Industrial equipment state monitoring method and system based on real-time data acquisition

Abstract

The application discloses an industrial equipment state monitoring method and system based on real-time data acquisition, and the method comprises the following steps: acquiring real-time response data corresponding to an industrial forming equipment; determining production process requirements, and selecting a plurality of pneumatic units to construct a symmetric reference unit group based on the production process requirements; acquiring unit response data of each symmetric reference unit group; establishing a gas-heat mapping relationship according to the unit response data; setting a thermocouple temperature measurement node in combination with the gas-heat mapping relationship, and acquiring a corresponding transient temperature difference signal; constructing a differential potential loop according to the transient temperature difference signal; utilizing the differential potential loop to perform zero-potential output on a differential potential signal of the thermocouple temperature measurement node of each pneumatic unit, and obtaining a zero-potential reference state; identifying an abnormal pneumatic unit in combination with the zero-potential reference state and the real-time response data; and generating a state monitoring report of the industrial forming equipment based on the abnormal pneumatic unit. The application can effectively improve the accuracy of local state monitoring of the forming equipment.

Description

Technical Field

[0001] This application relates to the field of industrial equipment information processing technology, and in particular to a method and system for industrial equipment status monitoring based on real-time data acquisition. Background Technology

[0002] With the rapid development of modern manufacturing, automated and intelligent production has become an important indicator of a country's industrial level. In the field of precision forming, the improvement of production efficiency and product quality places higher demands on the reliability of equipment. As a core production tool, the real-time monitoring of the stable operating status of forming equipment is crucial for ensuring production continuity, reducing maintenance costs, and improving product qualification rates.

[0003] Currently, the monitoring of the operating status of molding equipment mainly relies on multi-source sensor data acquisition and centralized signal processing and analysis. This typically involves deploying pressure sensors, vibration sensors, displacement sensors, or temperature sensors at key locations on the equipment to collect real-time data on load changes, structural vibrations, and thermal characteristics during the molding process. The data acquisition system then filters, extracts features, and applies thresholds to these signals, enabling a comprehensive assessment of the equipment's operating status and anomaly identification. However, these monitoring methods often struggle to capture subtle deterioration in critical functional units within the mold, such as pneumatic actuators. For example, under strong impact loads and complex working environments, the deterioration of the sealing performance of critical components like nitrogen springs inside the mold poses a potential risk of decreased molding accuracy and even equipment damage. Macroscopic data acquisition methods struggle to distinguish local anomalies, especially under strong impact loads and structural vibrations. Sensor signals are easily interfered with, masking early signs of localized deterioration and affecting the accuracy of the molding equipment's monitoring of localized equipment conditions.

[0004] There is currently no good solution to the above problems. Summary of the Invention

[0005] This application provides a method and system for monitoring the status of industrial equipment based on real-time data acquisition, which can improve the accuracy of monitoring the local status of forming equipment.

[0006] To achieve the above objectives, the embodiments of this application adopt the following technical solutions: Firstly, a method for monitoring the condition of industrial equipment based on real-time data acquisition is provided, the method comprising: When the slider in the industrial forming equipment comes into contact with the target sheet, the real-time response data of the industrial forming equipment is acquired. Determine the current production process requirements of the industrial forming equipment, and based on the production process requirements, select multiple pneumatic units within the industrial forming equipment to construct a symmetrical reference unit group; Obtain the element response data for each symmetric reference element group; Establish a gas-thermal mapping relationship based on the element response data of the symmetric reference element group; Thermocouple temperature measurement nodes are set in the aerodynamic unit corresponding to the symmetrical reference unit group based on the gas-thermal mapping relationship, and the corresponding transient temperature difference signal is obtained. A differential potential loop is constructed for the aerodynamic units within the same symmetrical reference unit group based on the transient temperature difference signal; When the sealing performance of the pneumatic units in the symmetrical reference unit group is consistent, the differential potential circuit is used to output the differential potential signal of the thermocouple temperature measurement node of each pneumatic unit to zero potential, so as to obtain the zero potential reference state. Abnormal pneumatic units are identified by combining zero potential reference state and real-time response data; A status monitoring report for industrial forming equipment is generated based on the abnormal pneumatic unit.

[0007] In another possible implementation of the first aspect, determining the current production process requirements of the industrial forming equipment and selecting multiple pneumatic units within the industrial forming equipment to construct a symmetrical reference unit group based on the production process requirements includes: The forming process type is determined based on the current production process requirements of the industrial forming equipment, and the forming stage information is determined through the forming process type. The forming stage information includes the blank holder stage, the drawing stage, and the final forming stage. By combining information from the forming stage, the process action mapping of the pneumatic units is performed to determine the dominant action type of each pneumatic unit; The aerodynamic units are grouped according to the dominant action type to obtain the equivalent unit set; During the operation of industrial forming equipment, the geometric parameters of the die surface within each equivalent unit set are acquired, and the material flow direction and flow resistance distribution characteristics of each equivalent unit set are determined based on the die surface geometric parameters. Initial symmetrical candidate unit groups are screened by combining material flow direction and flow resistance distribution characteristics, and symmetrical reference unit groups are determined by time-series screening of the initial symmetrical candidate unit groups based on forming stage information.

[0008] In another possible implementation of the first aspect, the method of selecting an initial symmetric candidate unit group by combining material flow direction and flow resistance distribution characteristics, and determining a symmetric reference unit group by combining forming stage information through time-series screening of the initial symmetric candidate unit group, includes: Collect the strain data of the target sheet material, and construct a material flow direction vector field based on the sheet material strain data and the material flow direction; A material flow resistance distribution field is constructed by combining the characteristics of flow resistance distribution, and the resistance symmetry axis is identified by using the material flow resistance distribution field; The aerodynamic unit is spatially symmetrically mapped by combining the drag symmetry axis, and the target unit group is selected according to the material flow direction vector field corresponding to the aerodynamic unit. When the forming stage is the edge pressing stage, the target unit group is subjected to time-series screening to obtain the first screening result; When the forming stage is the drawing stage, acquire the transient load data and displacement response data of the aerodynamic units within the target unit group; Transient load-displacement response curves are constructed by combining transient load data and displacement response data; The time warping distance and peak response phase difference are calculated based on the transient load-displacement response curve, and the target unit group is dynamically screened for consistency using the time warping distance and peak response phase difference to obtain the second screening result; When the forming stage is the final forming stage, determine the load response curve of the target unit group; The steady-state load mean and load convergence curve of the target unit group are extracted based on the load response curve. The consistency of the target unit group is verified based on the deviation of the steady-state load mean and the overlap of the load convergence curve to obtain the third screening result. The symmetric reference unit group was determined by combining the results of the first screening, the second screening, and the third screening.

[0009] In another possible implementation of the first aspect, the method of establishing a gas-thermal mapping relationship based on the element response data of a symmetric reference element group includes: The piston compression speed, effective cross-sectional area of ​​the cylinder, and cylinder pressure response sequence of the pneumatic unit are extracted from the unit response data. Using the moment of contact between the slider and the target plate as the time reference point, the peak range of compression speed is identified based on the piston compression speed of the pneumatic unit; The effective time domain window is determined by combining the time reference point and the peak compression rate range; The piston compression displacement change is determined by time integration based on the piston compression velocity within the effective time window, and the effective compression volume is calculated by combining it with the effective cross-sectional area of ​​the cylinder. By combining the effective compression volume and cylinder pressure response sequence, the gas adiabatic temperature rise sequence of the pneumatic unit within the effective time domain window is determined using a preset gas adiabatic state equation. The cylinder pressure response sequence and effective compression volume sequence are used as aerodynamic state characteristics, and the gas adiabatic temperature rise sequence is used as thermodynamic state characteristics. A gas-thermal state response model is constructed by combining the time-series dynamic correspondence between aerodynamic and thermodynamic state characteristics. The gas-thermal mapping relationship is determined by the gas-thermal state response model.

[0010] In another possible implementation of the first aspect, the method further includes: Extract the first aerodynamic state features and the second aerodynamic state features at the same time reference point from the first aerodynamic unit and the second aerodynamic unit in the symmetrical reference unit group, respectively. Input the first aerodynamic state features and the second aerodynamic state features into the gas-thermal state response model respectively to output the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence. Within the effective time window, calculate the dynamic temperature rise deviation sequence between the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence, and calculate the aerodynamic asymmetric pressure difference sequence between the first aerodynamic state characteristic and the second aerodynamic state characteristic. A symmetric error compensation function is constructed based on the aerodynamic asymmetric pressure difference sequence and the dynamic temperature rise deviation sequence; The symmetric error compensation function is used as a regularization constraint to iteratively update the model parameters of the gas-thermal state response model, resulting in the corrected gas-thermal state response model.

[0011] In another possible implementation of the first aspect, the method of constructing a differential potential loop for aerodynamic units within the same symmetrical reference unit group based on a transient temperature difference signal includes: The transient temperature rise response region corresponding to the outer wall of the pneumatic unit cylinder is identified by the transient temperature difference signal; In combination with the transient temperature rise response region, the first and second aerodynamic units in the symmetrical reference unit group are set with the same Seebeck coefficient; The transient temperature difference signal is converted into a differential potential signal using a thermoelectric conversion unit; The same polarity terminals of the first thermoelectric conversion unit and the second thermoelectric conversion unit are differentially connected, and the opposite polarity terminals are connected to the differential detection terminal to construct a differential potential loop for suppressing common-mode thermal interference.

[0012] In another possible implementation of the first aspect, the method further includes: Obtain the pressure regulation opening of each pneumatic unit; The potential signal output by the differential potential loop is acquired in time series, and the low-frequency thermal diffusion component of the time series potential signal is filtered out by frequency domain filtering to obtain a high-frequency differential potential signal. A reverse control mapping matrix is ​​constructed by combining the high-frequency differential potential signal and pressure regulation opening within the historical sampling period; Transient potential disturbance characteristics are constructed based on high-frequency differential potential signals, and the asymmetric deviation state between aerodynamic units is determined based on the transient potential disturbance characteristics. Based on the asymmetric deviation state, the deviation direction identifier and deviation intensity coefficient are extracted, and the deviation direction identifier and deviation intensity coefficient are added to the inverse control mapping matrix as control input variables. Based on the inverse control mapping matrix after adding control input variables, the corresponding feedforward compensation command is generated, and the pressure regulation opening of the pneumatic unit is adjusted using the feedforward compensation command.

[0013] In another possible implementation of the first aspect, the method of extracting the deviation direction identifier and deviation intensity coefficient based on the asymmetric deviation state includes: Extract the polarity characteristics and voltage peak value of the high-frequency differential potential signal within the current sampling period; Based on the polarity characteristics and the preset differential loop topology, the target dominant side with the highest heat generation in the symmetrical reference unit group is determined by matching, and the unit position attribute corresponding to the target dominant side is used as the deviation direction identifier. The transient thermal energy equivalent is obtained by time-domain integration of the absolute value of the high-frequency differential potential signal, so as to determine the transient thermal energy equivalent of the work done by the asymmetric compression of the aerodynamic unit. The deviation intensity coefficient is obtained by inputting the transient thermal energy equivalent and voltage peak value into the preset aerodynamic thermoelectric sensitivity function.

[0014] Secondly, this application provides a machine-readable storage medium storing instructions that cause a machine to execute the aforementioned industrial equipment status monitoring method based on real-time data acquisition.

[0015] Thirdly, this application provides an electronic device, comprising: The memory is configured to store instructions; and The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement the aforementioned method for monitoring the status of industrial equipment based on real-time data acquisition.

[0016] The above technical solution firstly involves simultaneously acquiring response data from the entire equipment and key local components at the instant the slider contacts the plate, enabling a refined representation of the operational status of each pneumatic unit under transient impact. Based on the forming process, a symmetrical reference unit group is selected to perform symmetrical comparative analysis of local structural areas, amplifying and identifying operational differences and improving the accuracy of local anomaly positioning. Data is independently acquired from each pneumatic unit and multi-source cross-verification is conducted. Response consistency analysis eliminates single-point errors, improving monitoring reliability and spatial coverage integrity. A gas-thermal mapping model is established to convert local pressure load changes into temperature responses, constructing a non-contact state monitoring channel and enhancing the interpretability of the mechanism. Based on this model, the temperature measurement node layout is optimized to match the temperature measurement positions with thermally sensitive and stress-critical areas, improving the sensitivity and signal-to-noise ratio of transient temperature difference signals. A differential potential loop is constructed to convert the local temperature response differences of different pneumatic units into potential difference signals, suppressing common-mode thermal interference and environmental temperature drift, achieving high-precision electrical performance for subtle thermal differences. By establishing a zero-potential reference when the sealing performance of each unit is consistent, the effects of installation errors and structural offsets are eliminated, improving the consistency and reliability of monitoring. The real-time collected local response data is differentially compared with the zero-potential reference to quantify the degree of operational deviation of each pneumatic unit, accurately detect and locate local sealing deterioration, aerodynamic imbalance, and structural abnormalities, and output a local condition monitoring report to provide data support for zonal maintenance and health management.

