separable signal output; in the final loosening stage, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are introduced again as state feedback quantities, a two-dimensional phase plane is constructed and the slope of the motion trajectory is extracted, the jamming state of the connection part is accurately identified, and then a coordinated loosening command with frequency, amplitude and phase all linked to the load state is generated, truly realizing full-cycle state drive from working condition assessment, process control to end execution.

[0016] In summary, this invention deeply integrates the manufacturing concepts of industrial automatic control system devices and intelligent heat treatment production line control, significantly improving the safety, reliability, and intelligence level of energy-saving building material manufacturing equipment in the mold replacement process. Through a collaborative loosening mechanism driven by multi-dimensional state perception, adaptive trajectory planning, full-process stability verification, and load feedback, it effectively avoids misoperation under unsafe conditions, suppresses attitude instability during pressure relief, greatly improves the success rate of loosening stuck molds, shortens the mold replacement cycle, and reduces energy consumption and equipment wear, providing key technical support for the high-quality, high-efficiency, and continuous production of energy-saving building materials. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 A flowchart of an automatic mold replacement control method for a block machine based on state feedback provided in an embodiment of this application; Figure 2This is a structural block diagram of the automatic mold changing control system for block making machines based on state feedback provided in an embodiment of this application; Figure 3 This is a block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0019] The automatic mold changing control method for block making machines based on state feedback, according to an embodiment of this application, is described below with reference to the accompanying drawings.

[0020] like Figure 1 As shown, the automatic mold changing control method for block making machines based on state feedback includes the following steps: Step 1: Collect multi-source state parameters during the operation of the block machine in real time. The multi-source state parameters include the cumulative number of working cycles of the mold, the temperature field distribution data of the inner wall of the mold cavity, the real-time pressure value of the hydraulic locking system, and the vibration spectrum characteristics of the vibration table.

[0021] In this embodiment, the cumulative number of working cycles of the mold is recorded in real time by a high-precision photoelectric encoder or limit switch installed on the mold lifting or locking mechanism. Each complete pressing-demolding action is counted as one working cycle, and the data is accumulated and stored by a PLC or edge controller. The temperature field distribution data of the inner wall of the mold cavity is acquired by a multi-point distributed thermocouple or infrared thermal imaging sensor array embedded in the side wall and bottom of the mold cavity. The sensors are arranged in layers along the height of the mold cavity to capture the non-uniform heat distribution in different areas caused by steam curing or environmental temperature differences. The sampling frequency is not less than 1. The system ensures the timeliness of thermal stress assessment by using a high-response piezoelectric pressure sensor installed in the inlet / outlet oil circuit of the locking cylinder. The output signal is filtered and amplified by the signal conditioning module before being input to the analog input module of the industrial control system, achieving millisecond-level dynamic pressure tracking. The vibration spectrum characteristics of the vibration table are acquired by a triaxial accelerometer fixed to the vibration table base or mold support frame, collecting the original vibration signal. This signal is then filtered against anti-aliasing and sent to an embedded FFT (Fast Fourier Transform) processing unit to extract spectral characteristic parameters such as the main frequency, harmonic components, and energy concentration frequency band in real time. All four types of parameters are uniformly and synchronously acquired by the central controller via the industrial bus and timestamped, forming a structured multi-source state data stream.

[0022] Step 2: Based on the temperature field distribution data of the inner wall of the mold cavity, execute a progressive composite judgment logic to identify the safe boundary conditions for mold separation under the current working conditions; Step 21: Obtain the maximum aggregate particle size parameter corresponding to the current production task and the mold cavity design clearance of the mold to be replaced. If the mold cavity design clearance is greater than 1.5 times the maximum aggregate particle size parameter, proceed to the next step; otherwise, determine that the safety boundary conditions are not met and terminate the replacement process.

[0023] When the cumulative number of working cycles of the mold reaches a preset threshold (e.g., 5000 times), the controller automatically triggers a mold pre-replacement request. This threshold is determined based on mold material fatigue life test data and historical failure statistics, aiming to proactively intervene before significant degradation of structural performance and avoid sudden fractures or deformations affecting product quality. Subsequently, the controller reads the maximum aggregate particle size parameter (e.g., 10 mm) from the current production task configuration file and retrieves the mold cavity design clearance (e.g., 18 mm) of the mold to be replaced from the mold digital twin model or equipment parameter table. The controller then checks whether the "mold cavity design clearance is greater than 1.5 times the maximum aggregate particle size." If the clearance is less than 1.5 times, large aggregate particles are prone to getting stuck between the mold wall and the aggregate during material distribution or vibration compaction, leading to localized stress concentration, surface scratches, or a sharp increase in demolding resistance. This not only accelerates mold wear but may also cause defects such as chipped edges and corners in the blocks. Setting 1.5 times as a safety boundary is an empirical critical value derived from extensive simulations of fluid concrete particle flow and on-site verification. This ensures smooth aggregate passage while also considering mold structural strength and product dimensional accuracy. This judgment must be strictly followed because the matching of aggregate particle size and mold cavity clearance is a physical prerequisite for the feasibility of mold replacement. If this condition is ignored and replacement is forced, the new mold, although installed, will be unable to produce qualified products, resulting in unnecessary downtime and wasted resources. Furthermore, this judgment can only be made at this stage, as the aggregate parameters are uniquely determined by the current production task, while the mold cavity clearance is an inherent property of the mold, neither of which can be dynamically adjusted during subsequent pressure relief or loosening stages. Therefore, only by conducting pre-verification at this point can we prevent incompatible molds from being put into operation from the source, ensuring the process rationality and production continuity of mold replacement operations.

[0024] Step 22: Obtain the current excitation main frequency and the natural frequency of the mold to be replaced. Determine the frequency difference based on the excitation main frequency and the natural frequency. If the frequency difference is within the preset resonance avoidance range, it is determined that the safety boundary conditions are not met and the replacement process is terminated; otherwise, proceed to the next step.

[0025] The current excitation frequency is obtained by extracting the frequency corresponding to the energy peak after the vibration spectrum characteristics of the vibration table collected in step 1 are processed by Fast Fourier Transform (FFT). This frequency reflects the actual operating frequency of the vibration motor under the current production task. The natural frequency of the mold to be replaced is pre-calibrated through finite element modal analysis or pre-shipment hammer test and stored as a structural attribute parameter in the mold digital file. The typical value range is 25 Hz to 45 Hz, depending on the mold mass, support stiffness and material damping characteristics. Subsequently, the controller calculates the absolute value of the difference between the two to obtain the frequency difference value and determines whether it falls within the preset resonance avoidance range - this range is set to ±3Hz, that is, if |excitation frequency − natural frequency| ≤ 3Hz, it is determined to be in the resonance risk zone. The reason for adopting this judgment logic is that when the excitation frequency approaches the mold's natural frequency, the system will experience a resonance amplification effect, leading to accelerated loosening of the locking mechanism bolts, fatigue leakage of hydraulic joints, and even mold displacement deviation or breakage. The ±3Hz avoidance bandwidth is determined based on extensive field vibration test data and structural response simulation: within this range, the acceleration response amplitude can surge to 3–5 times that of normal operating conditions, significantly exceeding the safety threshold. The advantage of setting this condition is that high-risk operating conditions can be proactively avoided before mold replacement, preventing severe vibrations induced by minor disturbances during subsequent depressurization or lifting processes, thereby ensuring the safety of personnel and equipment and avoiding misjudgments or mechanical damage caused by resonance. This judgment must be completed at this stage because: the excitation frequency is determined by the current production formula and vibration parameters, while the natural frequency is an inherent property of the mold; the combination of the two is only comparable during the mold replacement decision window. Once the depressurization or loosening stage begins, the vibration source may be shut down or changed, failing to accurately reflect the dynamic coupling risks during operation. Therefore, only by performing a preliminary verification based on the measured excitation frequency and the known natural frequency in step 32 can the source of resonance risk be effectively intercepted, ensuring that the entire automatic mold changing process is based on dynamic safety.

[0026] Step 23: Based on the temperature field distribution data of the inner wall of the mold cavity, obtain the stress amplitude of the maximum thermal stress concentration area, as well as the mold length and the difference between the average temperature of the mold cavity and the room temperature. Determine the shrinkage deformation based on the stress amplitude, mold length, and the difference between the average temperature of the mold cavity and the room temperature.

