Intelligent process management system based on technological parameter modeling

Abstract

The invention relates to the technical field of process modeling, in particular to an intelligent process management system based on process parameter modeling, which comprises a parameter acquisition module, a nested analysis module, a path construction module, a coupling deconstruction module and a structure compression module. According to the method, the hierarchy and expression accuracy of parameter organization are improved by collecting structured interface information and constructing a parameter mapping relation, dominant and attached parameters are set to form a control logic hierarchy, the clarity and controllability of process parameter response are enhanced, and the stability and anomaly adaptability of path combination are improved; the node frequency and sequence difference are compared to form an interference disassembly fragment, structural anomaly perception and local reconstruction efficiency are enhanced, a path tree structure is constructed through compression and recombination of a repeated path segment, the reusability and organization compactness of path information are improved, parameter logic nesting, path change tracking and interference analysis are achieved under multi-process cooperation, and the reliability of the system is improved. And the logic coupling identification capability, the structural adaptability and the expression precision of process modeling are enhanced.

Description

Technical Field

[0001] This invention relates to the field of process modeling technology, and in particular to an intelligent process management system based on process parameter modeling. Background Technology

[0002] The field of process modeling technology involves the technical activities of abstracting, analyzing, and modeling process flows, process parameters, and their interrelationships during manufacturing. Core aspects include process data acquisition, process feature identification, process behavior modeling, and rule extraction. The overall technical approach typically employs statistical analysis, rule induction, and model building to digitally represent and optimize the manufacturing process. Traditional intelligent process management systems based on process parameter modeling refer to information systems that collect historical parameter data from multiple process nodes during manufacturing and establish logical relationships between processes to achieve process scheduling and management. The technical challenge addressed by these systems is how to effectively organize various discrete process parameters and use them to generate reusable process flow descriptions. Traditional methods typically establish parameter association logic through process rule tables and use rule matching to achieve process flow judgment and management.

[0003] Existing technologies suffer from coarse-grained information organization and limited ability to construct logical connections during process modeling. Relying on historical parameter data to build process rule tables often only reflects empirical rules under static conditions, lacking the ability to map dynamic process responses in real time. Process parameters mostly exist in planar relationships, lacking a logical distinction between dominant and dependent parameters, resulting in ambiguous parameter control levels and difficulty in supporting progressive modeling of complex process logic. Identifying state abrupt changes or node branches in the process flow relies on rule-based judgment, which is difficult to adapt to high-frequency switching or multi-path combination scenarios, often leading to misjudgments or disordered path generation. For example, in multi-segment stamping paths, the lack of effective identification methods between frequent switching rhythms and structurally dense nodes easily leads to undetected interference and decreased process stability. Path reuse capability relies on manual rule summarization, lacking a systematic compression and integration strategy, resulting in redundant duplicate path information, affecting storage efficiency and response speed. These shortcomings limit the adaptability of existing systems in multi-process integration and highly dynamic manufacturing scenarios, making it difficult to support the digital reconstruction needs of complex process flows. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent process management system based on process parameter modeling.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: An intelligent process management system based on process parameter modeling includes: The parameter acquisition module acquires the structural information of the process interface in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, establishes the mapping structure between parameters, and generates process input parameter structure blocks. The nested parsing module sets the heat treatment temperature setpoint as the dominant parameter based on the process input parameter structure block, and sets the protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code as dependent parameters. It arranges the hierarchy according to the control order, constructs a nested relationship graph, and generates a nested logic combination model. Based on the nested logic combination model, the path construction module calls the mold cavity temperature, cooling channel opening and closing signals and gate status identifiers in the injection molding process, identifies discontinuous nodes and marks them as path branch points, sorts the nodes according to the state relationship, and generates multi-level path deduction structure segments. The coupling deconstruction module calls the multi-level path deduction structure segment, compares the node frequency of stamping frequency with the die closure amount, identifies interference points based on position differences, extracts and reconstructs path segments, and generates path interference deconstruction segment areas.

[0006] As a further embodiment of the present invention, the process input parameter structure block includes heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type, and start / stop time period code; the nested logic combination model includes dominant parameter, dependent parameter, parameter hierarchy, and nested relationship graph; the multi-level path deduction structure segment includes path stage, state discontinuity node, path branch point, path node sequence, and path segment; and the path interference decomposition segment area includes node occurrence frequency, structural interference point, and area to be reconstructed.

[0007] As a further aspect of the present invention, the parameter acquisition module includes: The process interface acquisition submodule acquires the process interface structure information in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, classifies and groups the parameters according to the processing batch, and generates a classified and grouped data structure. Based on the classification and grouping data structure, the parameter structure extraction submodule classifies and pairs the heat treatment temperature setpoint, protective atmosphere flow rate and start-stop time period code in each batch, filters the differences in proportional integral derivative control type and workpiece material code within the group, combines the correspondence between control parameters and materials, and establishes a parameter mapping structure. The input structure generation submodule calls the parameter mapping structure to integrate the heat treatment temperature setting, protective atmosphere flow rate, workpiece material code and start / stop time period code corresponding to each batch, and combines them to construct a unified input parameter structure form to generate the process input parameter structure block.