[0017] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0018] Figure 1 A flowchart illustrating an industrial equipment status monitoring method based on real-time data acquisition, provided in an embodiment of this application; Figure 2 This is a schematic diagram of a process for constructing a symmetric reference unit group, provided in an embodiment of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0020] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0021] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0022] Figure 1 The illustration shows a flowchart of an industrial equipment condition monitoring method based on real-time data acquisition according to an embodiment of this application. Figure 1 As shown in the figure, this application provides an industrial equipment status monitoring method based on real-time data acquisition, which may include the following steps.

[0023] S110. When the slider in the industrial forming equipment comes into contact with the target sheet, acquire the real-time response data corresponding to the industrial forming equipment. During the operation of industrial forming equipment, when the set stamping trajectory in the slider system descends and makes initial contact with the target sheet to form the stamping instant, the real-time response data of the forming equipment is synchronously acquired through a multi-source sensor array deployed in the slider mechanism, mold structure, and drive system. The real-time response data includes slider force data, mold reaction force data, machine body structure strain data, slider displacement and velocity change data, motor drive current data, and equipment vibration response data.

[0024] S120. Determine the current production process requirements of the industrial forming equipment, and based on the production process requirements, select multiple pneumatic units within the industrial forming equipment to construct a symmetrical reference unit group. The production work order basic information is read from the MES system or the equipment HMI to determine the current production process requirements of the industrial forming equipment. In this embodiment, the production process requirements refer to the set of process constraints and control targets related to the forming quality of the workpiece during the forming process. These requirements include at least one or more of the following: process type, target punching pressure range, slider speed curve parameters, blank holder force control strategy, and mold forming stage division information. These represent the comprehensive constraints on the equipment's stress state, motion state, and forming cycle time during the current processing. Next, based on the production process, multiple pneumatic units are selected within the industrial forming equipment to construct a symmetrical reference unit group. Specifically, firstly, the forming process type is determined based on the current production process requirements of the industrial forming equipment, and the corresponding forming stage information is determined according to the forming process type. The forming stage information includes at least one or more of the following: blank holder stage, drawing stage, and final forming stage. Secondly, the process action mapping of each pneumatic unit in the industrial forming equipment is performed in conjunction with the forming stage information to determine the dominant action type of each pneumatic unit in the current process. Based on the dominant action type, each pneumatic unit is classified and grouped to form multiple equivalent unit sets. During the operation of industrial forming equipment, the die surface geometric parameters corresponding to each equivalent unit set are acquired, and the flow direction and flow resistance distribution characteristics of the material during the forming process are determined based on the die surface geometric parameters. Subsequently, the equivalent unit set is screened for initial symmetrical candidate unit groups in combination with the material flow direction and flow resistance distribution characteristics to obtain candidate unit combinations that meet the constraints of spatial symmetry and flow consistency. Finally, the initial symmetrical candidate unit groups are screened and corrected for temporal constraints in combination with forming stage information to eliminate unit combinations that do not conform to the dynamic change characteristics of the current process stage, thereby determining the final symmetrical reference unit group, which is used as the benchmark reference structure for subsequent forming process state analysis and comparative evaluation.

[0025] S130. Obtain the element response data for each symmetric reference element group; In this embodiment, the symmetrical reference unit group refers to a set of units with symmetrical constraints formed after comprehensively screening multiple pneumatic actuators based on the spatial structural layout of the industrial forming equipment, the functional attributes of the pneumatic units, and the current forming process conditions. The symmetrical relationship includes at least one or more of geometric position symmetry, directional symmetry, and equivalent process load symmetry. Unit response data includes, but is not limited to, pneumatic pressure response data, actuation displacement response data, motion speed response data, mechanical load response data, and control electrical signal response data. Pneumatic pressure response data is used to represent the changes in internal pressure of the pneumatic unit over time during the forming process; actuation displacement response data is used to represent the displacement changes at the actuation end of the pneumatic unit; motion speed response data is used to represent the dynamic motion characteristics of the pneumatic unit during execution; mechanical load response data is used to represent the changes in output force or applied force of the pneumatic unit; and control electrical signal response data is used to represent the changes in voltage, current, or drive signal during the control execution process of the pneumatic unit.

[0026] S140. Establish the gas-thermal mapping relationship based on the element response data of the symmetric reference element group; Based on the unit response data of each pneumatic unit in the symmetric reference unit group during the forming process, the piston compression speed, effective cross-sectional area of ​​the cylinder, and cylinder pressure response sequence of the corresponding pneumatic unit are extracted. Taking the contact time between the slider and the target sheet as the time reference point, the piston compression speed of each pneumatic unit is analyzed in time series to identify the peak compression speed range. An effective time domain window is constructed by combining the time reference point and the peak compression speed range to limit the effective action stage in the pneumatic compression process. The piston compression speed is integrated over time within the effective time domain window to determine the piston compression displacement change, and the effective compression volume is calculated by combining the effective cylinder cross-sectional area. Furthermore, the effective compression volume and cylinder pressure response sequence are combined, and the gas adiabatic temperature rise sequence of the pneumatic unit within the effective time domain window is determined based on the preset gas adiabatic state equation. The cylinder pressure response sequence and the effective compression volume sequence are used as pneumatic state features, and the gas adiabatic temperature rise sequence is used as thermodynamic state features. The gas-thermal state response model is constructed by combining the time series dynamic correspondence between the pneumatic state features and the thermodynamic state features, and the gas-thermal mapping relationship of the pneumatic unit is determined based on the gas-thermal state response model.

[0027] S150. Based on the gas-thermal mapping relationship, thermocouple temperature measurement nodes are set in the aerodynamic unit corresponding to the symmetrical reference unit group, and the corresponding transient temperature difference signal is obtained. In this embodiment, the gas-thermal mapping relationship is a dynamic coupling correspondence established between aerodynamic state parameters and thermodynamic state parameters during the operation of the pneumatic unit. The aerodynamic state parameters include at least cylinder pressure response, effective compression volume, and piston compression process characteristic parameters; the thermodynamic state parameters include at least gas temperature rise changes, temperature response sequences, and heat change characteristics. The gas-thermal mapping relationship is used to represent the temperature evolution law caused by pressure and volume changes during gas compression in the pneumatic unit, and reflects the dynamic process of converting aerodynamic energy into thermal energy. Thermocouple temperature measurement nodes refer to temperature acquisition points deployed at key temperature measurement locations in the pneumatic unit. These nodes acquire temperature change information at corresponding locations by installing thermocouple sensors. Key temperature measurement locations are areas with sensitive response characteristics to aerodynamic state changes, determined based on the gas-thermal mapping relationship. Within the symmetrical reference unit group, each pneumatic unit has corresponding temperature measurement nodes with mirrored spatial positions to form symmetrical temperature measurement node pairs, used to acquire corresponding temperature response data and further construct transient temperature difference signals. In this embodiment, thermocouple temperature measurement nodes are set in the pneumatic unit corresponding to the symmetrical reference unit group based on the gas-thermal mapping relationship, including the following steps: A gas-thermal state response model is constructed by combining the time-series dynamic correspondence between aerodynamic and thermodynamic state characteristics to represent the coupling relationship between airflow and temperature rise distribution. The temperature rise gradient distribution of each aerodynamic unit under unit pressure change is calculated based on the gas-thermal mapping relationship, and the wall heat diffusion path and corresponding heat-sensitive area are determined based on the temperature rise gradient distribution. Symmetry difference analysis was performed on the thermally sensitive regions of each aerodynamic unit within the symmetrical reference unit group, and the target region with the largest thermal response difference was extracted as the preferred temperature measurement region. The local airflow vector field is determined based on the gas-thermal state response model, and the dominant airflow direction is obtained through the local airflow vector field. Within the preferred temperature measurement area, thermocouple temperature measurement nodes are determined based on the intersection of the dominant airflow direction and the wall heat diffusion path, and thermoelectric conversion units are set at the corresponding locations. The thermocouple temperature measurement nodes of the first and second pneumatic units within the symmetrical reference unit group are arranged in a spatially symmetrical manner to construct a symmetrical temperature measurement structure.

[0028] By constructing a gas-thermal state response model based on the temporal dynamic correspondence between aerodynamic and thermodynamic state characteristics, the temporal correlation between pressure response and volume change relative to temperature rise response can be analyzed, thus determining the temporal correspondence and its changing patterns between the aerodynamic change process and the thermodynamic response process. Subsequently, a mapping relationship between aerodynamic and thermodynamic state characteristics is established based on the temporal dynamic correspondence. Then, input and output data are modeled based on this temporal dynamic correspondence. By solving the model parameters, the model can characterize the dynamic influence of aerodynamic state changes on the thermodynamic state response. Through the gas-thermal state response model, the coupling relationship between airflow changes and temperature rise distribution can be expressed, thereby providing model support for subsequent aerodynamic unit state analysis and control strategy optimization.

[0029] In this embodiment, a gas-thermal mapping model between aerodynamic pressure and thermal response is first established to represent the temperature response characteristics of each aerodynamic unit under different pressure change conditions. Further, using a unit pressure change condition as a unified excitation input, a standardized pressure perturbation is applied to each aerodynamic unit to obtain the temperature response distribution data of the corresponding aerodynamic unit within the wall region, thus forming a temperature rise response field under unit pressure perturbation conditions. Based on the temperature rise response field, the spatial temperature gradient is calculated on the spatial discrete grid or finite element nodes to obtain the temperature rise gradient distribution corresponding to each aerodynamic unit, where the temperature rise gradient represents the rate and direction of temperature change in spatial location. Subsequently, based on the continuity and directional consistency of the gradient vector in the temperature rise gradient distribution, the heat transfer direction within the wall region is tracked and reconstructed to determine the main diffusion path of heat in the wall structure; simultaneously, combining the gradient magnitude and the pressure perturbation response intensity, the sensitivity of each spatial region is quantitatively evaluated to identify thermally sensitive regions with high response characteristics to pressure changes.

[0030] The aerodynamic units with symmetrical relationships in the structure are divided into symmetrical reference unit groups. Based on the thermal response results of each aerodynamic unit under unit pressure disturbance conditions, the corresponding thermally sensitive region distribution is obtained to represent the thermal response sensitivity of different spatial locations to pressure changes. Spatial symmetry mapping processing is performed on the thermally sensitive region distribution of each symmetrical aerodynamic unit, ensuring a one-to-one correspondence between the symmetrical units under a unified coordinate reference, thereby eliminating the influence of geometric position difference comparison analysis. Based on this, point-by-point difference calculation is performed on the thermal sensitivity values ​​at corresponding spatial locations to obtain the thermal response difference distribution field, representing the degree of deviation in thermal response characteristics between symmetrical units. Subsequently, spatial extreme value analysis or threshold screening processing is performed on the thermal response difference distribution field to identify the local region with the greatest difference, and this local region is determined as the target region. Finally, the target region is used as the preferred temperature measurement area for subsequent temperature sensor deployment or thermal state monitoring point setting, thereby achieving adaptive screening and optimized arrangement of key thermally sensitive locations based on symmetrical difference characteristics.

[0031] Based on the established gas-thermal state response model, pressure and temperature field distribution data within the target spatial region are acquired. State response analysis of local spatial units is then performed based on these pressure and temperature fields to represent the thermo-mechanical coupling response characteristics of each spatial location under aerodynamic disturbances. A local airflow vector field is constructed based on the coupling relationship between pressure and temperature gradients. The pressure gradient is used as the primary driving term, and the temperature gradient is introduced as a thermal coupling correction term to comprehensively reflect the gas flow behavior in space, thus obtaining the local airflow vector corresponding to each spatial sampling point. Subsequently, the spatial distribution of the local airflow vectors is calculated to form a local airflow vector field representing the flow characteristics within the target region. The vector direction represents the gas flow direction, and the vector magnitude represents the airflow intensity. Spatial aggregation and statistical processing are performed on the local airflow vector field, including weighted averaging, directional consistency analysis, or principal direction extraction, to eliminate the influence of local disturbances and random fluctuations, thereby determining the dominant airflow direction within the analysis region. The above method enables the inversion analysis of local airflow structure based on gas thermal state response information, thereby obtaining the dominant airflow direction information without the need for direct flow velocity measurement, providing data support and analysis basis for aerodynamic structure optimization design and thermal-fluid coupling state assessment.

[0032] Based on the identified preferred temperature measurement area, the distribution of the dominant local airflow direction and the wall heat diffusion path within this area are obtained. The dominant local airflow direction is obtained through directional convergence analysis of the local airflow vector field, while the wall heat diffusion path is obtained by tracing the heat conduction path along the direction of maximum change in the temperature gradient field. Furthermore, a spatial coupling analysis is performed on the dominant airflow direction and the wall heat diffusion path to identify their spatial intersection regions, which represent the concentrated coupling points of the pneumatic transport and heat conduction processes. Subsequently, the intersection regions are quantitatively evaluated. The intersection intensity is calculated based on the similarity or inner product relationship between the airflow direction vector and the heat diffusion direction vector, and the spatial location with the highest intersection intensity is selected as the target temperature measurement location. Finally, thermocouple temperature measurement nodes are set at the target temperature measurement locations, and thermoelectric conversion units are configured to convert and output temperature signals into electrical signals, thereby completing the optimized layout of key temperature measurement points based on the gas-thermal coupling characteristics.