[0027] First, from the temperature field distribution data collected by a multi-point temperature sensor array embedded in the inner wall of the mold cavity, the region with the largest temperature difference between adjacent measuring points and the duration of this temperature difference exceeding a preset time threshold (e.g., 30 seconds) is identified as the region of maximum thermal stress concentration. Then, based on the local temperature difference amplitude of this region, the thermal expansion coefficient of the mold material, and the structural constraint status of the region (e.g., whether it is near reinforcing ribs or locking surfaces), the corresponding equivalent thermal stress amplitude is obtained through a lookup table. This mapping relationship is formed by jointly calibrating the thermo-mechanical performance curves provided by the material supplier with historical service data. Next, the real-time readings of all effective temperature measuring points within the mold cavity are arithmetically averaged to obtain the average temperature of the mold cavity; simultaneously, the current room temperature value is obtained, and the difference between the two is used to obtain the overall temperature difference value. At the same time, the total design length of the mold to be replaced is retrieved from the mold parameter database. Based on this, the calculation process for the theoretical thermal shrinkage is as follows: Multiply the overall temperature difference by the total design length of the mold, and then multiply by the linear expansion coefficient of the mold material. The result is the expected thermal shrinkage along the length of the mold under ideal uniform cooling conditions. Subsequently, a dimensionless correction coefficient is introduced to adjust the theoretical thermal shrinkage. The adjustment method is to multiply the two, that is: corrected thermal shrinkage = theoretical thermal shrinkage × correction coefficient. This correction coefficient is greater than or equal to 1, and its specific value is obtained from a pre-established stress-residual deformation calibration table based on the equivalent thermal stress amplitude level corresponding to the aforementioned maximum thermal stress concentration area. This calibration table is constructed by performing three-dimensional laser scanning on molds of the same material and structural type under different thermal shock cycles, and statistically analyzing the ratio of the actual residual shrinkage after cooling to the theoretical value.

[0028] Step 24: If the shrinkage deformation is less than the preset tolerance threshold, the safety boundary condition is satisfied; otherwise, the safety boundary condition is not satisfied and the replacement process is terminated.

[0029] If the calculated shrinkage deformation is less than the preset tolerance threshold (e.g., 0.5 mm), the current mold state is determined to meet the safety boundary conditions, and the process is allowed to proceed to the next stage; otherwise, if the shrinkage deformation is greater than or equal to the threshold, the safety boundary conditions are determined not to be met, and the current replacement process is immediately terminated, while an alarm is triggered and the abnormal event is recorded.

[0030] The reason for using "whether the shrinkage deformation is less than the tolerance threshold" as the criterion is that when the mold is disassembled in a hot state, its actual geometric dimensions have not yet stabilized to the room temperature design state. If the expected shrinkage is too large, it will lead to uncontrollable mechanical interference during the subsequent release of the locking mechanism, the execution of the lifting action, or the installation of the new mold. For example, when the shrinkage deformation exceeds 0.5 mm, interference jamming may occur between the mold base and the locating pin due to the shrinkage after cooling. The hydraulic ejector rod forcibly ejecting the mold can easily cause the mold to tilt, the guide pillar to bend, or even break. In addition, if the new mold is installed according to the room temperature dimensions, but the old mold has not yet completed shrinkage, it may also cause mold closing misalignment, affecting the subsequent molding accuracy or even damaging the equipment. Therefore, a clear and quantifiable length threshold must be used as a safety boundary to intercept high-risk working conditions.

[0031] This decision logic offers significant engineering advantages. First, it simplifies the complex thermo-mechanical coupling problem into a single measurable and verifiable output (shrinkage deformation), facilitating integration into the logic decision modules of existing PLCs or edge controllers without requiring complex simulations or real-time solutions. Second, by setting a unified tolerance threshold, it enables automated evaluation of molds from different batches and of different materials under the same safety standards, improving production line compatibility and robustness. Third, by prematurely terminating high-risk mold change operations, it effectively avoids equipment damage, downtime, and product quality incidents caused by forced execution, significantly improving system reliability and production continuity.

[0032] More importantly, this judgment "can only be made here and in this way" for three reasons: First, shrinkage deformation is a comprehensive reflection of thermal history and structural response, and can only be accurately obtained after temperature field acquisition, stress region identification, and length parameter fusion are completed; the preceding steps cannot provide sufficient information. Second, the safety boundary must be verified before physical actions (such as depressurization, loosening, and lifting) are performed; once the execution phase begins, the system loses its ability to avoid risks. Third, the tolerance threshold is not arbitrarily set, but is determined based on the equipment's mechanical clearance margin, mold positioning accuracy requirements, and historical fault data statistics. For example, a 0.5 mm threshold corresponds to the upper limit of the safety of the guide column fit clearance of a typical die-casting machine (usually 0.3–0.6 mm); exceeding this value poses a risk of dry friction or seizure. Therefore, only by strictly comparing the measured shrinkage deformation with the engineering threshold in this step can we ensure efficiency while maintaining the safety baseline, achieving "predictive safety control" rather than "post-event remediation."

[0033] Step 3: Under the safety boundary conditions, multiple control indicators are combined with the real-time pressure value of the hydraulic locking system. Based on the entire process of pressure drop from the initial locking pressure to zero, multiple pressure relief stages are formed and together constitute the target pressure drop trajectory to generate corresponding reverse compensation commands.

[0034] Step 31: Obtain contact stress distribution data at each coordinate position of the mold locking surface based on the shrinkage deformation; determine the pressure center coordinates based on all coordinate positions of the mold locking surface; determine the effective bearing area of ​​the locking surface based on the contact stress distribution data.

[0035] Step 32: Determine the effective bearing area ratio and pressure center offset distance based on the effective bearing area and pressure center coordinates; First, the mold locking surface is divided into several discrete coordinate position units, each corresponding to a local region with known temperature and structural response. Based on the previously obtained shrinkage deformation, combined with the elastic modulus and Poisson's ratio of the mold material in this region, the contact stress value borne by each coordinate position unit in the locking state is determined by looking up a table, thus forming contact stress distribution data covering the entire locking surface. Subsequently, all coordinate position units are traversed, and the coordinate value of each unit is multiplied by its corresponding contact stress value and then accumulated. The accumulated result is then divided by the sum of the contact stress values ​​of all units to obtain the horizontal and vertical coordinates of the pressure center. At the same time, an effective contact stress threshold is set. This threshold is pre-calibrated based on the surface hardness and allowable crushing strength of the mold material. The total area covered by all coordinate position units in the contact stress distribution data that are greater than or equal to this threshold is accumulated to obtain the effective bearing area of ​​the locking surface.

[0036] Divide the effective bearing area obtained in step 31 by the theoretical total area of ​​the mold locking surface, and the resulting ratio is the effective bearing area ratio; at the same time, obtain the coordinates of the geometric center of the mold locking surface, and calculate the straight-line distance between the coordinates of the geometric center and the coordinates of the pressure center obtained in step 31. This distance is the pressure center offset distance.

[0037] Step 33: If the effective bearing area ratio is greater than or equal to a preset area ratio threshold and the pressure center offset distance is less than or equal to a preset distance threshold, then enter the uniform contact pressure relief mode. The entire pressure drop process from the initial locking pressure to zero is divided into three pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: rapid pressure relief stage, pressure holding and balancing stage, and slow pressure relief stage; the first target pressure drop trajectory is formed based on the continuous pressure intervals corresponding to the three pressure relief stages. The initial locking pressure is defined as the rapid pressure relief stage, the pressure is maintained for 3-5 seconds at the initial locking pressure, and the pressure is maintained for 0.6 times the initial locking pressure. The initial locking pressure is defined as the pressure holding and balancing stage. The initial locking pressure is defined as the slow pressure relief stage.

[0038] If the effective bearing area ratio is greater than or equal to a preset area ratio threshold and the pressure center offset distance is less than or equal to a preset distance threshold, the mold locking surface is determined to be in a uniform and stable contact state, and the uniform contact pressure relief mode is entered. The rationale behind this judgment is that the effective bearing area ratio reflects the proportion of the locking surface that actually participates in bearing the load. If it reaches or exceeds the preset area ratio threshold (e.g., 85%), it indicates sufficient contact between the mold and the supporting structure, with no obvious local overhangs or edge warping. Simultaneously, if the pressure center offset distance does not exceed the preset distance threshold (e.g., 2% of the locking surface characteristic dimension, typically 3 mm), it indicates good load distribution symmetry and no significant off-center loading. These two conditions together constitute the engineering criterion for "uniform contact." Only under this state will the mold not tilt, slip, or jam during pressure relief due to sudden changes in local stress or torque imbalance. Therefore, both threshold requirements must be met simultaneously to activate the pressure relief strategy optimized for uniform working conditions.