[0033] The first and second aerodynamic units within the symmetrical reference unit group are identified as spatially symmetrical paired units, and a spatial correspondence is established between them. This symmetrical mapping relationship includes mirror symmetry mapping or rotational symmetry mapping. Within the first aerodynamic unit, based on the distribution of heat-sensitive areas, the coupling characteristics of the dominant airflow direction and the wall heat diffusion path, the optimal placement location of the thermocouple temperature measurement nodes is determined. This optimal placement location is then transformed to the corresponding spatial location in the second aerodynamic unit through the symmetrical mapping relationship to obtain the coordinates of the paired temperature measurement nodes. Subsequently, thermocouple temperature measurement nodes are set at the corresponding spatial locations in the first and second aerodynamic units, and a temperature measurement loop is constructed through lead connections and signal acquisition units, thus forming a symmetrical temperature measurement structure with one-to-one spatial correspondence. This method achieves a symmetrical temperature measurement node matching placement, ensuring spatial consistency and comparability of thermal response data between symmetrical units, and providing fundamental data support for subsequent thermal anomaly identification and condition assessment based on symmetrical difference analysis.

[0034] S160. Construct a differential potential loop for the aerodynamic units within the same symmetrical reference unit group based on the transient temperature difference signal. During the operation of industrial forming equipment, the temperature change sequence of each pneumatic unit is analyzed in time based on the acquired transient temperature difference signal. Combined with the structural layout parameters of the pneumatic unit, the temperature change characteristics are mapped to the corresponding spatial position on the outer wall of the cylinder to identify the transient temperature rise response area on the outer wall of the cylinder. Within the transient temperature rise response region, a first thermoelectric conversion unit and a second thermoelectric conversion unit are respectively deployed on the first and second pneumatic units within the same symmetrical reference unit group. The first and second thermoelectric conversion units are constructed using thermoelectric materials with the same Seebeck coefficient and satisfy the condition of mirror symmetry in spatial position. The first and second thermoelectric conversion units are used to perform thermoelectric conversion on the temperature changes at corresponding positions to convert the transient temperature difference signal into an electric potential signal. Furthermore, the same polarity terminals of the first and second thermoelectric conversion units are connected, and their opposite polarity terminals are respectively connected to the differential detection terminal to construct a differential electric potential loop. This allows the common-mode electric potential signals generated under the same ambient temperature change or overall temperature rise conditions to cancel each other out, retaining only the differential electric potential signal caused by the thermal response difference between the symmetrical pneumatic units, which is used to represent the asymmetric change characteristics of the pneumatic unit's operating state.

[0035] S170. When the sealing performance of the pneumatic units in the symmetrical reference unit group is consistent, the differential potential circuit is used to output the differential potential signal of the thermocouple temperature measurement node of each pneumatic unit to zero potential, so as to obtain the zero potential reference state. When the sealing performance of the pneumatic units within the symmetrical reference unit group is consistent, a differential potential loop is used to output a zero potential from the potential signals of the thermocouple temperature measurement nodes of each pneumatic unit. Under this consistent operating condition, each pneumatic unit is in an equivalent pneumatic load and equivalent thermal response state, and its temperature rise distribution tends to be consistent, thus making the temperature difference between each thermocouple temperature measurement node approach zero, and the corresponding differential potential signal also approaches zero. Based on this, the output result of the differential potential loop is defined as the zero potential reference state, used to represent the reference potential state of the system under conditions of no structural anomalies and no differences in sealing performance. For each differential loop, its output potential is adjusted to 0V through a hardware zeroing circuit. After establishing a zero potential reference, the system enters normal monitoring mode. When a pneumatic unit leaks, its temperature drops, and the corresponding differential loop outputs a negative potential signal. When the absolute value of the output signal exceeds a preset threshold, the system triggers an alarm. The magnitude of the signal is approximately proportional to the amount of leakage and can be used to assess the severity of the leak. By comparing the signals from multiple temperature measurement points, the specific location of the leak can also be accurately located.

[0036] S180, combining zero potential reference state and real-time response data to identify abnormal pneumatic units; Next, the differential potential signal corresponding to the thermocouple temperature measurement node of each pneumatic unit is acquired in real time as real-time response data. The real-time response data is then aligned with a pre-established zero potential reference state to obtain the potential deviation sequence of each pneumatic unit. Based on the potential deviation sequence, the deviation amplitude and deviation change characteristics of each pneumatic unit within the current sampling period are extracted to construct an anomaly judgment index representing the operating status of the pneumatic unit. The anomaly judgment index is compared with a preset anomaly threshold. When the anomaly judgment index exceeds the preset anomaly threshold, the corresponding pneumatic unit is determined to be in an abnormal state. The spatial location attribute of the abnormal pneumatic unit is determined by combining the spatial symmetry relationship of the symmetrical reference unit group, thereby realizing the anomaly identification and location of the pneumatic units of the forming equipment. The preset anomaly threshold can be obtained by statistically analyzing the standard deviation of the differential potential signal or the average deviation value between symmetrical units to represent the upper limit of the system's potential fluctuation under normal operating conditions.

[0037] S190, Generate status monitoring reports for industrial forming equipment based on abnormal pneumatic units.

[0038] Subsequently, the information of the identified abnormal pneumatic units is summarized to obtain the unit number, anomaly type, anomaly intensity coefficient, and anomaly occurrence time, and an anomaly information set is constructed. Next, based on the anomaly intensity coefficient, each abnormal pneumatic unit is graded to form different levels of anomaly state classification results. Further, combining the spatial structural relationship of the symmetrical reference unit group, each abnormal pneumatic unit is mapped to its corresponding spatial location area in the forming equipment to achieve visualized location of the anomaly. The temporal distribution characteristics and spatial correlations among multiple abnormal pneumatic units are analyzed to determine whether there are local or systemic anomaly features. Finally, based on the anomaly information set, anomaly level classification results, and spatial location mapping results, a structured status monitoring report is generated to represent the operating status and anomaly distribution of the forming equipment, thus providing a basis for equipment maintenance and operating status assessment.

[0039] In this embodiment, in a specific application scenario, this solution can be applied to automotive stamping production equipment. When the stamping slider contacts the body sheet to form the stamping moment, the system acquires the real-time response data of the industrial forming equipment in real time, and selects multiple pneumatic units to construct a symmetrical reference unit group according to the process stages such as edge pressing, drawing, and final forming. The system extracts unit response data such as piston compression speed, cylinder effective cross-sectional area, and cylinder pressure response from each symmetrical reference unit group, establishes a gas-thermal mapping relationship using the gas adiabatic state equation, and sets thermocouple temperature measurement nodes on the outer wall of the pneumatic unit to acquire transient temperature difference signals. Differential potential loops are constructed within the same symmetrical reference unit group through differential connections at the same polarity terminals. When the sealing performance of the pneumatic units within the group is consistent, a zero potential reference state is output. If a pneumatic unit causes temperature rise and differential potential deviation from the zero potential reference due to sealing leakage, abnormal air resistance, or pressure regulation deviation, the system identifies the abnormal pneumatic unit by combining real-time response data and generates a status monitoring report. If necessary, feedforward compensation commands are generated based on high-frequency differential potential signals and pressure regulation opening, thereby ensuring the balance of blank holder force and the stability of forming quality during the automotive forming process.

[0040] The above scheme first collects equipment response data at the instant the slider contacts the plate, capturing transient operating conditions and improving data timeliness and anomaly detection accuracy. Based on process requirements, multiple pneumatic units are selected to form a symmetrical reference group, enhancing detection targeting and narrowing the fault range through symmetrical comparison, thus improving positioning accuracy. Independent response data from each unit is collected to achieve multi-source verification, avoiding single-point failures and expanding monitoring coverage. A gas-thermal mapping relationship is established, converting pressure load changes into thermal effects, providing a physical correlation channel for non-electrical contact monitoring and enhancing mechanism interpretability. Based on this mapping, the temperature measurement node layout is optimized to match sampling positions with heat-sensitive areas, improving the signal-to-noise ratio and sensitivity of transient temperature difference signals. A differential potential loop is constructed to suppress common-mode thermal interference and environmental temperature drift, achieving high-precision potential representation of weak temperature differences and improving signal stability. Under the condition of consistent pneumatic unit sealing performance, a zero-potential reference state is established to achieve system self-calibration, eliminating manufacturing errors and initial biases, and ensuring reference consistency. By comparing and analyzing real-time response data with a zero-potential reference, the relative deviation of abnormal units is identified, improving the detection accuracy and false alarm resistance of anomalies such as seal deterioration and aerodynamic imbalance. The final output is an equipment status monitoring report, achieving a structured representation and traceable recording of operating status, providing data support for maintenance decisions and health management.

[0041] Figure 2 This is a schematic diagram illustrating a process for constructing a symmetric reference unit group, provided as an embodiment of this application. Figure 2 As shown, in one embodiment, the current production process requirements of the industrial forming equipment are determined, and based on these requirements, multiple pneumatic units are selected within the industrial forming equipment to construct a symmetrical reference unit group, including: S210. Determine the forming process type based on the current production process requirements of the industrial forming equipment, and determine the forming stage information through the forming process type. The forming stage information includes the blank holder stage, the drawing stage, and the final forming stage. S220. Combine the forming stage information to perform process action mapping on the pneumatic units and determine the dominant action type of each pneumatic unit. S230. Group the aerodynamic units according to the dominant action type to obtain the equivalent unit set; S240. During the operation of the industrial forming equipment, the geometric parameters of the die surface within each equivalent unit set are obtained, and the material flow direction and flow resistance distribution characteristics of each equivalent unit set are determined based on the die surface geometric parameters. S250. Based on the material flow direction and flow resistance distribution characteristics, an initial symmetrical candidate unit group is selected, and based on the forming stage information, the initial symmetrical candidate unit group is time-series-selected to determine the symmetrical reference unit group.

[0042] First, the process requirements information corresponding to the production task currently being performed by the forming equipment is obtained. This information includes the sheet material type, mold configuration parameters, and forming target parameters. Second, the process requirements information is analyzed to extract process features representing the stamping deformation method and process path. Further, based on these process features, the process type of the current forming process is identified, determining the corresponding forming process type. Then, according to the forming evolution process corresponding to the forming process type, the forming process is divided into the blank holder stage, the drawing stage, and the final forming stage, thereby generating forming stage information. In this embodiment, the forming stage information includes the blank holder stage, the drawing stage, and the final forming stage. The blank holder stage refers to the stage in which the blank holder applies a clamping force to the edge of the sheet metal, constraining the sheet metal in the initial stage of mold closure, thereby controlling the material flow state, preventing wrinkling, and stabilizing the boundary conditions of the sheet metal. The drawing stage refers to the stage in which, based on the blank holder constraint, the punch continues to descend, causing the sheet metal to undergo plastic flow in the mold cavity and gradually being drawn into the mold, realizing the overall extension of the material and the initial shaping of the shape. The final forming stage refers to the stage in which, after the basic completion of drawing, the punch continues to load, causing the sheet metal to conform to the final mold surface and completing the dimensional correction and local detail shaping.

[0043] Subsequently, it is determined whether the unit is in a working state at this stage, i.e., whether there is pressure output and whether stroke action is generated. For units in a working state, all possible roles they may undertake are matched from the process action type library. For example, in the blank-pressing stage, all blank-pressing cylinders may undertake blank-pressing action, and some local ejector rods may undertake pre-positioning action; in the drawing stage, blank-pressing cylinders may undertake drawing pressure holding action, and bottom ejector rods may undertake drawing assistance action; in the final forming stage, local blank-pressing cylinders may undertake shaping action, and nitrogen cylinders may undertake anti-springback action. For each unit's possible actions at this stage, its process contribution is calculated separately, and the one with the highest contribution is the dominant action type of the unit at this stage. The contribution can be determined by calculating the pressure contribution ratio, the action area ratio, and the material flow influence weight. The pressure contribution ratio is the proportion of the pressure output by the unit to achieve a certain action to its total output pressure at this stage; the action area ratio is the proportion of the die surface area covered by the unit to achieve a certain action to its total action area; the material flow influence weight can be calculated by using finite element simulation or historical data to calculate the influence coefficient of the unit's pressure change on the material flow velocity in the corresponding area. Based on their contribution, each pneumatic unit is sorted or categorized to determine the dominant role of each pneumatic unit in the current forming stage.

[0044] Pneumatic units are grouped according to their dominant function type to obtain an equivalent unit set. Specifically, the dominant function type of each pneumatic unit in the current forming stage is first determined, based on the mapping relationship of the pneumatic unit's process function in different forming stages. Next, pneumatic units with the same dominant function type are categorized to form corresponding functional category sets. Furthermore, pneumatic units within each functional category set are aggregated to construct corresponding equivalent unit sets, ensuring consistency or equivalence in the process function of pneumatic units within the same equivalent unit set. Finally, the overall function characteristics of the pneumatic units are uniformly represented based on the equivalent unit sets, thereby achieving equivalent expression and grouping control from the single pneumatic unit level to the functional set level.