[0039] The core reason for adopting a three-stage pressure relief strategy is to balance efficiency and safety. The rapid pressure relief stage is used to quickly release most of the non-critical locking force, shortening the overall mold changeover time; the pressure holding and balancing stage maintains an intermediate pressure level briefly, allowing the thermal stress inside the mold and the micro-deformation of the contact interface to reach mechanical equilibrium again, avoiding local rebound or sudden gap changes caused by a sudden drop in pressure; the slow pressure relief stage precisely controls the unloading rate in the low pressure range, preventing the sudden release of residual friction force from causing mold jumping or positioning failure. This segmented strategy avoids the cycle time loss caused by slow pressure relief throughout the entire process, and also avoids the mechanical impact risk that may be induced by a single-stage rapid pressure relief.

[0040] This classification method is irreplaceable for the following reasons: First, the pressure relief behavior must be strictly matched with the contact state. If this mode is used in non-uniform contact, the pressure holding stage cannot compensate for the torque caused by the off-center load, and will instead solidify the unfavorable state. Second, the pressure ranges of the three stages are not arbitrarily set, but are critical points determined based on a large number of field tests: 0.6 times the initial locking pressure is the inflection point for most die-casting / injection molds to transition from macroscopic locking to microscopic fit maintenance. Below this value, the contact stiffness decreases significantly, and the pressure relief rate needs to be reduced. The pressure holding platform is set as a constant pressure range (i.e., the upper and lower limits are the same) to ensure that the system has a clear steady-state window for adaptive adjustment. If this classification is not followed, contact stability cannot be maintained during dynamic unloading.

[0041] The two thresholds are based on the following criteria: the preset area ratio threshold (e.g., 85%) is derived from the minimum effective contact requirement under typical wear and assembly tolerances of the mold locking surface. Below this value, edge stress concentration can easily lead to surface crushing. The preset distance threshold (e.g., 3 mm) is determined based on a combination of the locking surface geometry and the equipment's guiding accuracy. Exceeding this offset will cause an overturning moment due to excessive lever arm, even if the total bearing area is sufficient, threatening disassembly and assembly safety. Both thresholds are calibrated through a combination of historical fault data inversion and bench testing, possessing strong engineering constraints and cannot be arbitrarily relaxed.

[0042] Step 34, otherwise, enter the non-uniform pressure relief mode: divide the entire pressure drop process from the initial locking pressure to zero into two pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: local stress equalization pressure relief stage and global rapid pressure relief stage; based on the continuous pressure intervals corresponding to the two pressure relief stages, form a second target pressure drop trajectory; define [initial locking pressure, x times the initial locking pressure] as the local stress equalization pressure relief stage, and define [x times the initial locking pressure, 0] as the pressure holding balance stage, where x is between 0.3 and 0.5.

[0043] If the effective bearing area ratio is less than a preset area ratio threshold, or the pressure center offset distance is greater than a preset distance threshold, the mold locking surface is determined to be in a non-uniform contact state, and a non-uniform pressure relief mode is entered. The identification logic of this mode is based on the following engineering fact: when the effective bearing area of ​​the locking surface is insufficient or the pressure center is significantly offset, it indicates that the mold has problems such as local suspension, edge warping, asymmetrical thermal deformation, or assembly misalignment, resulting in load concentration in a few areas. Directly and rapidly releasing pressure in this state will cause the mold to tilt, slip, or even jam in the guide structure due to the sudden unloading of local high-stress areas. Therefore, a controllable low-speed unloading stage must first be implemented to gradually release energy from high-stress areas, while allowing low-stress or non-contact areas to "catch up" with deformation coordination, thereby achieving dynamic equilibrium of the contact state. Only after completing this equilibrium process can subsequent rapid pressure relief be safely entered. Therefore, it is necessary to identify the non-uniform working condition based on the aforementioned two contact state indicators and switch to a specially designed pressure relief strategy.

[0044] The non-uniform pressure relief mode is divided into two stages: the first stage is the local stress equalization pressure relief stage, where the pressure slowly decreases from the initial locking pressure to x times the initial locking pressure, where x ranges from 0.3 to 0.5; the second stage is the global rapid pressure relief stage, where the pressure rapidly decreases from x times the initial locking pressure to zero. This division has clear physical significance and operational advantages. The core objective of the local stress equalization pressure relief stage is to eliminate the stress gradient on the contact surface. By maintaining a sufficiently long unloading time in the higher pressure range (retaining 30%–50% of the original locking force), the material undergoes slight but sufficient elastic rebound and interface redistribution, avoiding local collapse-type unloading. Once the pressure drops below x times the initial locking pressure, the contact stiffness has been significantly reduced, and the overall structure tends to be flexible, allowing for safe rapid pressure relief to improve efficiency. This two-stage strategy maximizes the compression of mold-changing cycle time under non-uniform working conditions while ensuring safety.

[0045] This identification and division method is irreplaceable for several reasons: First, the pressure-holding platform in the uniform pressure relief mode cannot be used under non-uniform contact conditions because the offset of the pressure center means that even if pressure is held, torque balance cannot be achieved, which will prolong the high-risk state. Second, the x-value must be limited to the range of 0.3–0.5. If it is higher than 0.5, the residual pressure is too large, which cannot effectively trigger the contact response in the low-stress zone, resulting in insufficient equalization. If it is lower than 0.3, the residual locking force is too small when entering rapid pressure relief, which is insufficient to maintain the stability of the mold posture and is prone to shaking at the end of the pressure relief. This range has been verified by a large number of hot demolding tests and is the optimal compromise between stress equalization effect and structural stability. In addition, the number of stages cannot exceed two, otherwise the control complexity will increase dramatically and the marginal benefits will decrease; they cannot be combined into a single stage, otherwise equalization and efficiency cannot be balanced. Therefore, this two-stage division is the only solution that balances safety, effectiveness, and feasibility under non-uniform conditions.

[0046] Based on the three continuous pressure intervals defined by the uniform contact pressure relief mode, the starting and ending points of each stage are determined sequentially: the rapid pressure relief stage starts from the initial locking pressure and linearly decreases to 0.6 times the initial locking pressure; the pressure holding and balancing stage maintains a constant pressure of 0.6 times the initial locking pressure for 3 to 5 seconds; the slow pressure relief stage starts from 0.6 times the initial locking pressure and linearly decreases to zero at a lower slope than the rapid pressure relief stage. Subsequently, the pressure-time relationships of these three stages are spliced ​​together in chronological order to form a continuous, piecewise linear pressure decrease curve, which is the first target pressure decrease trajectory. This trajectory serves as the setpoint sequence for pressure closed-loop regulation in the control system, guiding the hydraulic or pneumatic actuators to unload precisely at a predetermined rate and platform.

[0047] When forming the second target pressure reduction trajectory, the pressure boundaries of each stage are first determined based on the two continuous pressure intervals defined by the non-uniform pressure relief mode: the local stress equalization pressure relief stage starts from the initial locking pressure and linearly decreases to x times the initial locking pressure at a relatively low rate, where x is a specific value between 0.3 and 0.5, which can be dynamically selected according to the severity of the current effective bearing area ratio or the pressure center offset distance; the global rapid pressure relief stage starts from x times the initial locking pressure and linearly decreases to zero at a relatively high rate. Then, the pressure-time periods of these two stages are connected end-to-end to form a continuous broken-line pressure reduction path containing a slow-decrease segment and a rapid-decrease segment; this path is the second target pressure reduction trajectory. This trajectory also serves as the pressure control command for the actuator, ensuring that stress equalization is prioritized under non-uniform contact conditions, followed by efficient completion of the remaining unloading.

[0048] Step 35: Based on the global rapid pressure relief phase of the second target pressure drop trajectory and the three pressure relief phases of the first target pressure drop trajectory, a basic reverse compensation command is generated; a dynamic reverse compensation command is generated based on the local stress equilibrium pressure relief phase; both the dynamic reverse compensation command and the basic reverse compensation command consist of phase and amplitude. The process of obtaining the phase and amplitude of the local stress equilibrium pressure relief phase is as follows: the phase of the local stress equilibrium pressure relief phase is obtained through the pressure center coordinates; the current locking pressure is obtained; and the amplitude of the local stress equilibrium pressure relief phase is obtained based on the pressure center offset distance and the current locking pressure.

[0049] Dynamic reverse compensation command (only used in the local stress equalization and pressure relief phase of the second target pressure drop trajectory): Phase acquisition process: The locking surface is a rectangular plane, and its geometric center coordinates are predefined as the reference origin. The system acquires the horizontal and vertical coordinates of the pressure center in real time. The offsets of the horizontal and vertical coordinates are calculated by subtracting the corresponding horizontal and vertical coordinate values ​​of the geometric center, respectively, to obtain the lateral offset and longitudinal offset. Using the geometric center as the pole and the positive lateral direction of the locking surface as the zero-degree reference direction, the angle of the offset direction is determined using the four-quadrant arctangent method: if the lateral offset is zero and the longitudinal offset is positive, the phase is 90 degrees; if the lateral offset is zero and the longitudinal offset is negative, the phase is 270 degrees; if the longitudinal offset is zero and the lateral offset is positive, the phase is 0 degrees; if the longitudinal offset is zero and the lateral offset is negative, the phase is 180 degrees; in other cases, the ratio of the longitudinal offset to the lateral offset is first calculated, and then the acute angle value is obtained through the arctangent function. Then, the quadrant to which the offset belongs is determined according to the positive and negative combination of the lateral and longitudinal offsets, and the acute angle value is converted into the corresponding standard azimuth angle, ranging from 0 degrees to 360 degrees. This angle is the phase of the dynamic reverse compensation command.

[0050] Amplitude acquisition process: The system reads the current locking pressure value and obtains the calculated pressure center offset distance (i.e., the Euclidean distance from the pressure center to the geometric center). The current locking pressure value and the pressure center offset distance are used as two input parameters to index a two-dimensional preset amplitude mapping table. This mapping table is established through bench testing during the equipment commissioning phase: under different locking pressure levels, different eccentric loads are applied, and the critical offset distance at which the mold exhibits attitude instability during unloading is measured, and the minimum reverse drive amount required at this point is recorded; this drive amount is normalized and filled into the corresponding position in the table as the amplitude. During operation, the control system locates the four nearest neighbor data points in the mapping table based on the current locking pressure and offset distance, and calculates the accurate amplitude output using bilinear interpolation. This amplitude represents the reverse action intensity that the actuator needs to apply in the phase direction.

[0051] Basic reverse compensation command (used for the three stages of the first target pressure drop trajectory and the global rapid pressure relief stage of the second target pressure drop trajectory): Phase setting: In uniform contact pressure relief mode, since the effective bearing area ratio meets the standard and the pressure center offset distance does not exceed the limit, the system defaults to symmetrical contact state. Therefore, the basic reverse compensation command phase of the three stages is uniformly set to zero degrees, that is, no directional compensation is introduced, and it is only used to offset the inherent inertia of the system or pipeline hysteresis.

[0052] In the global rapid depressurization phase of the non-uniform depressurization mode, although the overall state is non-uniform, this phase has entered the low pressure range. The main risk is residual vibration rather than overturning. Therefore, the phase value used at the end of the local stress equilibrium depressurization phase is adopted to maintain the consistency of the compensation direction and avoid abrupt changes.

[0053] Amplitude Setting: For the three stages of the first target pressure drop trajectory, the amplitude is fixed according to the stage type: the amplitude is set to a low value in the rapid pressure relief stage to suppress hydraulic shock; the amplitude is set to zero in the pressure holding and balancing stage because the pressure is constant and no compensation is needed; the amplitude is set to a medium-low value in the slow pressure relief stage to smoothly transition to zero pressure. These amplitudes are determined by standard mold testing before the equipment leaves the factory and are fixed in the control parameter library.

[0054] For the global rapid depressurization phase of the second target pressure drop trajectory, its amplitude is taken as 50% of the final output amplitude of the local stress equilibrium depressurization phase, in order to retain some directional suppression capability and avoid over-driving in the low-pressure area.

[0055] If the phase is not based on the pressure center coordinates, the compensation direction may deviate from the actual off-center load direction, which will not only be ineffective but may even exacerbate the imbalance. If the amplitude is not considered in conjunction with the current locking pressure and offset distance, the required compensation intensity cannot be accurately quantified. For example, under the same offset distance, the overturning moment generated by high locking pressure is much greater than that under low pressure conditions, and they must be treated differently. Therefore, the phase must originate from the pressure center position, and the amplitude must be jointly determined by the pressure and offset distance. This is a necessary and sufficient condition for achieving effective, safe, and adaptive reverse compensation.

[0056] Step 4: Based on the multiple pressure relief stages divided by the adaptive target pressure drop trajectory and the corresponding reverse compensation commands, graded pressure relief and reverse vibration control are executed in a coordinated manner. After the pressure relief is completed, a mold separability confirmation signal is generated based on the stability index of the mold throughout the pressure relief process. Step 41: After entering each pressure relief stage, execute the pressure relief action and reverse compensation command, and monitor the real-time pressure value of the hydraulic locking system. If the real-time pressure value of the hydraulic locking system exceeds the continuous pressure range of the current pressure relief stage, pause the current pressure relief action, maintain the current real-time pressure value of the hydraulic locking system until it returns to the allowable range, and then continue the pressure relief and reverse compensation command operation of the next stage.

[0057] Upon entering any depressurization stage, the control system simultaneously performs two operations: first, it drives the hydraulic actuator to perform a depressurization action according to the target pressure drop trajectory corresponding to the current stage; second, it simultaneously outputs the corresponding reverse compensation command (including phase and amplitude) for that stage, which is applied to the mold-locking guide or auxiliary leveling mechanism. Simultaneously, the system monitors the actual pressure value of the hydraulic locking system in real time at a sampling frequency of no less than fifty times per second.

[0058] If the monitored real-time pressure value exceeds the upper or lower limit of the continuous pressure range defined for the current pressure relief stage (for example, in the local stress equalization pressure relief stage, the pressure should be between the initial locking pressure and x times the initial locking pressure; if the measured pressure is higher than the initial locking pressure or lower than x times the initial locking pressure, it is considered to be out of bounds), the pressure relief action should be immediately suspended, the pressure relief valve should be closed or the unloading output of the servo pump should be stopped, and the pressure value in the current hydraulic chamber should be kept unchanged; at the same time, the reverse compensation command of the current stage should be continuously output until the real-time pressure value returns to the pressure range allowed for the current stage; only after the pressure stabilizes and meets the range constraints can the remaining pressure relief action of this stage be continued until the stage is completed, and then the pressure relief and compensation operation of the next stage can be started.

[0059] The target pressure drop trajectory is a carefully designed safe unloading path based on the mold contact state, material elastic response, and system dynamics. If the actual pressure deviates from the preset range for this stage, it indicates an anomaly in the system, such as an excessively rapid response of the pressure relief valve, sudden changes in hydraulic oil temperature leading to viscosity changes, external vibration interference, or sensor drift. If pressure is continued according to the original plan at this time, it may lead to sudden changes in local stress, loss of mold posture control, or jamming of the locking mechanism. Especially in non-uniform pressure relief mode, excessive pressure can easily cause irreversible mechanical damage.

[0060] By adopting a closed-loop fault-tolerant mechanism of "pausing when exceeding the limit and resuming after stabilization", the pressure relief process is transformed from an open-loop execution to a controlled feedback process, which significantly improves the robustness of the system. This avoids interruption of the entire mold changing process due to instantaneous disturbances and prevents forced pressure relief under dangerous conditions. At the same time, maintaining the current pressure instead of emergency reset preserves the established stress equilibrium state and avoids fatigue accumulation caused by repeated loading and unloading.

[0061] The essence of dividing the pressure relief phase is to discretize the continuous unloading process into several control windows with clear physical meanings. Each window corresponds to specific mechanical behavior requirements (such as stress equalization, pressure stabilization, or rapid release). Once the pressure deviates from this window, the phase function fails. For example, if the pressure is too low during the local stress equalization phase, it means that rapid pressure relief has begun before equalization is complete, thus losing its protective function; if the pressure is too high, it indicates that pressure relief has not started or that pressure has increased in reverse, violating the purpose of unloading. Therefore, the pressure range of the current phase must be used as the only legal operating domain. Any deviation from this range represents a control mismatch, and only by pausing and waiting for recovery can the preconditions for subsequent phases be ensured. Other judgment methods (such as monitoring only the rate of pressure change or the cumulative pressure relief) cannot directly reflect whether the phase function has been effectively executed, and therefore do not possess the same level of safety and logical completeness.

[0062] Step 42: The angular displacement and angular velocity of the mold are sampled synchronously at a fixed control cycle to obtain the angular displacement data sequence and angular velocity data sequence arranged in time sequence.

[0063] Simultaneously with the pressure relief operation, the control system synchronously monitors the mold's attitude state at a preset fixed control cycle. This control cycle is strictly consistent with the main control cycle of hydraulic pressure control and reverse compensation command output, with a typical value of twenty milliseconds, ensuring strict alignment of multi-source data on the time axis. Specifically, the system acquires the mold's spatial attitude information in real time using high-precision tilt sensors or inertial measurement units installed at the four corners of the mold clamping mechanism or near the central rotation axis. At the arrival of each control cycle, the data acquisition module simultaneously triggers the reading of angular displacement and angular velocity signals: angular displacement is obtained by synthesizing the pitch and roll angles directly output by the sensors to obtain the mold's overall tilt angle relative to the horizontal reference plane; angular velocity is obtained by synthesizing the angular rate measured by the sensor's built-in gyroscope after coordinate transformation to obtain the mold's total angular velocity at the current moment. The acquired angular displacement and angular velocity values ​​are stored in two independent first-in-first-out buffer queues according to the sampling time sequence, forming a strictly chronologically arranged sequence of angular displacement and angular velocity data. Each data point in the sequence has a timestamp index corresponding to the control cycle, ensuring that it can be accurately matched with other timing signals such as pressure data and compensation instructions during subsequent processing.