[0045] During the operation of industrial forming equipment, the geometric parameters of the die surface within each equivalent unit set are acquired, and the material flow direction and flow resistance distribution characteristics of each equivalent unit set are determined based on the die surface geometric parameters. In this embodiment, the die surface geometric parameters include, but are not limited to, curvature distribution parameters, local slope parameters, normal vector variation parameters, surface roughness parameters, and surface height gradient parameters, which are used to represent the geometric shape and variation characteristics of the die surface in the spatial dimension. A three-dimensional digital model of the die surface is performed, and the die surface geometric parameters are discretized to obtain a spatial geometric field including curvature distribution, normal vector, and local slope. Then, spatial differentiation operations are performed on the die surface geometric field to construct a geometric gradient field, which is used to represent the spatial variation trend of the die surface. Subsequently, geometric parameters of the die surface are extracted within each equivalent unit set, including local curvature parameters, normal gradient parameters, relative equivalent gap parameters, local tilt angle parameters, and contact density parameters per unit area. These geometric parameters are then coupled and mapped to transform the geometric changes of the die surface into flow resistance contribution components. Specifically, a contraction-type resistance contribution is formed based on the equivalent gap parameters to represent the local resistance enhancement caused by flow channel compression; a directional resistance contribution is formed based on the curvature and normal gradient parameters to represent the additional resistance effect caused by flow path bending and directional adjustment; and a shear decomposition-type resistance contribution is formed based on the tilt angle parameters to represent the anisotropic resistance distribution caused by flow direction re-decomposition. Based on this, the resistance contribution components within the same equivalent unit set are spatially superimposed to form the local flow resistance distribution characteristics of the corresponding equivalent unit set. The resistance distribution is then recursively corrected through the geometric continuity between adjacent equivalent unit sets, ensuring that the resistance distribution reflects the continuous evolution characteristics of the flow propagating along the die surface, thus obtaining the spatial distribution result of flow resistance within the entire die surface area.

[0046] The die surface is divided into equivalent elements based on the geometric center datum or the axis of symmetry datum. Equivalent elements that satisfy the geometric mirror relationship are paired to form an initial symmetric candidate element group. Subsequently, material flow direction information is extracted for each equivalent element. The corresponding flow direction vector is determined by the displacement evolution trajectory or streamline direction of the material during the forming process. The flow direction consistency analysis is performed on the element pairs in the initial symmetric candidate element group. When the angle between the flow direction vectors of two elements meets the preset consistency condition, it is determined that they have symmetric consistency at the flow behavior level. The preset consistency condition may include flow direction consistency condition, flow resistance consistency condition, and flow path continuity condition, etc., which can be set according to the enterprise requirements. Furthermore, a flow resistance distribution feature is constructed for each equivalent element. This feature is based on the local geometric constraint information of the die surface, including the spatial combination of contraction resistance contribution, turning resistance contribution, and shear decomposition resistance contribution. The resistance distribution consistency of element pairs in the initial symmetric candidate element group is compared. When the resistance distribution morphology and gradient change characteristics of the two pairs are consistent, the candidate symmetric element group is retained. On this basis, forming stage information is introduced to perform time-series screening on the retained candidate symmetric element groups. The forming process is divided into the blanking stage, the drawing stage, and the final forming stage. The flow state and resistance response sequence of each equivalent element at different stages are determined to form the corresponding stage time-series feature sequence. By comparing the consistency of flow direction and flow resistance distribution of the same candidate symmetric element group in each forming stage, when the symmetric evolution relationship is consistent in all or key forming stages, the element group is determined as the symmetric reference element group, thus obtaining a standard reference element group that simultaneously satisfies geometric symmetry, flow behavior symmetry, and forming time-series stability.

[0047] By constructing a symmetrical reference unit group, the adaptive capability of the symmetrical reference unit group to dynamically optimize with the process can be realized. This can significantly improve the accuracy and robustness of pressure equalization control during the forming process of large automotive body panels, effectively suppress forming defects such as local cracking and wrinkling caused by uneven pressure distribution, improve surface quality and dimensional accuracy, and reduce the risk of mold wear and equipment failure.

[0048] In one embodiment, an initial symmetric candidate unit group is screened by combining material flow direction and flow resistance distribution characteristics, and a symmetric reference unit group is determined by combining forming stage information through time-series screening of the initial symmetric candidate unit group, including: S301. Collect the strain data of the target sheet material and construct a material flow direction vector field based on the sheet material strain data and the material flow direction. S302. Construct a material flow resistance distribution field by combining the characteristics of flow resistance distribution, and use the material flow resistance distribution field to identify the resistance symmetry axis; S303. Combine the drag symmetry axis to perform spatial symmetry mapping on each aerodynamic unit, and select the target unit group according to the material flow direction vector field corresponding to the aerodynamic unit; S304. When the forming stage is the edge pressing stage, the target unit group is subjected to time-series screening to obtain the first screening result. S305. When the forming stage is the drawing stage, acquire the transient load data and displacement response data of the pneumatic unit in the target unit group. S306. Construct transient load-displacement response curves by combining transient load data and displacement response data; S307. Calculate the time warping distance and peak response phase difference based on the transient load-displacement response curve, and use the time warping distance and peak response phase difference to perform dynamic consistency screening of the target unit group to obtain the second screening result; S308. When the forming stage is the final forming stage, determine the load response curve of the target unit group; S309. Based on the load response curve, extract the steady-state load mean and load convergence curve of the target unit group, and perform consistency verification on the target unit group according to the deviation of the steady-state load mean and the overlap of the load convergence curve to obtain the third screening result. S310. Combine the first screening results, the second screening results, and the third screening results to determine the symmetric reference unit group.

[0049] The strain state of the target sheet is collected during the forming process to obtain the sheet strain data covering the surface of the sheet. First, during or after forming, the degree of deformation in each region of the sheet is obtained by measurement, such as principal strain, secondary strain, thickness reduction, local strain concentration areas, etc., as well as the main direction of material extension or migration during deformation, such as flow from the edge pressing area to the drawing area, and expansion from the high resistance area to the low resistance area. Subsequently, the surface of the target sheet is spatially discretized to form several strain distribution nodes, each corresponding to a local region on the sheet surface and possessing corresponding strain data. Further, the local flow direction of the material is determined based on the strain change characteristics of each strain distribution node. This is achieved by calculating the strain gradient relationship between adjacent regions and correcting the flow direction in conjunction with the principal strain direction, thus obtaining a flow direction vector representing the local expansion trend of the material. Based on this, the flow direction vectors of each node undergo spatial consistency processing. By constraining and smoothing the continuity of the flow directions of adjacent nodes, abrupt changes in direction caused by local noise are eliminated, thereby forming a spatially continuously distributed material flow direction information. Finally, the flow direction vectors of all nodes are spatially integrated and reconstructed to construct a material flow direction vector field covering the entire target sheet area, used to represent the overall flow trend and spatial distribution characteristics of the material expanding from low-strain regions to high-strain regions during the forming process.

[0050] Based on the geometric information of the mold surface and the material flow characteristics, the flow resistance distribution parameters of each discrete region are obtained. These parameters include contraction resistance, turning resistance, and shear resistance. The resistance parameters of each discrete region are correlated and mapped with their spatial coordinates to form a discrete resistance dataset. Subsequently, the discrete resistance dataset is spatially continuous. A material flow resistance distribution field is constructed through neighborhood interpolation or spatial expansion, ensuring that each spatial location within the target region corresponds to a unique resistance distribution characteristic value, thus representing the degree of flow obstruction and its changing trend at different spatial locations. Multiple candidate symmetric reference axes are set within the material flow resistance distribution field, and the resistance distribution field is mirrored based on each candidate symmetric reference axis to obtain the resistance distribution matching relationship at corresponding positions on both sides of the axis. By comparing the consistency of the resistance values ​​and their gradient change characteristics at the mirrored positions, the symmetry of the resistance distribution of different candidate symmetric reference axes is evaluated. Finally, when the resistance distribution shape and changing trend on the left and right sides corresponding to a certain candidate symmetric reference axis have the highest consistency, that axis is determined as the material flow resistance symmetry axis.

[0051] Next, based on the resistance symmetry axis determined by the material flow resistance distribution area, spatial mirror mapping is performed on each aerodynamic unit within the target area to construct an initial symmetrical aerodynamic unit pair. Using the resistance symmetry axis as a spatial reference, each aerodynamic unit is mapped to its corresponding symmetrical position on the other side of the axis, and the corresponding mirror aerodynamic unit is determined, thus forming a candidate symmetrical unit group. The average material flow direction information inside each aerodynamic unit is extracted based on the material flow direction vector field to represent the overall flow behavior characteristics of the material within each aerodynamic unit. The consistency analysis of the flow direction characteristics of the two units in the candidate symmetrical unit group is performed to determine whether their material flow direction meets the symmetry consistency condition. Then, the environmental consistency check of the candidate symmetrical unit group is performed in conjunction with the material flow resistance distribution field to identify whether the flow resistance environment of the aerodynamic units on both sides is consistent, thereby eliminating non-true symmetrical relationship units caused by local resistance differences; finally, a target aerodynamic unit group that simultaneously satisfies the spatial mirror relationship, material flow direction consistency, and resistance distribution environment consistency is obtained.

[0052] When the forming stage is the edge-pressing stage, the target unit group is subjected to time-series screening to obtain the first screening result. In this embodiment, the first screening result is the set of target unit groups determined based on the consistency of material flow direction evolution and the consistency of flow resistance change during the time-series screening process in the edge-pressing stage. Specifically, it is a subset of target unit groups that meet the synchronous trend of flow direction change and the consistent trend of flow resistance change within the time range of the edge-pressing stage. Specifically, the forming process is divided into multiple stages, and the time range corresponding to the edge-pressing stage is determined. Within this time range, the time-series state data of each pneumatic unit in the target unit group is extracted. The time-series state data includes material flow direction information and flow resistance change information. Then, the time-step consistency analysis of the flow direction evolution sequence and flow resistance evolution sequence of each pneumatic unit in the target unit group in the edge-pressing stage is performed to determine whether the flow behavior and resistance change of two pneumatic units in the same unit group have synchronous evolution characteristics in the edge-pressing stage. When the target unit group meets the consistent trend of flow direction change and the consistent trend of flow resistance change within the edge-pressing stage, the target unit group is determined as a unit group that maintains consistent behavior in the edge-pressing stage, and the first screening result is output.

[0053] When the forming stage is the drawing stage, transient load data and displacement response data of the pneumatic units within the target unit group are acquired. In this embodiment, the transient load data is used to represent the time-varying characteristics of the contact pressure, friction force, and combined force experienced by the pneumatic unit during the drawing process. The displacement response data is used to represent the spatial position change and deformation evolution characteristics of the pneumatic unit under load. The transient load data includes, but is not limited to, contact pressure data, normal load data, tangential friction force data, combined force data, and load change rate data of the pneumatic unit during the drawing stage. The contact pressure data is used to represent the local interaction strength between the die and the sheet metal. The normal load data is used to represent the force component perpendicular to the die surface. The tangential friction force data is used to represent the interface sliding resistance characteristics. The combined force data is used to represent the comprehensive load level experienced by the pneumatic unit. The load change rate data is used to represent the transient evolution characteristics of the load over time. Displacement response data includes, but is not limited to, spatial displacement data, displacement increment data, displacement velocity data, displacement acceleration data, and displacement trajectory offset data of the pneumatic unit during the drawing stage. Among them, spatial displacement data is used to represent the positional change state of the pneumatic unit in three-dimensional space, displacement increment data is used to represent the displacement change amplitude between adjacent time steps, displacement velocity data is used to represent the speed of material flow response, displacement acceleration data is used to represent the acceleration or deceleration characteristics of displacement change, and displacement trajectory offset data is used to represent the degree of deviation of the actual movement path of the pneumatic unit from the expected flow path.

[0054] Transient load-displacement response curves are constructed by combining transient load data and displacement response data. The transient load data and displacement response data collected from each aerodynamic unit in the target unit group during the drawing stage are time-synchronized to ensure a one-to-one correspondence between load data and displacement data at the same time step. Then, the transient load data and displacement response data corresponding to each time step are paired and mapped to form a load-displacement correspondence data sequence. Based on this data sequence, transient load-displacement response curves are constructed to represent the dynamic coupling relationship between load changes and displacement response of the aerodynamic unit during the drawing stage, thereby achieving a continuous description of the mechanical response characteristics of the aerodynamic unit.

[0055] Subsequently, the transient load-displacement response curves corresponding to each aerodynamic unit in the target unit group are standardized to eliminate the influence of dimensional differences and sampling biases. Then, the time warping distance between the curves of different aerodynamic units is calculated using the dynamic time warping method to represent the degree of consistency between them in the overall dynamic response morphology. Simultaneously, the time information corresponding to the peak load response and peak displacement response is extracted from the transient load-displacement response curves of each aerodynamic unit, and the peak response phase difference between different aerodynamic units is calculated to represent the synchronicity of the key dynamic response occurrence moments. Based on this, the time warping distance and peak response phase difference are used as a joint criterion to perform dynamic consistency screening of the target unit group. When the target unit group has high consistency in the overall response morphology and maintains synchronous characteristics in the key response timing, it is determined to be a unit group that meets the dynamic consistency requirements, and a second screening result is output. In this embodiment, the target unit group in the second screening result exhibits a stable force-displacement coupling relationship during the transient load-displacement response process, its corresponding time warping distance is within the consistency constraint range, and the peak response phase difference meets the synchronicity requirements, thus indicating that the unit group has consistent force behavior and deformation response characteristics during the drawing stage. The target element group exhibits a stable force-displacement coupling relationship during the transient load-displacement response process. Its corresponding time warp distance is within the consistency constraint range and the peak response phase difference meets the synchronization requirement, which indicates that the element group has consistent stress behavior and deformation response characteristics during the drawing stage.

[0056] When the forming stage is the final forming stage, the load response curves of the target unit group are determined. First, the time interval corresponding to the final forming stage is determined. Within this time interval, load data of each pneumatic unit in the target unit group are collected and processed. The load data includes contact pressure, normal load, and combined force. Next, time-series analysis is performed on the collected load data to identify the response interval after the load change tends to stabilize or reaches its peak. The load change process within this response interval is continuously recorded and filtered to eliminate the influence of initial impact and local fluctuations on the results. Subsequently, the load data of each pneumatic unit in the final forming stage are mapped according to the time series or forming stroke sequence to construct the corresponding load response curves, representing the force evolution and stable convergence characteristics of the pneumatic unit in the final forming stage.

[0057] Subsequently, time-series analysis was performed on the load response curve to identify the steady-state region where the load tends to stabilize. Specifically, the load time-series data of the pneumatic unit during the final forming stage was acquired, and the load change between adjacent time steps was calculated to determine the magnitude of load change over time. Continuity analysis of the load change was performed to determine whether the load change tended to level off. When the load change remained within a preset range without significant abrupt changes within a continuous time interval, this time interval was determined as the starting interval for the load to enter a steady state, and the time range from this interval to the final forming end time was defined as the steady-state region. The preset range could be determined through laboratory calibration. Under standard experimental conditions, forming conditions with known acceptable forming quality were selected, and the load time-series data of the final forming stage of the forming process was collected. The steady-state region data after the load entered a steady state was extracted, and the distribution characteristic parameters of load fluctuation in the steady state were obtained. Based on the distribution characteristic parameters, the statistical range or upper limit of load change was determined, and this was used as the preset range for determining whether the load tended to stabilize. Within the steady-state range, the average load of each aerodynamic unit is calculated to obtain the steady-state load mean. Based on the steady-state load mean of each aerodynamic unit within the same target unit group, the degree of deviation is calculated to determine the consistency of the final stress state. The convergence process segment of the load transitioning from the fluctuating state to the steady-state range is extracted from the load response curve to form a load convergence curve. The load convergence curves of each aerodynamic unit within the same target unit group are compared and analyzed to determine the overlap and consistency of their convergence processes. The consistency of the target unit group is verified by combining the deviation of the steady-state load mean and the overlap of the load convergence curves. When the target unit group meets the consistency requirements in both the final stress level and the convergence process, it is selected as the third screening result.

[0058] Finally, a set operation is performed on the three types of screening results to determine the target unit group that simultaneously satisfies the consistency of material flow in the edge pressing stage, the consistency of dynamic response in the drawing stage, and the consistency of steady-state stress in the final forming stage. Based on this, the consistency of the obtained intersection results is checked to eliminate unstable unit groups that only meet the conditions in a single stage or local stage, thereby obtaining a symmetrical reference unit group that maintains consistent characteristics throughout the forming process, realizing a method for determining a stable symmetrical structure throughout the forming process.

[0059] By defining a set of symmetrical reference units, the real-time accuracy of pressure control and process stability during the forming process of automotive body panels can be significantly improved. This effectively suppresses sheet flow imbalance and forming defects caused by inconsistent responses of symmetrical units, thereby improving the forming quality and production yield of large and complex curved surface parts.

[0060] In one embodiment, establishing a gas-thermal mapping relationship based on the element response data of a symmetric reference element group includes: S410. Extract the piston compression speed, cylinder effective cross-sectional area, and cylinder pressure response sequence of the pneumatic unit from the unit response data. S420: Using the moment of contact between the slider and the target plate as the time reference point, the peak range of compression speed is identified based on the piston compression speed of the pneumatic unit. S430. Determine the effective time domain window by combining the time reference point and the peak compression rate range; S440. The piston compression velocity within the effective time domain window is integrated over time to determine the change in piston compression displacement, and the effective compression volume is calculated by combining the effective cross-sectional area of ​​the cylinder. S450: Combining the effective compression volume and cylinder pressure response sequence, the gas adiabatic temperature rise sequence of the pneumatic unit within the effective time domain window is determined using a preset gas adiabatic state equation. S460, The cylinder pressure response sequence and effective compression volume sequence are used as aerodynamic state characteristics, and the gas adiabatic temperature rise sequence is used as thermodynamic state characteristics. S470. Construct an aero-thermal state response model by combining the time-series dynamic correspondence between aerodynamic state characteristics and thermodynamic state characteristics; S480. Determine the gas-thermal mapping relationship through the gas-thermal state response model.

[0061] The piston compression speed, effective cross-sectional area of ​​the cylinder, and cylinder pressure response sequence of the pneumatic unit are extracted using unit response data. In this embodiment, unit response data refers to a set of multi-source dynamic response information collected from the pneumatic unit during the forming process. This data represents the comprehensive changes in the force and motion states of the pneumatic unit at different forming stages. It includes displacement response data, load response data, and time synchronization identification data. The displacement response data represents the trajectory and deformation process of the pneumatic unit in space. The load response data represents the dynamic changes in contact pressure, normal force, and combined load experienced by the pneumatic unit during operation. The time synchronization identification data aligns the correspondence between different types of response data under the same time reference. Based on the displacement response data of the pneumatic unit, the piston displacement change sequence over time is obtained, and the time derivative of the displacement sequence is calculated to determine the piston compression speed, thereby representing the dynamic motion characteristics of the pneumatic unit. Subsequently, based on the correspondence between the load response data of the pneumatic unit and the internal pressure of the cylinder, and combined with the structural parameters of the pneumatic unit or through inverse calculations using the coupling relationship between load and pressure, the effective cross-sectional area of ​​the cylinder is determined to represent the equivalent pressure-bearing area in which the pneumatic unit participates in the force action. At the same time, based on the pressure sensing device installed in the cylinder to obtain the pressure change data over time, or based on the load data and the effective cross-sectional area to infer the cylinder pressure, a cylinder pressure response sequence is constructed to represent the dynamic pressure change characteristics of the pneumatic unit during operation.

[0062] Using the moment of contact between the slider and the target plate as the time reference point, the peak compression speed range is identified based on the piston compression speed of the pneumatic unit. That is, the moment when the slider first contacts the target plate is used as the unified time reference point. Then, the rate of change of piston displacement with time is calculated based on the displacement response data of the pneumatic unit to obtain the piston compression speed curve. On this basis, time series analysis is performed on the piston compression speed curve, and a sliding time window is used to comprehensively evaluate the mean, maximum and fluctuation characteristics of the speed to identify the continuous time interval in which the piston compression speed is high and relatively stable, and this continuous time interval is determined as the peak compression speed range.

[0063] The effective time domain window is determined by combining the time reference point and the peak compression speed range. In this embodiment, the effective time domain window refers to the continuous time range used to represent the main dynamic response behavior of the pneumatic unit during the forming process, with the moment of contact between the slider and the target sheet as the time reference point and the peak compression speed range of the pneumatic unit piston as the reference point. This time range covers the key working stage of the pneumatic unit from accelerated compression to high-speed force and gradually tends to stabilize. Within this window, the displacement response, load response and pressure response of the pneumatic unit are all in the main change range. Specifically, firstly, the instant when the slider contacts the target plate is taken as the time reference point, defined as the time zero point, and used as the starting boundary of the effective time domain window. Then, based on the piston compression speed change curve of the pneumatic unit, the peak compression speed range is identified, and the end time of this peak range is determined, where the end time is the time point corresponding to the compression speed falling back from the peak to a preset proportional threshold. Based on the end time of the peak range, a preset delay time is introduced to extend the time boundary backward, thereby including the peak compression speed influence stage and the subsequent pressure stabilization transition stage in the analysis scope. Finally, using the time reference point as the starting boundary and the end time of the extended peak range as the ending boundary, an effective time domain window covering the initial response stage, rapid compression stage, and thermodynamic response evolution stage of pneumatic compression is constructed. In this embodiment, the preset proportional threshold for identifying the peak compression speed range and the preset delay time for extending the peak range boundary are both obtained based on statistical data from multiple sets of pneumatic unit response data under historical normal operating conditions. By performing time alignment processing on the piston compression speed change curves during multiple ramming processes, the main change segments near the peak compression speed in each response are extracted, and the speed percentage change range corresponding to the transition from rapid increase to significant decrease in compression speed is statistically analyzed. The boundary features with stable consistency in the statistical results are used as the basis for determining the current peak compression speed range. By analyzing the time evolution process of cylinder pressure response and gas temperature rise after the peak compression speed range in the above historical data, the time distribution characteristics corresponding to the slowdown of pressure change and the entry of temperature rise response into a stable transition stage are statistically analyzed, and the time extension range after the peak range is determined based on the time distribution characteristics, thereby obtaining a time extension boundary that matches the actual evolution process of aerodynamic and thermodynamic response. Finally, the peak compression speed range and its extended time range are both derived from the statistical regularity of historical data.

[0064] Next, time-series data of the piston compression velocity of the pneumatic unit are extracted within the time range corresponding to the effective time-domain window, and the time-series data is integrated to obtain the cumulative displacement change of the piston along the compression direction within the effective time-domain window. The integration operation represents the overall motion displacement characteristics of the piston during the pneumatic compression process. Further, the effective cross-sectional area of ​​the cylinder is obtained by combining the cylinder structural parameters, and based on the geometric correspondence between the piston compression displacement change and the effective cross-sectional area of ​​the cylinder, the volumetric compression change of the gas within the effective time-domain window is calculated. This is achieved by multiplying the effective cross-sectional area of ​​the cylinder by the piston compression displacement change within the effective time-domain window, resulting in the gas volumetric compression change, i.e., the effective compression volume, thus determining the effective compression volume. The effective compression volume represents the actual spatial changes of the gas during the critical compression stage.

[0065] By combining the effective compressible volume and cylinder pressure response sequence, the adiabatic temperature rise sequence of the gas within the effective time domain window of the pneumatic unit is determined. This involves constructing an instantaneous gas volume sequence based on the previously calculated change in effective compressible volume, which can be arranged chronologically. Furthermore, by combining the pressure response sequence obtained from the cylinder pressure sensor, the dynamic relationship between gas pressure and volume within the effective time domain window is determined. Further, using the gas pressure, volume, and temperature at the initial moment as a reference state, and based on a preset gas adiabatic state equation, the correspondence between pressure, volume, and temperature is established. This preset gas adiabatic state equation represents the gas approximately satisfying adiabatic change conditions during rapid compression. The expression for the preset gas adiabatic state equation is shown below: Where P is the cylinder pressure; V is the instantaneous gas volume; γ is the adiabatic index (approximately 1.4 for air); and C is a constant that remains unchanged during the same adiabatic process. C is determined by the product of the initial pressure and the initial volume and is used to represent the state invariant of the gas during the same adiabatic process.

[0066] Subsequently, the instantaneous temperature value of the gas at each sampling moment is calculated based on the synchronous change relationship between pressure and volume, and the temperature change relative to the initial state is further calculated, thereby forming the gas adiabatic temperature rise sequence of the pneumatic unit within the effective time domain window. The gas adiabatic temperature rise sequence is used to represent the thermodynamic response evolution characteristics of the gas during the compression process.

[0067] Next, the cylinder pressure response sequence and the effective compression volume sequence are used as pneumatic state features. The cylinder pressure response sequence reflects the dynamic pressure response of the pneumatic unit over time within the effective time window, while the effective compression volume sequence represents the instantaneous volume change of the gas caused by piston movement during compression. Together, they describe the pressure-volume coupling change characteristics of the pneumatic unit. Furthermore, the adiabatic temperature rise sequence of the gas, calculated based on a preset adiabatic gas equation of state, is used as a thermodynamic state feature. This sequence reflects the dynamic temperature evolution of the gas during compression due to volume compression and pressure changes, thus representing the thermodynamic response process of the pneumatic unit within the effective time window.