[0064] Step 43: After depressurization is complete and the hydraulic pressure drops to zero, execute the following composite judgment logic. If none of the composite judgment logics are satisfied, proceed to step 54; otherwise, determine that the mold posture is unstable and do not generate a mold separability confirmation signal. The composite judgment logic includes three sub-conditions: a. Determine the range of continuous sampling times corresponding to each pressure relief stage. If the absolute value of the angular displacement corresponding to any sampling time within the range of continuous sampling times is greater than the first safety threshold corresponding to the current pressure relief stage; b. If the absolute value of the angular velocity measured at each of at least three consecutive sampling times is greater than the preset second safety threshold. c. Within the sampling time range corresponding to the pressure holding balance section or the slow pressure relief section, identify local extreme points based on the angular displacement data sequence. Starting from the sampling time corresponding to the local extreme point, if there is another sampling time with an angular displacement value that has the opposite sign to the angular displacement value of the local extreme point, and the angular displacement change amplitude between the two is greater than the preset third safety threshold.

[0065] After the hydraulic locking system completes all depressurization stages and real-time monitoring confirms that the hydraulic pressure has stabilized and dropped to zero, the composite judgment logic for mold posture stability is immediately initiated. This logic includes three parallel sub-conditions (a, b, c). If any sub-condition is met, the mold posture is determined to be unstable, and no mold separation confirmation signal is generated; only when all three sub-conditions are not met will step 54 be entered.

[0066] Sub-condition a: Staged angular displacement exceeding limit judgment For each pressure relief stage, a predefined range of continuous sampling times is defined (e.g., the k-th stage starts from the m-th control cycle and ends at the n-th control cycle). Within this range, the absolute value of the angular displacement at each sampling time in the angular displacement data sequence is checked point by point to see if it exceeds the first safety threshold specifically set for that stage. If any point exceeds the threshold, then sub-condition a is met.

[0067] The permissible deflection of the mold attitude varies at different decompression stages. For example, in the local stress equalization decompression stage, the system actively applies reverse compensation, allowing for a small but controllable tilt; while in the slow decompression stage, near-perfect equilibrium is required, allowing for even smaller deflections. Therefore, differentiated thresholds must be set for each stage and rigorously verified within the corresponding time window to reflect the mechanical constraints of each stage. This enables refined, stage-adaptive attitude monitoring, avoiding misjudgments or omissions caused by a "one-size-fits-all" threshold. For example, if the strictest threshold is uniformly used, false alarms may occur frequently in the early non-uniform stages; if a lenient threshold is used, minor instabilities cannot be captured in later stages. Stage-based judgment balances safety and process efficiency because the risk of mold overturning is strongly correlated with the current decompression stage. Angular displacement values ​​outside the stage context have no clear safety significance. Only by embedding angular displacement limits into the time window in which they occur and matching the physical expectations of that window can a valid criterion be established. Other methods (such as global maximum angular displacement) cannot distinguish between "controllable off-center loading" and "dangerous instability."

[0068] The initial safety threshold for each stage is obtained through bench testing. Under standard mold conditions and typical off-center loading, the maximum angular displacement that does not induce jamming or residual stress in each stage is recorded, and then multiplied by a safety factor (usually 0.7–0.8) to form a conservative but feasible threshold. For example, the threshold for the local stress equalization stage is set at 0.15 degrees, and for the slow pressure relief stage at 0.05 degrees.

[0069] Sub-condition b: Continuous high angular velocity judgment Traverse the angular velocity data sequence throughout the entire depressurization process, and search for the existence of at least three consecutive sampling moments (i.e., spanning at least two complete control cycles, such as 60 milliseconds) where the absolute value of the angular velocity at each moment is greater than the preset second safety threshold. If such a moment exists, then sub-condition b is true.

[0070] Instantaneous angular velocity spikes may be caused by sensor noise or external vibrations and are not dangerous; however, if the angular velocity remains consistently high, it indicates that the mold is accelerating its overturning, and the system's damping or compensation mechanism has failed, which is a precursor to irreversible instability. Random interference is filtered out through "continuous multi-point" constraints, focusing on real dynamic instability; simultaneously, "no fewer than three" constraints ensure that the judgment has a time-cumulative effect, conforming to the inertial response characteristics of the mechanical system and improving the robustness of the criterion. A single instantaneous angular velocity cannot distinguish between disturbances and trends, and an excessively long continuous window (such as ten points) will delay the alarm. Three continuous points constitute a response window of approximately 60 milliseconds in a 20-millisecond period, satisfying both industrial real-time requirements and capturing the initial acceleration phase; this is the minimum reliable criterion length proven effective in engineering practice.

[0071] The second safety threshold is derived by inversely calculating the mold's rotational inertia and the maximum damping torque of the hydraulic system. The upper limit of the velocity corresponding to the maximum angular acceleration that may occur under safe unloading conditions is calculated, and 80% of this is taken as the threshold. A typical value is 0.8 degrees / second. This value is much higher than the thermal drift or micro-vibration levels during normal pressure relief, but lower than the critical overturning velocity.

[0072] Sub-condition c: Angular displacement sign reversal and large oscillation judgment This process is only executed within the sampling time range corresponding to the pressure holding and balancing phase or the slow pressure relief phase. First, all local extreme points (i.e., the angular displacement value of a point is greater than or less than the two points immediately before and after it) are identified in the angular displacement data sequence during this time period. For each local extreme point, starting from its sampling time, all subsequent sampling points are searched: if there exists a point whose angular displacement has the opposite sign to that extreme point (one positive and one negative), and the sum of their absolute values ​​(i.e., the amplitude of change) is greater than the preset third safety threshold, then sub-condition c is met.

[0073] The pressure holding and slow pressure release phases should be in a quasi-static equilibrium state. If the angular displacement direction reverses and the change is significant, it indicates that the system is experiencing oscillating behavior of "springback-overshoot-reverse tilting" during unloading. This is a typical manifestation of insufficient support stiffness or excessive compensation, which can easily lead to mold misalignment or guide pillar damage. Instead of focusing solely on static deflection, the system should directly capture non-monotonic unloading behavior to identify potential dynamic instability. This criterion is highly sensitive to dangerous conditions where "the angular displacement appears small but there is repeated swaying." In the quasi-static phase, the ideal angular displacement should monotonically approach zero or maintain a small, constant offset. Any sign reversal violates physical expectations. Merely detecting extreme values ​​cannot detect oscillations; merely detecting the total change may ignore intermediate processes. Only by combining "opposite sign + excessive amplitude" can harmful oscillations be accurately identified.

[0074] The third safety threshold is set at 50% of the maximum non-interference swing angle allowed by the mold guide clearance. For example, if the maximum swing angle corresponding to the guide post fit clearance is 0.2 degrees, then the threshold is set to 0.1 degrees. This value ensures that even if small oscillations occur, the mechanical limit will not be triggered, preserving sufficient safety margin.

[0075] Step 44: Calculate the energy entropy value of the angular displacement data sequence during the entire depressurization process. If the energy entropy value is less than the preset entropy threshold, generate the mold separability confirmation signal; otherwise, do not generate it.

[0076] After confirming that the hydraulic pressure has dropped to zero and that none of the composite judgment logic in step 53 has been triggered, the system performs energy entropy calculation on the angular displacement data sequence collected throughout the depressurization process. This calculation process is a purely textual numerical processing flow: First, the angular displacement data sequence is divided into several continuous segments in chronological order, each segment containing the same number of data points; second, the sum of the squares of the absolute values ​​of all data points within each segment is calculated as the energy value of that segment; next, the energy values ​​of all segments are added together to obtain the total energy, and the energy value of each segment is divided by the total energy to obtain the energy percentage of each segment; subsequently, for each non-zero energy percentage, the logarithm to base 2 is taken, and this logarithm is multiplied by the corresponding energy percentage. All product results are then summed and their negative values ​​are taken. The final value obtained is the energy entropy value of the angular displacement data sequence.

[0077] If the energy entropy value is less than the preset entropy threshold, the system generates a mold separation confirmation signal; otherwise, no signal is generated. Energy entropy comprehensively reflects the concentration of energy distribution during the entire depressurization process. If the mold unloads smoothly and its attitude is stable, most of the energy is concentrated in the initial stage or a very small range, resulting in a highly concentrated energy distribution and a low entropy value. If there are multiple disturbances, oscillations, or sudden deflections, the energy is dispersed over multiple time periods, tending towards a uniform distribution, and the entropy value increases significantly. Therefore, energy entropy is an effective global indicator for measuring the overall stability of the unloading process.