[0068] The cylinder pressure response sequence, effective compression volume sequence, and gas adiabatic temperature rise sequence are aligned along their time axes to ensure comparability of various state characteristics under a unified time reference. In this embodiment, the three time-aligned sequences are denoted as P(tk), V(tk), and T(tk), respectively, where tk is the k-th sampling time after alignment. Subsequently, the gas adiabatic compression equation is used as the basic physical framework to construct a gas thermal state response model. Specifically, the following steps are taken: using the start time tstart of the pneumatic unit's operation as a reference, the active gas compression segment before the pressure response reaches steady state is selected as the effective time domain window [ta, tb]. Within this window, the gas volume changes significantly, and the thermodynamic response characteristics are obvious. Next, based on the fundamental thermodynamic laws of adiabatic processes, at any aligned time tk within the effective time domain window, the gas temperature rise change can be directly calculated from the cylinder pressure and effective compression volume at the corresponding time using the following deterministic functional relationship, the specific expression of which is shown below: Where T(P(tk),V(tk)) represents the thermodynamic state feature calculated from the aerodynamic state features at the aligned sampling time tk within the effective time domain window, i.e., the absolute temperature at that time, which is a function value determined by the two independent variables of pressure and volume. P(tk) is the pressure value in the acquired cylinder pressure response sequence; V(tk) is the volume value in the effective compression volume sequence, which is another input aerodynamic state feature; T ref The reference absolute temperature is the starting time of the effective time domain window as the reference reference point; P ref Is with T ref The reference absolute pressure at the same reference reference time is selected from the cylinder pressure response sequence at that moment as the reference starting point for pressure changes. γ is the adiabatic index of the gas, also called the specific heat ratio, and for diatomic gases such as air, a typical value is about 1.4. It is a known physical constant characterizing the thermodynamic properties of a gas.

[0069] The above mapping relationship is then defined as a pneumatic-thermal state response model. The inputs to the model are the aligned aerodynamic state characteristics—cylinder pressure P(tk) and effective compression volume V(tk), where V(tk) is implicit in the reference quantity P. ref During the establishment process, or through the adiabatic relationship PV γ The simultaneous solution of `=const` outputs the aligned thermodynamic state characteristics, specifically the gas adiabatic temperature rise sequence T(tk). This involves analyzing the temporal correlation between pressure response and volume change relative to temperature rise response based on the temporal relationship between aerodynamic and thermodynamic state characteristics. This analysis aims to determine the temporal correspondence and its changing patterns between the aerodynamic and thermodynamic response processes. Subsequently, based on this temporal dynamic correspondence, a mapping relationship is established between aerodynamic and thermodynamic state characteristics, allowing gas temperature rise changes to be represented by the cylinder pressure response sequence and effective compression volume sequence at corresponding times. Finally, a gas-thermal state response model is constructed using this mapping relationship to represent the evolution of the thermodynamic response of the aerodynamic unit within the effective time window caused by aerodynamic changes, thereby achieving a quantitative description of the coupling relationship between the aerodynamic and thermodynamic states.

[0070] Under a unified time reference, the aerodynamic state characteristic sequence is input into the gas-thermal state response model to obtain the corresponding thermodynamic response output sequence, thus determining the influence path and mechanism of pressure and volume changes on temperature rise. The mapping relationship is structured so that the corresponding gas temperature rise state can be deduced at any given time based on the cylinder pressure response and effective compression volume, thereby forming a deterministic correspondence between aerodynamic and thermodynamic states, i.e., the gas-thermal mapping relationship. This relationship represents the quantitative correlation rules between aerodynamic changes and thermodynamic responses of the aerodynamic unit within the effective time window.

[0071] By establishing a gas-thermal mapping relationship, the prediction accuracy and real-time sensing capability of the thermal effect of the pneumatic system during the forming process can be significantly improved, effectively avoiding pressure drift and decreased forming stability caused by temperature rise, and providing a reliable basis for gas-thermal coordinated control for high-precision forming of automotive body panels.

[0072] In one embodiment, the method further includes: S510. Extract the first aerodynamic state features and the second aerodynamic state features at the same time reference point from the first aerodynamic unit and the second aerodynamic unit in the symmetrical reference unit group, respectively. S520. Input the first aerodynamic state features and the second aerodynamic state features into the gas-thermal state response model respectively, and output the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence. S530. Within the effective time domain window, calculate the dynamic temperature rise deviation sequence between the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence, and calculate the aerodynamic asymmetric pressure difference sequence between the first aerodynamic state characteristic and the second aerodynamic state characteristic. S540. Construct a symmetric error compensation function based on the aerodynamic asymmetric pressure difference sequence and the dynamic temperature rise deviation sequence; S550. The symmetric error compensation function is used as a regularization constraint term to iteratively update the model parameters of the gas-thermal state response model, resulting in the corrected gas-thermal state response model.

[0073] The first and second aerodynamic state features at the same time reference point are extracted from the first and second aerodynamic units in the symmetrical reference unit group, respectively. First, a unified time reference point is determined based on key triggering events in the forming process so that the aerodynamic units in the symmetrical reference unit group can achieve time alignment under the same time reference system. Further, at the time corresponding to the unified time reference point, aerodynamic parameters such as cylinder pressure response, effective compression volume, and piston motion state of the first aerodynamic unit are synchronously collected and extracted to form the first aerodynamic state features. At the same time, the corresponding aerodynamic parameters of the second aerodynamic unit at the same time reference point are synchronously collected and extracted to form the second aerodynamic state features. By synchronously acquiring the state data of aerodynamic units at different symmetrical positions at the same time reference point, a unified representation of the aerodynamic response features of each aerodynamic unit under the same working condition is achieved.

[0074] The first aerodynamic state feature and the second aerodynamic state feature are respectively input into the gas-thermal state response model, and the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence are output. In other words, the first aerodynamic state feature and the second aerodynamic state feature are time-aligned and data-standardized to ensure that the two sets of aerodynamic state features are comparable under the conditions of unified time reference and unified dimensions. Subsequently, the processed first aerodynamic state characteristics are input into a pre-constructed aero-thermal state response model. The thermodynamic response process of the first aerodynamic unit at each time sampling point is deduced through the aero-thermal state response model, resulting in a first predicted adiabatic temperature rise sequence corresponding to the first aerodynamic state characteristics. Simultaneously, the processed second aerodynamic state characteristics are input into the aero-thermal state response model. Based on the same model structure and model parameters, the thermodynamic response process of the second aerodynamic unit is deduced, resulting in a second predicted adiabatic temperature rise sequence corresponding to the second aerodynamic state characteristics. By performing model deductions on the aerodynamic state characteristics of different symmetrical aerodynamic units, independent prediction outputs of the thermodynamic response characteristics of each aerodynamic unit are achieved, thus providing basic data support for subsequent thermodynamic response difference analysis and aero-thermal consistency assessment based on symmetrical reference unit groups.

[0075] Next, within the effective time domain window, the first and second predicted adiabatic temperature rise sequences are aligned along the time axis. At each time sampling point, the temperature rise values ​​at corresponding moments are calculated to obtain a dynamic temperature rise deviation sequence that changes over time. This sequence represents the dynamic evolution of the thermodynamic response differences between symmetrical aerodynamic units. Further, cylinder pressure response data are extracted from the first and second aerodynamic state characteristics, and the pressure response values ​​at the same time sampling points are calculated to form an aerodynamic asymmetric pressure difference sequence. This sequence represents the differences in aerodynamic input at the symmetrical aerodynamic units. By simultaneously constructing the dynamic temperature rise deviation sequence and the aerodynamic asymmetric pressure difference sequence within the effective time domain window, a temporal consistency representation of the aerodynamic state differences and thermodynamic response differences of the aerodynamic units is achieved, thus providing fundamental data support for subsequent aero-thermal consistency analysis and symmetry assessment.

[0076] Within a unified time reference and effective time window, the aerodynamic asymmetric pressure difference sequence and the dynamic temperature rise deviation sequence are time-aligned to form a one-to-one correspondence between input and output data pairs. Subsequently, using the aerodynamic asymmetric pressure difference sequence as the input variable and the dynamic temperature rise deviation sequence as the output variable, a sample set of mapping relationships between the two is constructed. Based on this sample set, a function fitting analysis of the input-output relationship is performed. The parameters can be solved using a least-squares fitting method to ensure that the fitting function satisfies the constraint of minimizing the sum of squared input-output errors overall. Specifically, using the aerodynamic asymmetric pressure difference sequence as the independent variable and the dynamic temperature rise deviation sequence as the dependent variable, an error function is constructed. The parameters of the fitting function are determined through iterative optimization or analytical solution, enabling the fitting function to represent the corresponding change relationship between the aerodynamic asymmetric pressure difference and the dynamic temperature rise deviation. This achieves a quantitative description of the error propagation law of the symmetric aerodynamic unit, obtaining a mapping function model representing the influence of aerodynamic asymmetry changes on the thermodynamic response deviation, allowing changes in aerodynamic asymmetric pressure difference to be derived correspondingly to changes in dynamic temperature rise deviation. By solving and optimizing the parameters of the mapping function model, a symmetrical error compensation function is finally formed to describe the error transmission relationship between symmetrical aerodynamic units, thereby realizing the quantitative representation and compensation correction of the thermodynamic response deviation caused by aerodynamic asymmetry.

[0077] A symmetric error compensation function is used as a regularization constraint to iteratively update the parameters of the gas-thermal state response model. The deviation between the predicted output of the gas-thermal state response model and the actual adiabatic temperature rise of the gas is used as a data fitting error term to represent the model's fit to the thermodynamic response data. The symmetric error compensation function, constructed based on the aerodynamic asymmetric pressure difference sequence and the dynamic temperature rise deviation sequence, is introduced as a regularization constraint term to represent the inherent constraint relationship between aerodynamic changes and thermodynamic response deviations under aerodynamic asymmetry. Subsequently, the data fitting error term and the regularization constraint term are weighted and combined to construct a total loss function. The parameters of the gas-thermal state response model are iteratively optimized and updated based on the total loss function, so that the model satisfies the constraint relationship represented by the symmetric error compensation function while reducing prediction errors. Finally, a modified gas-thermal state response model is obtained, which balances data fitting consistency and symmetry constraint requirements when outputting the predicted adiabatic temperature rise of the gas, thereby improving the model's accuracy and stability in representing the gas-thermal coupling response process of the aerodynamic unit.

[0078] By iteratively updating the model parameters, a corrected gas-thermal state response model is obtained, which can significantly improve the model's prediction accuracy and long-term stability, effectively reduce pressure control deviation and forming quality fluctuations caused by uneven temperature distribution, and enhance the balance and reliability of multi-pneumatic unit collaborative control during the forming process of automotive body panels.

[0079] In one embodiment, a differential potential loop is constructed for the pneumatic units within the same symmetrical reference unit group based on the transient temperature difference signal, including: S610: Identify the transient temperature rise response area corresponding to the outer wall of the pneumatic unit cylinder through transient temperature difference signal; S620, combining the transient temperature rise response region, the first and second aerodynamic units in the symmetrical reference unit group are set with the same Seebeck coefficient as the first and second aerodynamic units. S630: The transient temperature difference signal is converted into a differential potential signal using a thermoelectric conversion unit; S640. The same polarity terminal of the first thermoelectric conversion unit and the same polarity terminal of the second thermoelectric conversion unit are differentially connected, and the opposite polarity terminal is connected to the differential detection terminal to construct a differential potential circuit for suppressing common-mode thermal interference.

[0080] Multiple temperature acquisition points are deployed at different spatial locations on the outer wall of the pneumatic unit cylinder to obtain temperature data that changes over time at each point. Further, the temperature data at each point within a continuous sampling period are differentially processed to obtain the corresponding transient temperature difference signal, which represents the rate of temperature change over time at each spatial location. Subsequently, the transient temperature difference signals at each point are analyzed, and points whose transient temperature difference changes exceed preset criteria are selected as high-response points. Further, spatial neighborhood association analysis and clustering are performed on the high-response points to determine the boundaries of high-response regions with spatial continuity, thereby identifying the transient temperature rise response region corresponding to the outer wall of the pneumatic unit cylinder. The transient temperature rise response region represents the spatial distribution characteristics of the thermal response of the cylinder outer wall under transient operating conditions and provides basic data support for subsequent thermal field evolution analysis and gas-thermal coupling relationship modeling.

[0081] By combining the transient temperature rise response region, a first thermoelectric conversion unit and a second thermoelectric conversion unit with the same Seebeck coefficient are set for the first and second pneumatic units in the symmetrical reference unit group. This means that the thermal response spatial distribution range of the pneumatic unit cylinder outer wall is determined based on the aforementioned identified transient temperature rise response region, and the first and second pneumatic units are symmetrically divided within the thermal response space. A first thermoelectric conversion unit is placed within the corresponding transient temperature rise response region of the first pneumatic unit, and a second thermoelectric conversion unit is placed at the corresponding symmetrical position of the second pneumatic unit. The first and second thermoelectric conversion units are made of the same thermoelectric material and have the same Seebeck coefficient to ensure consistent thermoelectric potential output characteristics under the same temperature difference conditions. Through this setup, the temperature changes of the first and second pneumatic units within the same thermal response region can be represented by an electrical signal with consistent thermoelectric conversion characteristics, thus providing a unified and comparable measurement basis for subsequent analysis of the thermal response differences of the symmetrical pneumatic units and modeling of the gas-thermal coupling relationship.

[0082] The transient temperature difference signal is converted into a differential potential signal using a thermoelectric conversion unit. First, based on the first and second thermoelectric conversion units arranged in the symmetrical reference unit group, the transient temperature difference signals of the first and second pneumatic units in the transient temperature rise response region are obtained respectively. Further, the transient temperature difference signal is converted into the corresponding first and second thermoelectric potential signals based on the Seebeck effect using the first and second thermoelectric conversion units, where the thermoelectric potential signal and the temperature difference signal satisfy a thermoelectric conversion relationship. Subsequently, the first and second thermoelectric potential signals are differentially processed to obtain the differential potential signal, which represents the electrical signal difference characteristics of the symmetrical pneumatic unit during the transient thermal response process.

[0083] The same-polarity terminals of the first and second thermoelectric conversion units are differentially connected, and the opposite-polarity terminals are connected to a differential detection terminal to construct a differential potential loop for suppressing common-mode thermal interference. Specifically, the first and second thermoelectric conversion units are respectively set within the transient temperature rise response regions corresponding to the first and second aerodynamic units of the symmetrical reference unit group, and the positive and negative output terminals of each thermoelectric conversion unit are determined. Further, the same-polarity terminals of the first and second thermoelectric conversion units are connected accordingly to form a common reference potential node, and the opposite-polarity terminals of each thermoelectric conversion unit are led out as differential output terminals. Subsequently, the differential output terminals are connected to a differential amplification unit. In this embodiment, the differential amplification unit can be a differential amplifier or a differential voltage detection module, which has a non-inverting input terminal, an inverting input terminal, and a signal output terminal. The potential difference signal generated by the opposite polarity terminals of the first and second thermoelectric conversion units is used as a differential input signal and connected to the in-phase and inverting input terminals respectively. It is amplified and output by a differential amplifier or module to obtain the potential difference signal between the first and second thermoelectric conversion units. By constructing a differential potential loop through the above connection method, the common-mode thermal signals generated under the same environmental thermal disturbance are mutually canceled in the circuit, and only the differential potential signal caused by the thermal response difference of the symmetrical aerodynamic units is retained. This effectively suppresses common-mode thermal interference and improves the stability and measurement accuracy of the thermal response difference detection of the symmetrical aerodynamic units.

[0084] By constructing a differential potential loop, the accuracy and reliability of transient temperature difference signal detection can be significantly improved, avoiding misjudgments caused by environmental thermal disturbances, and providing high-fidelity sensor data support for thermal state monitoring and pressure equalization control of pneumatic systems in industrial forming equipment.

[0085] In one embodiment, the method further includes: S710, Obtain the pressure regulation opening of each pneumatic unit; S720: The potential signal output by the differential potential circuit is acquired in time series, and the low-frequency thermal diffusion component of the time series potential signal is filtered by frequency domain filtering to obtain a high-frequency differential potential signal. S730: Construct an inverse control mapping matrix by combining the high-frequency differential potential signal and pressure regulation opening within the historical sampling period; S740: Construct transient potential disturbance characteristics based on high-frequency differential potential signals, and determine the asymmetric deviation state between aerodynamic units based on the transient potential disturbance characteristics; S750. Extract the deviation direction identifier and deviation intensity coefficient based on the asymmetric deviation state, and add the deviation direction identifier and deviation intensity coefficient as control input variables to the inverse control mapping matrix; S760 generates corresponding feedforward compensation commands based on the inverse control mapping matrix after adding control input variables, and uses the feedforward compensation commands to adjust the pressure regulation opening of the pneumatic unit.

[0086] First, the pressure regulation opening degree of each pneumatic unit is obtained. In this embodiment, the pressure regulation opening degree indicates the extent to which the pneumatic unit regulates the gas pressure in its current operating state. The pressure regulation execution degree parameter can be obtained based on the control information or status feedback information of the pressure regulation actuator corresponding to each pneumatic unit.

[0087] Secondly, the potential signal output from the differential potential loop is acquired in a time series, and the low-frequency thermal diffusion component is filtered out by frequency domain filtering to obtain a high-frequency differential potential signal. This involves continuous time sampling of the potential signal output from the differential potential loop to obtain a time-varying potential signal, followed by frequency domain transformation to convert it from a time-domain signal to a frequency-domain signal, which can be achieved using the Fast Fourier Transform (FFT) method. Subsequently, low-frequency components related to the thermal diffusion process and ambient temperature rise drift are identified in the frequency-domain signal, and these components are suppressed or filtered out by setting frequency domain filtering conditions, while retaining the high-frequency components representing transient thermal response changes. Finally, the filtered frequency-domain signal undergoes inverse transformation to recover the high-frequency differential potential signal, which is used to represent the rapid thermal response changes of the pneumatic unit under transient conditions. This effectively suppresses low-frequency thermal diffusion interference and the influence of ambient thermal drift, improving the accuracy and stability of the differential potential signal in representing transient thermal response characteristics.

[0088] An inverse control mapping matrix is ​​constructed by combining the high-frequency differential potential signal and pressure regulation opening within historical sampling periods. The pressure regulation opening sequence and high-frequency differential potential signal sequence corresponding to each pneumatic unit are acquired within multiple historical sampling periods, and time alignment is performed on the two types of sequences to form historical sample data pairs with a one-to-one correspondence. Furthermore, the pressure regulation opening sequence is used as the input variable, and the high-frequency differential potential signal sequence is used as the output variable to structure the historical sample data and construct an input-output relationship sample set. Subsequently, the pressure regulation opening sequence acquired within the historical sampling period is used as input data, and the high-frequency differential potential signal sequence acquired at the corresponding time is used as output data. The input and output data are time-aligned to form sample data pairs with a one-to-one correspondence. Further, the sample data pairs are matrix-organized to construct a mapping relationship structure between the input matrix and the output matrix, which is used to represent the overall correspondence between the pressure regulation opening and the high-frequency differential potential signal. Then, based on the mapping relationship structure, the relationship between the input matrix and the output matrix is ​​parameterized. By minimizing the output error of the input matrix after mapping transformation, the deviation between the output matrix and the actual high-frequency differential potential signal sequence is minimized, thereby obtaining the inverse control mapping matrix that represents the correspondence between the two.

[0089] Next, transient potential disturbance characteristics are constructed based on high-frequency differential potential signals, and the asymmetric deviation state between aerodynamic units is determined according to these characteristics. First, the high-frequency differential potential signal output from the differential potential loop is acquired via time series to obtain electrical signal data representing the transient thermal response changes of the aerodynamic units. Then, feature extraction processing is performed on the high-frequency differential potential signal to obtain transient potential disturbance characteristics, including potential amplitude characteristics, time change rate characteristics, and energy distribution characteristics, to represent the intensity of change and dynamic fluctuation characteristics of the electrical signal under transient conditions. Subsequently, a comparative analysis is performed based on the transient potential disturbance characteristics corresponding to the first and second aerodynamic units, calculating the characteristic differences between the two at each time sampling point and within the statistical window to form an asymmetric deviation performance quantity. The consistency of the electrical signal response between aerodynamic units is evaluated using the asymmetric deviation performance quantity, thereby determining the asymmetric deviation state between the aerodynamic units, which is used to represent the operational consistency and structural response balance of the aerodynamic system under transient conditions.

[0090] The variation in asymmetric deviation between pneumatic units is determined, and the sign of this variation is assigned to obtain the deviation direction identifier. For example, the state difference between the first and second pneumatic units is obtained based on their corresponding asymmetric deviation states. This state difference represents the degree of response difference between the two pneumatic units at the same time or within the same sampling window. When the state difference is greater than zero, the first pneumatic unit is determined to be in an enhanced state relative to the second pneumatic unit, and the deviation direction identifier is recorded as positive. When the state difference is less than zero, the first pneumatic unit is determined to be in a weakened state relative to the second pneumatic unit, and the deviation direction identifier is recorded as negative. When the state difference is equal to zero, the two pneumatic units are determined to be in a symmetrical equilibrium state, and the deviation direction identifier is recorded as zero. Subsequently, quantitative indicators of transient potential disturbance characteristics are extracted, such as the peak-to-valley difference in potential, signal distortion, inter-unit potential phase difference, and disturbance fluctuation variance. By comparing the quantitative indicators with the benchmark statistical parameters extracted from historical sampling data to represent the normal fluctuation range of the system, the difference between the two can be calculated to determine the actual deviation. The intensity coefficient is equal to the ratio of the actual deviation to the benchmark statistical parameter. The deviation intensity coefficient represents the degree of asymmetry between pneumatic units. Subsequently, the deviation direction indicator and the deviation intensity coefficient are introduced as extended control variables into the input variable set of the inverse control mapping matrix. Based on the extended input-output relationship, the matrix and parameters are updated and solved again to obtain the updated inverse control mapping matrix. In other words, based on the original input-output sample set constructed with pressure regulation opening and high-frequency differential potential signal, the deviation direction indicator and deviation intensity coefficient extracted from the asymmetric deviation state are introduced as new control input variables to expand the input feature dimension, thus forming an extended input-output sample set. The extended input-output sample set is then expressed in a matrix form, expanding the input variables from a single control quantity to a multi-dimensional input vector containing pressure regulation opening, deviation direction indicator, and deviation intensity coefficient, and establishing a mapping relationship with the corresponding high-frequency differential potential signal output vector. Subsequently, by solving for the correspondence between the extended input and output matrices, the mapping matrix is ​​updated while minimizing the output error.

[0091] Finally, based on the inverse control mapping matrix after adding control input variables, a corresponding feedforward compensation command is generated, and the pressure regulation opening of the pneumatic unit is adjusted using the feedforward compensation command. First, the pressure regulation opening, deviation direction indicator, and deviation intensity coefficient under real-time operating conditions are obtained, and the above parameters are constructed into an extended input vector. Further, the extended input vector is input into the updated inverse control mapping matrix for calculation to obtain the corresponding predicted output result, and the feedforward compensation amount is determined based on the deviation relationship between the predicted output result and the target control state. For example, at a certain moment, the pressure regulation opening of the first pneumatic unit is 60%, the deviation direction indicator is +1, and the deviation intensity coefficient is 0.35. Then, the above parameters are constructed into an extended input vector U(t), S(t), K(t) = 0.60, +1, 0.35, and input into the updated inverse control mapping matrix for calculation, resulting in a predicted output potential response value that is 0.12 units higher than the target reference value (for example only). At this point, the feedforward compensation amount is determined based on the deviation between the predicted output and the target control state. For example, if an 8% increase in pressure regulation opening compensation is calculated, this compensation amount is added to the current pressure regulation opening to generate a feedforward compensation control command of 68%, which is then output to the pressure regulation actuator of the pneumatic unit. This achieves feedforward correction control for asymmetric deviation states, thereby improving the consistency and stability of the system's transient response. Subsequently, the feedforward compensation amount is superimposed with the current pressure regulation opening to generate a feedforward compensation control command. Finally, the feedforward compensation control command is output to the pressure regulation actuator of the pneumatic unit to adjust its pressure regulation opening, thereby achieving feedforward compensation control for asymmetric deviation states of the pneumatic unit and improving the stability and control accuracy of the pneumatic system's transient response.

[0092] By generating feedforward compensation commands based on transient potential disturbance characteristics to adjust the pressure regulation opening in real time, rapid suppression and active balancing of thermo-mechanical coupling disturbances are achieved. This can significantly improve the dynamic consistency of pressure distribution and forming stability of multi-pneumatic units in industrial forming equipment, effectively avoid local pressure drift and sheet forming defects caused by transient thermal imbalance, and improve the quality consistency and process reliability of large cover part forming production.

[0093] In one embodiment, extracting the deviation direction identifier and deviation intensity coefficient based on the asymmetric deviation state includes: S810: Extract the polarity characteristics and voltage peak value of the high-frequency differential potential signal within the current sampling period; S820: Based on polarity characteristics and preset differential loop topology, the target dominant side with the highest heat generation in the symmetrical reference unit group is determined by matching, and the unit position attribute corresponding to the target dominant side is used as the deviation direction identifier. S830. The transient thermal energy equivalent is obtained by time-domain integration of the absolute value of the high-frequency differential potential signal, so as to determine the transient thermal energy equivalent of the asymmetric work done by the pneumatic unit during compression. S840: Input the transient thermal energy equivalent and voltage peak value into the preset aerodynamic thermoelectric sensitivity function to obtain the deviation intensity coefficient.

[0094] Within the current sampling period, the high-frequency differential potential signal output by the differential potential loop is continuously sampled. Polarity analysis is performed on the voltage sequence data, and its polarity characteristics are determined by calculating the positive and negative distribution of the voltage signal within this sampling period, thus representing the overall offset direction of the potential signal within this sampling period. Subsequently, the maximum instantaneous amplitude within this sampling period is obtained from the voltage sequence data to determine the voltage peak value, which represents the maximum transient fluctuation intensity of the signal within the current sampling period.

[0095] Next, the polarity characteristics of the high-frequency differential potential signal within the current sampling period are acquired to represent the positive and negative directions of the differential potential signal. Furthermore, based on a pre-defined differential potential loop topology, the polarity correspondence between the first and second thermoelectric conversion units in the circuit is clarified, thus establishing a mapping relationship between the potential signal polarity and the spatial position of the pneumatic units. Subsequently, matching analysis is performed based on the polarity characteristics and mapping relationship to determine the side of the pneumatic unit that generates a higher temperature rise within the current sampling period, and this side is identified as the target dominant side with the highest heat generation. Finally, the positional attributes of the pneumatic unit corresponding to the target dominant side are extracted and encoded as deviation direction identifiers to represent the asymmetric deviation direction between pneumatic units, thereby providing a directional basis for subsequent control variable construction and control strategy adjustment.