[0078] Compared to criteria that rely solely on local peaks or instantaneous changes, energy entropy integrates time-energy information throughout the entire process, exhibiting higher sensitivity to weak but persistent unstable behavior and inherent robustness to occasional noise. Furthermore, this criterion does not require pre-setting specific waveform patterns, making it applicable to various molds and operating conditions, and highly versatile. This is because after completing the staged local judgment (step 53), a comprehensive stability measure covering the entire cycle without blind spots is still needed to capture risks that are difficult to identify through local thresholds, such as cross-stage cumulative effects or low-frequency drift. Energy entropy precisely quantifies "orderliness" in an information-theoretic way: smooth unloading equates to high order and low entropy, while instability equates to disorder and high entropy. Other global indicators (such as variance and peak-to-valley difference) cannot distinguish periods of concentrated energy or reflect dynamic evolution characteristics, thus lacking equivalent discriminative power.

[0079] This threshold was determined through statistical analysis of historical data under extensive standard operating conditions. In hundreds of normal unloading experiments, the energy entropy of each angular displacement sequence was calculated, and the maximum value was multiplied by a safety factor of 1.2 to serve as the upper limit threshold. Typical entropy thresholds range from 1.5 to 2.0. This setting ensures that all known safe operating conditions can be met, while effectively intercepting abnormal oscillations or hysteresis responses.

[0080] Step 5: In response to the confirmation signal that the mold can be separated, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are collected to construct a two-dimensional phase plane to obtain the slope of the motion trajectory, determine the connection status of the mold connection part, and generate a targeted collaborative loosening scheme to remove the old mold and replace the mold to be replaced.

[0081] Step 51: In response to the mold separability confirmation signal, the real-time displacement value of the lifting mechanism and the real-time current value of the drive motor are collected in real time.

[0082] In this embodiment, after the mold separability confirmation signal is generated, the status monitoring process of the lifting mechanism is immediately triggered. The control system synchronously starts two high-precision acquisition channels: one channel is connected to a displacement sensor installed on the piston rod of the lifting cylinder or the mechanical push rod, which is used to obtain the current physical displacement value of the lifting mechanism in real time; the other channel is connected to the current detection module of the drive motor (such as a Hall current sensor or the current feedback interface built into the frequency converter), which is used to obtain the effective value of the phase current or the total current amplitude of the drive motor in real time.

[0083] Data acquisition is performed using the same fixed control cycle as the main control system, typically twenty milliseconds, to ensure strict alignment of displacement and current data on the time axis. Within each control cycle, the system simultaneously triggers sampling, analog-to-digital conversion, and numerical reading of two signals, storing the obtained real-time displacement and current values ​​in chronological order into independent buffer queues, forming a synchronously aligned displacement-current time-series data pair. This data pair is continuously updated until the lifting action is completed or the system enters the next control phase.

[0084] Step 52: Perform central difference calculation on the displacement values ​​at three consecutive sampling times to obtain the displacement change rate at the current time. Calculate the relative change rate between the current value at the current sampling time and the current value at the previous time, and label it as the current fluctuation rate.

[0085] From the displacement data sequence, extract three consecutive displacement values ​​at the current sampling time, the previous sampling time, and the next sampling time. Subtract the previous displacement value from the next displacement value, and then divide the difference by the time length of two sampling periods. The result is the displacement change rate at the current moment. Simultaneously, extract the current value at the current sampling time and the current value at the previous sampling time from the current data sequence. Calculate the absolute value of the difference between the two, and then divide this absolute value by the current value at the previous sampling time. If the current value at the previous time is zero, the current value is used as the substitute denominator to avoid division by zero. The result is the current fluctuation rate at the current moment, and this is labeled as the current fluctuation rate.

[0086] Step 53: Construct a two-dimensional phase plane based on the displacement change rate and current fluctuation rate; construct a data set based on the displacement change rate and current fluctuation rate and mark it as the working point; fit the working point of the most recent five consecutive sampling times into a motion trajectory and calculate the slope of the motion trajectory.

[0087] The displacement change rate obtained at the current sampling moment is used as the abscissa and the current fluctuation rate as the ordinate, combined into a data set, which is then marked as the operating point at the current moment. Subsequently, the system extracts the operating points corresponding to the five most recent consecutive sampling moments, including the current moment, from the cache, arranges them in chronological order, forming a sequence containing five ordered operating points. Based on these five operating points, the linear least squares method is used to fit a straight line to their distribution on the two-dimensional phase plane, obtaining a fitted straight line that best represents the overall trend of the sequence. Finally, the system calculates the slope of this fitted straight line on the two-dimensional phase plane, that is, the change in the ordinate corresponding to each unit change in the abscissa from left to right; the resulting value is the slope of the motion trajectory.

[0088] Step 54: If the absolute difference between the current slope and the slope of the previous window is greater than the preset sudden change threshold and the current fluctuation rate is greater than the empirical threshold, then it is determined that there is mechanical jamming at the mold connection part; otherwise, it is determined that there is normal free separation at the mold connection part.

[0089] Obtain the slope of the motion trajectory calculated in the current window and compare it with the slope obtained in the previous sliding window, calculating the absolute value of the difference between the two. If the absolute difference is greater than a preset abrupt change threshold, and the current fluctuation rate at the current sampling moment is greater than a set empirical threshold, then it is determined that there is mechanical jamming at the mold connection part; otherwise, it is determined that the mold connection part is in a normal free separation state.

[0090] The judgment is made using a dual condition of slope change and current fluctuation rate. During the lifting process, if jamming occurs at the mold connection point, the lifting mechanism will encounter unexpected resistance, leading to a sudden increase in the load on the drive motor, manifested as a significant increase in current fluctuation rate. Simultaneously, the displacement change rate decreases rapidly or even stagnates due to obstruction, causing a drastic deflection of the working point's trajectory in the phase plane and a significant jump in the slope of the motion trajectory. Single indicators (such as looking only at current or displacement) are easily affected by noise, starting inertia, or load fluctuations. The slope, however, reflects the changing trend of the dynamic relationship between displacement and current coupling. Combining this with current fluctuation rate can effectively eliminate instantaneous disturbances and accurately identify genuine jamming events.

[0091] On the one hand, the structural changes in the dynamic characteristics of the system are captured by the slope of the phase plane trajectory, which has high sensitivity to abnormal modes. On the other hand, the introduction of current fluctuation rate as an auxiliary criterion enhances the ability to perceive abnormal energy input. The two together constitute a dual verification mechanism of "abnormal dynamic response + abnormal energy consumption", which greatly improves the accuracy and robustness of jamming identification and avoids false shutdowns or missed alarms. In the mold separation stage, normal free separation is characterized by a steady increase in displacement, small fluctuations in current, and a smooth phase plane trajectory with a stable slope. However, once jamming occurs, the physical constraints will break the original electromechanical coupling balance, which will inevitably cause both dynamic hysteresis of displacement and surge in current. If only displacement stagnation is relied upon for judgment, low-speed uniform lifting may be misjudged as jamming; if only current increase is relied upon for judgment, transient motor start-up or power grid fluctuations may be misjudged as faults. Only by understanding the logic and relationship between slope mutation (reflecting dynamic mismatch) and excessive current fluctuation rate (reflecting load abnormality) can the specific fault mode of "mechanical jamming" be uniquely identified, and interference from other operating conditions be eliminated.

[0092] The mutation threshold was derived from statistical analysis of extensive normal separation experimental data. During hundreds of fault-free lifting operations, the maximum difference in slope between adjacent windows was calculated, and then multiplied by a 1.3 safety margin to ensure that no alarms were triggered under normal operating conditions. The empirical threshold was determined based on a measured comparison of the motor's rated no-load current and typical jamming current, using twice the upper limit of normal fluctuation as the judgment boundary. Both thresholds have undergone field calibration and long-term operational verification, covering practical factors such as equipment aging and temperature drift, while possessing sufficient fault detection sensitivity, demonstrating engineering feasibility and stability.

[0093] Step 55: Based on the mechanical jamming, generate reciprocating motion commands for the lifting mechanism and hydraulic pulse pressure commands, and start and stop them synchronously for a duration of 1 to 3 seconds; if during the execution process, the displacement change rate returns to the normal range and the current fluctuation rate is less than or equal to the empirical threshold, then terminate the current loosening operation in advance. The reciprocating motion command of the lifting mechanism is as follows: the servo driver of the lifting mechanism outputs a sine or triangular waveform position command. When the current fluctuation rate is between 0.1 and 0.3, the frequency of the sine or triangular waveform is set to 5 Hz; when it is between 0.3 and 0.6, the frequency of the sine or triangular waveform is set to 10 Hz; when it is greater than 0.6, the frequency of the sine or triangular waveform is set to 20 Hz; and the amplitude of the sine or triangular waveform is fixed at 0.3 mm.

[0094] Hydraulic pulse pressure command: Control the hydraulic proportional valve to output a pulse pressure signal with the same frequency and phase as the reciprocating motion command of the lifting mechanism; obtain the amplitude reference, generate an incremental term based on the current fluctuation rate, and obtain the amplitude of the pulse pressure based on the sum of the amplitude reference and the incremental term.

[0095] When generating a hydraulic pulse pressure command, the system first reads a preset amplitude reference value. This reference value is the minimum effective pulse pressure amplitude required for the equipment to release slight jamming under standard operating conditions. Then, it calculates an increment based on the current current fluctuation rate: if the current fluctuation rate is less than 0.3%, the increment is zero; if the current fluctuation rate is between 0.3 and 0.6%, the increment is 20% of the amplitude reference value; if the current fluctuation rate is greater than 0.6%, the increment is 50% of the amplitude reference value. Adding the amplitude reference value and the increment yields the amplitude of the current pulse pressure. Based on this, the system generates a pulse pressure signal with the same frequency and phase as the reciprocating motion command of the lifting mechanism and outputs it to the hydraulic proportional valve.

[0096] Step 56: Based on the normal free separation state, control the lifting mechanism to rise according to the preset constant speed curve until the maximum separation stroke is reached.

[0097] The constant speed curve is defined as a continuous ascent from the current displacement position at a set constant speed until the pre-configured maximum separation stroke position is reached. In practice, the current actual displacement value is first read as the starting point, and then the maximum separation stroke value and constant ascent speed value are retrieved from the parameter storage area. Subsequently, the servo driver generates a corresponding uniform speed position command sequence based on this constant speed value and tracks its execution in real time, ensuring that the speed fluctuation of the lifting mechanism does not exceed the allowable tolerance range throughout the entire ascent process. During the ascent, the displacement feedback value is continuously monitored. When the actual displacement value is detected to be equal to or exceed the maximum separation stroke value, the position command output is immediately stopped, and the lifting mechanism is locked at the current position, completing the separation action. If an abnormal decrease in the displacement change rate or an excessive current fluctuation rate occurs during the ascent process, the constant speed ascent process is interrupted and fault response processing is initiated; however, this situation should not occur under normal free separation conditions.

[0098] When this application determines that there is mechanical jamming at the mold connection point, a loosening strategy must be adopted that can effectively relieve the jamming without damaging the equipment. In this case, simply applying continuous unidirectional thrust can easily lead to drive overload or structural damage, while random vibration or blind pressure increase lacks specificity and may exacerbate the malfunction. Therefore, a reciprocating motion command for the lifting mechanism and a hydraulic pulse pressure command are generated, ensuring that they start and stop synchronously, with an execution duration set to 1 to 3 seconds. An adaptive termination mechanism is also introduced: once the displacement change rate returns to the normal range and the current fluctuation rate falls below the empirical threshold, the operation is immediately terminated. This strategy breaks the static friction balance through small-amplitude, high-frequency reciprocating motion, while a hydraulic proportional valve outputs a pulse pressure signal of the same frequency and phase, forming a combined electromechanical-hydraulic vibration-impact excitation that precisely acts on the jammed interface. This operating method has significant advantages: First, it avoids rigid jamming, replacing static force with dynamic excitation, greatly reducing the risk of overload; second, it responds quickly, completing the unblocking process within seconds; third, it has intelligent adaptive capabilities, automatically matching parameters according to the degree of jamming, achieving gentle vibration for light jamming and heavy impact for heavy jamming; fourth, it has a safety exit mechanism to prevent invalid operations. The reason this method is the only option is that mechanical jamming is essentially static friction lock between microscopic contact surfaces. Only through the coordinated action of controllable periodic micro-displacement and transient fluid impact can its mechanical balance be efficiently disrupted without exceeding limits. Other methods cannot simultaneously guarantee safety, effectiveness, and controllability. Accordingly, when the normal free separation state is confirmed, it is necessary to switch to the preset constant-speed curve ascent mode until the maximum separation stroke is reached. This is because there is no abnormal resistance at this point, and the separation action should be completed in the most stable and efficient way, ensuring process control, stable cycle time, reasonable energy consumption, and avoiding unnecessary vibration interference with subsequent positioning accuracy. If vibration or pulse modes are still used in this state, it will actually damage the system stability.

[0099] like Figure 2 As shown, this embodiment also discloses an automatic mold changing control system for a block machine based on state feedback, including the following modules: The data acquisition module is used to collect real-time data on the temperature field distribution of the inner wall of the mold cavity and the real-time pressure value of the hydraulic locking system during the operation of the block machine. The data analysis module is used to perform progressive composite judgment logic based on the temperature field distribution data of the inner wall of the mold cavity to identify the safe boundary conditions for mold separation under the current working conditions. The data processing module is used to meet multiple control indicators that combine the real-time pressure value of the hydraulic locking system under the safety boundary conditions. Based on the entire process of pressure drop from the initial locking pressure to zero, it divides the pressure drop into multiple pressure relief stages and forms a target pressure drop trajectory to generate corresponding reverse compensation commands. The evaluation module is used to coordinate the execution of graded pressure relief and reverse vibration control based on the multiple pressure relief stages divided by the adaptive target pressure drop trajectory and the corresponding reverse compensation commands. After the pressure relief is completed, the module generates a mold separability confirmation signal based on the stability index of the mold throughout the pressure relief process. The control module is replaced in response to the mold separability confirmation signal. It collects the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor to construct a two-dimensional phase plane to obtain the slope of the motion trajectory. It determines the connection status of the mold connection parts and generates a targeted collaborative loosening scheme to remove the old mold and replace the mold to be replaced.

[0100] To implement the above embodiments, this application also proposes an electronic device. Please see [link to relevant documentation]. Figure 3 , Figure 3 This is a schematic diagram of the structure of the electronic device provided in an embodiment of this application. For example... Figure 3 As shown, the electronic device 500 includes: a processor 501 and a memory 502 communicatively connected to the processor 501; the memory 502 stores computer-executable instructions; the processor 501 executes the computer-executable instructions stored in the memory to implement the method provided in the foregoing embodiments.

[0101] To implement the above embodiments, this application also proposes a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the methods provided in the foregoing embodiments.

[0102] To implement the above embodiments, this application also proposes a computer program product, including a computer program that, when executed by a processor, implements the methods provided in the foregoing embodiments.

[0103] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for automatic mold changing control of a block machine based on state feedback, characterized in that, Includes the following steps: Step 1: Real-time acquisition of temperature field distribution data on the inner wall of the mold cavity and real-time pressure value of the hydraulic locking system during the operation of the block machine; Step 2: Based on the temperature field distribution data of the inner wall of the mold cavity, execute a progressive composite judgment logic to identify the safe boundary conditions for mold separation under the current working conditions; Step 3: Under the safety boundary conditions, multiple control indicators are combined with the real-time pressure value of the hydraulic locking system. Based on the entire process of pressure drop from the initial locking pressure to zero, multiple pressure relief stages are formed and together constitute the target pressure drop trajectory to generate corresponding reverse compensation commands. Step 4: Based on the multiple pressure relief stages divided by the adaptive target pressure drop trajectory and the corresponding reverse compensation commands, graded pressure relief and reverse vibration control are executed in a coordinated manner. After the pressure relief is completed, a mold separability confirmation signal is generated based on the stability index of the mold throughout the pressure relief process. Step 5: In response to the confirmation signal that the mold can be separated, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are collected to construct a two-dimensional phase plane to obtain the slope of the motion trajectory, determine the connection status of the mold connection part, and generate a targeted collaborative loosening scheme to remove the old mold and replace the mold to be replaced.

2. The automatic mold changing control method for block making machines based on state feedback according to claim 1, characterized in that, Based on the three-dimensional working condition safety verification matrix, a progressive composite judgment logic is executed to identify the safety boundary conditions for mold separation under the current working condition, including the following steps: Step 21: Obtain the maximum aggregate particle size parameter corresponding to the current production task and the mold cavity design clearance of the mold to be replaced. If the mold cavity design clearance is greater than 1.5 times the maximum aggregate particle size parameter, proceed to the next step; otherwise, determine that the safety boundary conditions are not met and terminate the replacement process. Step 22: Obtain the current excitation frequency and the natural frequency of the mold to be replaced. Determine the frequency difference based on the excitation frequency and the natural frequency. If the frequency difference is within the preset resonance avoidance range, it is determined that the safety boundary conditions are not met and the replacement process is terminated; otherwise, proceed to the next step. Step 23: Based on the temperature field distribution data of the inner wall of the mold cavity, obtain the stress amplitude of the maximum thermal stress concentration area, as well as the mold length and the difference between the average temperature of the mold cavity and the room temperature. Determine the shrinkage deformation based on the stress amplitude, mold length, and the difference between the average temperature of the mold cavity and the room temperature. Step 24: If the shrinkage deformation is less than the preset tolerance threshold, the safety boundary condition is satisfied; otherwise, the safety boundary condition is not satisfied and the replacement process is terminated.