[0096] The absolute value of the high-frequency differential potential signal is integrated in the time domain to obtain the transient thermal energy equivalent, thereby determining the work done by the asymmetric compression of the pneumatic unit. Specifically, the high-frequency differential potential signal output from the differential potential loop is acquired within a preset effective time domain window, forming corresponding time series data. The absolute value of the high-frequency differential potential signal is processed to eliminate the influence of the potential signal's directionality on energy performance, obtaining a sequence of absolute potential values ​​reflecting transient amplitude changes. Subsequently, the absolute potential value sequence is integrated in the effective time domain window. Under discrete sampling conditions, the absolute potential value at each sampling moment is accumulated with the sampling time interval to obtain the corresponding transient thermal energy equivalent. Through this processing, the transient potential difference signal is converted into an equivalent thermal energy index that can represent the energy accumulation effect, thus representing the transient thermal energy equivalent corresponding to the work difference caused by asymmetry during compression of the pneumatic unit.

[0097] The transient thermal energy equivalent and voltage peak value are input into a preset aerodynamic thermoelectric sensitivity function to obtain the deviation intensity coefficient. First, the transient thermal energy equivalent calculated within the effective time domain window and the voltage peak value extracted within the current sampling period are obtained, respectively used to represent the cumulative effect of asymmetric work and the transient peak response characteristics during the compression process of the aerodynamic unit. Further, the transient thermal energy equivalent and voltage peak value are introduced as input variables into the preset aerodynamic thermoelectric sensitivity function for fusion calculation. The specific expression of the preset aerodynamic thermoelectric sensitivity function is shown below: Among them, K dev It is the deviation strength coefficient; Q eq It is the transient thermal energy equivalent obtained by time-domain integration of the absolute value of the high-frequency differential potential signal; V peak It is the peak voltage of the extracted high-frequency differential potential signal; V ref It is the preset system reference voltage, which is usually calibrated as the maximum tolerance voltage for normal allowable fluctuations; It is the basic thermal energy sensitivity weighting coefficient, used to convert thermal energy into a basic control compensation ratio; It is the transient peak penalty weighting coefficient; I is the nonlinear sensitivity exponent, which can take the value of 2 or a real number greater than 1, and is used to measure the transient peak value in V. reak More than V ref At that time, the deviation intensity coefficient is amplified rapidly and non-linearly; the specific value can be set according to the company's accuracy requirements. It can be set directly based on the allowable basic thermal deviation compensation ratio of the equipment under normal operating conditions. First, determine the compensation ratio per unit Q when the equipment has no sudden failures. eq The corresponding upper limit of the safety compensation opening change is calculated by dividing the target compensation ratio by Q under normal operating conditions. eq The average value is used to obtain λ1. To amplify sudden voltage peak deviations, calibration can be performed based on typical fault conditions. The most typical fault scenario for the equipment can be selected, and V under that scenario can be collected. peak The required compensation intensity is determined by back-calculating the data. Subsequently, the corresponding deviation intensity coefficient is obtained by evaluating the aerothermal-electric sensitivity function. Through this process, the multi-dimensional aerothermal response characteristics are integrated into a single quantitative index, thereby achieving a comprehensive representation of the asymmetry between aerodynamic units, which is then used for subsequent control variable construction and control strategy optimization.

[0098] By extracting the deviation direction identifier and deviation intensity coefficient, the refinement and dynamic response capability of the asymmetric deviation state can be significantly improved, providing high-fidelity and high-time-efficiency control input for the generation of feedforward compensation commands, and effectively improving the accuracy and stability of pressure equalization control of the pneumatic system of industrial forming equipment.

[0099] This application also provides a machine-readable storage medium storing instructions that cause a machine to execute the above-described industrial equipment status monitoring method based on real-time data acquisition.

[0100] This application also provides an electronic device, including: The memory is configured to store instructions; and The processor is configured to retrieve instructions from memory and, when executing instructions, to implement the aforementioned method for monitoring the condition of industrial equipment based on real-time data acquisition.

[0101] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0103] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0104] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0105] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0106] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0107] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0108] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0109] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for monitoring the condition of industrial equipment based on real-time data acquisition, characterized in that, include: When the slider in the industrial forming equipment comes into contact with the target sheet, the real-time response data of the industrial forming equipment is acquired. Determine the current production process requirements of the industrial forming equipment, and based on the production process requirements, select multiple pneumatic units within the industrial forming equipment to construct a symmetrical reference unit group; Obtain the element response data for each symmetric reference element group; Establish a gas-thermal mapping relationship based on the element response data of the symmetric reference element group; Thermocouple temperature measurement nodes are set in the aerodynamic unit corresponding to the symmetrical reference unit group based on the gas-thermal mapping relationship, and the corresponding transient temperature difference signal is obtained. A differential potential loop is constructed for the aerodynamic units within the same symmetrical reference unit group based on the transient temperature difference signal; When the sealing performance of the pneumatic units in the symmetrical reference unit group is consistent, the differential potential circuit is used to output the differential potential signal of the thermocouple temperature measurement node of each pneumatic unit to zero potential, so as to obtain the zero potential reference state. Abnormal pneumatic units are identified by combining zero potential reference state and real-time response data; A status monitoring report for industrial forming equipment is generated based on the abnormal pneumatic unit.

2. The method according to claim 1, characterized in that, The process of determining the current production process requirements of the industrial forming equipment and selecting multiple pneumatic units within the industrial forming equipment to construct a symmetrical reference unit group based on these requirements includes: The forming process type is determined based on the current production process requirements of the industrial forming equipment, and the forming stage information is determined through the forming process type. The forming stage information includes the blank holder stage, the drawing stage, and the final forming stage. By combining information from the forming stage, the process action mapping of the pneumatic units is performed to determine the dominant action type of each pneumatic unit; The aerodynamic units are grouped according to the dominant action type to obtain the equivalent unit set; During the operation of industrial forming equipment, the geometric parameters of the die surface within each equivalent unit set are acquired, and the material flow direction and flow resistance distribution characteristics of each equivalent unit set are determined based on the die surface geometric parameters. Initial symmetrical candidate unit groups are screened by combining material flow direction and flow resistance distribution characteristics, and symmetrical reference unit groups are determined by time-series screening of the initial symmetrical candidate unit groups based on forming stage information.

3. The method according to claim 2, characterized in that, The process of selecting an initial symmetrical candidate unit group by combining material flow direction and flow resistance distribution characteristics, and then performing time-series screening of the initial symmetrical candidate unit group to determine a symmetrical reference unit group by combining forming stage information, includes: Collect the strain data of the target sheet material, and construct a material flow direction vector field based on the sheet material strain data and the material flow direction; A material flow resistance distribution field is constructed by combining the characteristics of flow resistance distribution, and the resistance symmetry axis is identified by using the material flow resistance distribution field; The aerodynamic unit is spatially symmetrically mapped by combining the drag symmetry axis, and the target unit group is selected according to the material flow direction vector field corresponding to the aerodynamic unit. When the forming stage is the edge pressing stage, the target unit group is subjected to time-series screening to obtain the first screening result; When the forming stage is the drawing stage, acquire the transient load data and displacement response data of the aerodynamic units within the target unit group; Transient load-displacement response curves are constructed by combining transient load data and displacement response data; The time warping distance and peak response phase difference are calculated based on the transient load-displacement response curve, and the target unit group is dynamically screened for consistency using the time warping distance and peak response phase difference to obtain the second screening result; When the forming stage is the final forming stage, determine the load response curve of the target unit group; The steady-state load mean and load convergence curve of the target unit group are extracted based on the load response curve. The consistency of the target unit group is verified based on the deviation of the steady-state load mean and the overlap of the load convergence curve to obtain the third screening result. The symmetric reference unit group was determined by combining the results of the first screening, the second screening, and the third screening.

4. The method according to claim 1, characterized in that, The step of establishing a gas-thermal mapping relationship based on the unit response data of the symmetric reference unit group includes: The piston compression speed, effective cross-sectional area of ​​the cylinder, and cylinder pressure response sequence of the pneumatic unit are extracted from the unit response data. Using the moment of contact between the slider and the target plate as the time reference point, the peak range of compression speed is identified based on the piston compression speed of the pneumatic unit; The effective time domain window is determined by combining the time reference point and the peak compression rate range; The piston compression velocity within the effective time window is integrated over time to determine the change in piston compression displacement, and the effective compression volume is calculated by combining it with the effective cross-sectional area of ​​the cylinder. By combining the effective compression volume and cylinder pressure response sequence, the gas adiabatic temperature rise sequence of the pneumatic unit within the effective time domain window is determined using a preset gas adiabatic state equation. The cylinder pressure response sequence and effective compression volume sequence are used as aerodynamic state characteristics, and the gas adiabatic temperature rise sequence is used as thermodynamic state characteristics. A gas-thermal state response model is constructed by combining the time-series dynamic correspondence between aerodynamic and thermodynamic state characteristics. The gas-thermal mapping relationship is determined by the gas-thermal state response model.

5. The method according to claim 4, characterized in that, The method further includes: Extract the first aerodynamic state features and the second aerodynamic state features at the same time reference point from the first aerodynamic unit and the second aerodynamic unit in the symmetrical reference unit group, respectively. Input the first aerodynamic state features and the second aerodynamic state features into the gas-thermal state response model respectively to output the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence. Within the effective time window, calculate the dynamic temperature rise deviation sequence between the first predicted adiabatic temperature rise sequence and the second predicted adiabatic temperature rise sequence, and calculate the aerodynamic asymmetric pressure difference sequence between the first aerodynamic state characteristic and the second aerodynamic state characteristic. A symmetric error compensation function is constructed based on the aerodynamic asymmetric pressure difference sequence and the dynamic temperature rise deviation sequence; The symmetric error compensation function is used as a regularization constraint to iteratively update the model parameters of the gas-thermal state response model, resulting in the corrected gas-thermal state response model.

6. The method according to claim 1, characterized in that, The construction of a differential potential loop for aerodynamic units within the same symmetrical reference unit group based on transient temperature difference signals includes: The transient temperature rise response region corresponding to the outer wall of the pneumatic unit cylinder is identified by the transient temperature difference signal; In combination with the transient temperature rise response region, the first and second aerodynamic units in the symmetrical reference unit group are set with the same Seebeck coefficient; The transient temperature difference signal is converted into a differential potential signal using a thermoelectric conversion unit; The same polarity terminals of the first thermoelectric conversion unit and the second thermoelectric conversion unit are differentially connected, and the opposite polarity terminals are connected to the differential detection terminal to construct a differential potential loop for suppressing common-mode thermal interference.

7. The method according to claim 6, characterized in that, The method further includes: Obtain the pressure regulation opening of each pneumatic unit; The potential signal output by the differential potential loop is acquired in time series, and the low-frequency thermal diffusion component of the time series potential signal is filtered out by frequency domain filtering to obtain a high-frequency differential potential signal. A reverse control mapping matrix is ​​constructed by combining the high-frequency differential potential signal and pressure regulation opening within the historical sampling period; Transient potential disturbance characteristics are constructed based on high-frequency differential potential signals, and the asymmetric deviation state between aerodynamic units is determined based on the transient potential disturbance characteristics. Based on the asymmetric deviation state, the deviation direction identifier and deviation intensity coefficient are extracted, and the deviation direction identifier and deviation intensity coefficient are added to the inverse control mapping matrix as control input variables. Based on the inverse control mapping matrix after adding control input variables, the corresponding feedforward compensation command is generated, and the pressure regulation opening of the pneumatic unit is adjusted using the feedforward compensation command.

8. The method according to claim 7, characterized in that, The step of extracting the deviation direction identifier and deviation intensity coefficient based on the asymmetric deviation state includes: Extract the polarity characteristics and voltage peak value of the high-frequency differential potential signal within the current sampling period; Based on the polarity characteristics and the preset differential loop topology, the target dominant side with the highest heat generation in the symmetrical reference unit group is determined by matching, and the unit position attribute corresponding to the target dominant side is used as the deviation direction identifier. The transient thermal energy equivalent is obtained by time-domain integration of the absolute value of the high-frequency differential potential signal, so as to determine the transient thermal energy equivalent of the work done by the asymmetric compression of the aerodynamic unit. The deviation intensity coefficient is obtained by inputting the transient thermal energy equivalent and voltage peak value into the preset aerodynamic thermoelectric sensitivity function.

9. A machine-readable storage medium, characterized in that, The machine-readable storage medium stores instructions for causing the machine to execute an industrial equipment condition monitoring method based on real-time data acquisition according to any one of claims 1 to 8.

10. An electronic device, characterized in that, include: The memory is configured to store instructions; as well as The processor is configured to retrieve the instructions from the memory and, when executing the instructions, to implement an industrial equipment status monitoring method based on real-time data acquisition according to any one of claims 1 to 8.

Images