3. The automatic mold changing control method for block making machines based on state feedback according to claim 2, characterized in that, To meet the aforementioned safety boundary conditions and combine multiple control indicators of the mold locking surface, the entire pressure drop process from the initial locking pressure to zero is divided into multiple pressure relief stages, which together constitute the target pressure drop trajectory, in order to generate corresponding reverse compensation commands, including the following steps: Step 31: Obtain contact stress distribution data at each coordinate position of the mold locking surface based on the shrinkage deformation; determine the pressure center coordinates based on all coordinate positions of the mold locking surface; determine the effective bearing area of ​​the locking surface based on the contact stress distribution data. Step 32: Determine the effective bearing area ratio and pressure center offset distance based on the effective bearing area and pressure center coordinates; Step 33: If the effective bearing area ratio is greater than or equal to a preset area ratio threshold and the pressure center offset distance is less than or equal to a preset distance threshold, then enter the uniform contact pressure relief mode. The entire pressure drop process from the initial locking pressure to zero is divided into three pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: rapid pressure relief stage, pressure holding and balancing stage, and slow pressure relief stage; the first target pressure drop trajectory is formed based on the continuous pressure intervals corresponding to the three pressure relief stages. Step 34, otherwise, enter the non-uniform pressure relief mode: divide the entire pressure drop process from the initial locking pressure to zero into two pressure relief stages and corresponding continuous pressure intervals; the pressure relief stages are: local stress equalization pressure relief stage and global rapid pressure relief stage; based on the continuous pressure intervals corresponding to the two pressure relief stages, a second target pressure drop trajectory is formed; Step 35: Based on the global rapid depressurization stage of the second target pressure drop trajectory and the three depressurization stages of the first target pressure drop trajectory, a basic reverse compensation command is generated; a dynamic reverse compensation command is generated based on the local stress equilibrium depressurization stage; both the dynamic reverse compensation command and the basic reverse compensation command consist of phase and amplitude.

4. The automatic mold changing control method for block making machines based on state feedback according to claim 3, characterized in that, The initial locking pressure is defined as the rapid pressure relief stage, the pressure is maintained for 3-5 seconds at the initial locking pressure, and the pressure is maintained for 0.6 times the initial locking pressure. The initial locking pressure is defined as the pressure holding and balancing stage. The initial locking pressure is defined as the slow pressure relief stage.

5. The automatic mold changing control method for block making machines based on state feedback according to claim 4, characterized in that, [Initial locking pressure, x times the initial locking pressure] is defined as the local stress equalization and pressure relief stage, and [x times the initial locking pressure, 0] is defined as the pressure holding and balance stage, where x is between 0.3 and 0.

5.

6. The automatic mold changing control method for block making machines based on state feedback according to claim 5, characterized in that, Based on the multiple depressurization stages defined by the adaptive target pressure drop trajectory, and the corresponding reverse compensation commands, graded depressurization and reverse vibration control are executed in coordination. After depressurization is completed, a mold separability confirmation signal is generated based on the mold's stability index throughout the depressurization process, including the following steps: Step 41: After entering each pressure relief stage, execute the pressure relief action and reverse compensation command, and monitor the real-time pressure value of the hydraulic locking system. If the real-time pressure value of the hydraulic locking system exceeds the continuous pressure range of the current pressure relief stage, pause the current pressure relief action, maintain the current real-time pressure value of the hydraulic locking system until it returns to the allowable range, and then continue the pressure relief and reverse compensation command operation of the next stage. Step 42: The angular displacement and angular velocity of the mold are synchronously sampled at a fixed control cycle to obtain the angular displacement data sequence and angular velocity data sequence arranged in time sequence; Step 43: After depressurization is complete and the hydraulic pressure drops to zero, execute the following composite judgment logic. If none of the composite judgment logics are satisfied, proceed to step 54; otherwise, determine that the mold posture is unstable and do not generate a mold separability confirmation signal. The composite judgment logic includes three sub-conditions: a. Determine the range of continuous sampling times corresponding to each pressure relief stage. If the absolute value of the angular displacement corresponding to any sampling time within the range of continuous sampling times is greater than the first safety threshold corresponding to the current pressure relief stage; b. If the absolute value of the angular velocity measured at each of at least three consecutive sampling times is greater than the preset second safety threshold. c. Within the sampling time range corresponding to the pressure holding balance section or the slow pressure relief section, identify local extreme points based on the angular displacement data sequence. Starting from the sampling time corresponding to the local extreme point, if there is another sampling time with an angular displacement value that has the opposite sign to the angular displacement value of the local extreme point, and the angular displacement change amplitude between the two is greater than the preset third safety threshold. Step 44: Calculate the energy entropy value of the angular displacement data sequence during the entire depressurization process. If the energy entropy value is less than the preset entropy threshold, generate the mold separability confirmation signal; otherwise, do not generate it.

7. The automatic mold changing control method for block making machines based on state feedback according to claim 6, characterized in that, In response to the confirmation signal that the mold can be separated, the displacement change rate of the lifting mechanism and the current fluctuation rate of the drive motor are collected to construct a two-dimensional phase plane to obtain the slope of the motion trajectory. The connection status of the mold connection parts is determined and a targeted coordinated loosening scheme is generated to remove the old mold and replace the mold to be replaced. The process includes the following steps: Step 51: In response to the mold separability confirmation signal, the real-time displacement value of the lifting mechanism and the real-time current value of the drive motor are collected in real time. Step 52: Perform central difference calculation on the displacement values ​​at three consecutive sampling times to obtain the displacement change rate at the current time. Calculate the relative change rate between the current value at the current sampling time and the current value at the previous time, and mark it as the current fluctuation rate. Step 53: Construct a two-dimensional phase plane based on the displacement change rate and current fluctuation rate; construct a data set based on the displacement change rate and current fluctuation rate and mark it as the working point; fit the working points of the five most recent consecutive sampling times into a motion trajectory and calculate the slope of the motion trajectory; Step 54: If the absolute difference between the current slope and the slope of the previous window is greater than the preset sudden change threshold and the current fluctuation rate is greater than the empirical threshold, then it is determined that there is mechanical jamming at the mold connection part; otherwise, it is determined that there is normal free separation at the mold connection part. Step 55: Based on the mechanical jamming, generate reciprocating motion commands for the lifting mechanism and hydraulic pulse pressure commands, and start and stop them synchronously for a duration of 1 to 3 seconds; if during the execution process, the displacement change rate returns to the normal range and the current fluctuation rate is less than or equal to the empirical threshold, then terminate the current loosening operation in advance. Step 56: Based on the normal free separation state, control the lifting mechanism to rise according to the preset constant speed curve until the maximum separation stroke is reached.

8. The automatic mold changing control method for block making machines based on state feedback according to claim 3, characterized in that, The process of obtaining the phase and amplitude of the local stress equalization and pressure relief stage is as follows: the phase of the local stress equalization and pressure relief stage is obtained through the pressure center coordinates, the current locking pressure is obtained, and the amplitude of the local stress equalization and pressure relief stage is obtained based on the pressure center offset distance and the current locking pressure.

9. The automatic mold changing control method for block making machines based on state feedback according to claim 7, characterized in that, The reciprocating motion command of the lifting mechanism is as follows: the servo driver of the lifting mechanism outputs a sine or triangular waveform position command. When the current fluctuation rate is between 0.1 and 0.3, the frequency of the sine or triangular waveform is set to 5 Hz; when it is between 0.3 and 0.6, the frequency of the sine or triangular waveform is set to 10 Hz; when it is greater than 0.6, the frequency of the sine or triangular waveform is set to 20 Hz; and the amplitude of the sine or triangular waveform is fixed at 0.3 mm.

10. The automatic mold changing control method for block making machines based on state feedback according to claim 9, characterized in that, Hydraulic pulse pressure command: Control the hydraulic proportional valve to output a pulse pressure signal with the same frequency and phase as the reciprocating motion command of the lifting mechanism; obtain the amplitude reference, generate an incremental term based on the current fluctuation rate, and obtain the amplitude of the pulse pressure based on the sum of the amplitude reference and the incremental term.

Images