[0008] As a further aspect of the present invention, the nested parsing module includes: The parameter identification submodule extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code based on the process input parameter structure block, arranges the parameters in the order of their appearance, identifies the hierarchical relationship between control parameters and dependent parameters, and obtains the parameter hierarchical identification value. The control response parsing submodule calls the parameter level identification value, takes the heat treatment temperature setpoint as the dominant parameter, analyzes the difference between the protective atmosphere flow response time period and the temperature change time node, calculates the response ratio and determines whether it meets the control sequence requirements, and obtains the response sequence judgment coefficient. The combined modeling submodule calls the response order judgment coefficient, filters data segments that meet the control order requirements, establishes a parameter nesting mapping matrix, combines proportional-integral-derivative control types and start-stop time period codes, constructs a multi-dimensional combined structure, obtains the nested topological responsivity of parameter combination relationships, and establishes a nested logic combination model.

[0009] As a further aspect of the present invention, the path construction module includes: The path recognition submodule, based on the nested logic combination model, calls three main parameters: mold cavity temperature, cooling channel opening and closing signal, and gate status identifier, to extract the node status in the path stage. By judging whether there are jump differences between states, comparing the change rate and state amplitude combination range, it filters out nodes with discontinuous states and generates a sequence of discontinuous state nodes. The path splicing submodule calls the state discontinuous node sequence, logically identifies the state change nodes according to the time sequence, and compares the combination order between nodes according to the temporal relationship of the gate state identifier, splicing them to form a hierarchical path segment structure, and obtaining the path segment splicing sequence. The deduction generation submodule compares and classifies the structural mapping combinations between path segments based on the path segment splicing sequence and the mold cavity temperature and cooling channel opening and closing signals. It identifies segment nodes with significant differences in path branches, filters structural combinations with strong path extensibility, and obtains multi-level path deduction structural segments.

[0010] As a further aspect of the present invention, the coupling deconstruction module includes: The frequency comparison submodule obtains the die closure amount data and node occurrence frequency in the stamping process based on the multi-level path deduction structure segment. Based on the frequency distribution corresponding to the node, it calls the stamping frequency value and the node occurrence frequency for difference comparison, filters the node set whose frequency deviation exceeds the stamping reference frequency difference value, establishes the occurrence frequency offset record of the corresponding node, and generates the occurrence frequency offset sequence. The path difference extraction submodule extracts the index position of the involved node in the path structure according to the occurrence frequency offset sequence, compares the index difference of corresponding nodes between paths, determines whether it exceeds the tolerance range of the interference path, extracts the path node combination with the difference exceeding the limit, obtains the path segment in the structure, and obtains the path interference segment index interval. The interference zone generation submodule, based on the path interference segment index interval, calls the structural node order and spacing of the corresponding path segment, identifies the concentration degree and minimum interval of interference nodes in the segment, filters path segments whose concentration exceeds the interference baseline interval, obtains the node range, and generates the path interference decomposition segment area.

[0011] As a further aspect of the present invention, the system further includes: The structure compression module decomposes the segment area according to the path interference, extracts the welding current, welding gun position code and welding speed in the welding process, determines whether the main control parameters in the continuous path segment are completely overlapping, marks the repeated path structure segment, integrates it into the path tree structure and updates the index order, and generates a set of reconfigurable path segment structures. The reconfigurable path segment structure set includes repeating path segment structures, path tree structures, and index order.

[0012] As a further aspect of the present invention, the structure compression module includes: The path decomposition submodule decomposes the path interference segment area, collects the start and end position codes of the path segments and the distribution information of the number of segments, indexes and marks the segments according to the segment intervals, extracts the welding process dataset in the path segments, including the welding current sequence, the spatial position coordinates of the welding torch and the welding speed trajectory, forms the path segmentation parameter table, and obtains the segment path parameter set. The main control parameter identification submodule calls the segment path parameter set, constructs the segment main control parameter group based on welding current, welding gun position code and welding speed, obtains the parameter arrangement order and position interval boundary in the continuous path segment, compares the parameter values ​​within the interval and matches the sequence of the segment path parameter group, obtains the path segment parameter overlap value, determines whether the overlap reaches the overlap threshold, filters the segment number that meets the overlap condition, obtains the path segment parameter overlap discrimination result, and generates a repeating structure segment index set. The structural segment integration submodule, based on the repeated structural segment index set, rearranges the indexes of the repeated structural segments according to the original path order, adjusts the paragraph hierarchy and index mapping order of the path tree structure, reconstructs the path tree structure, integrates all paragraph numbers and main control parameter order information, and generates a reconstructable path paragraph structure set.

[0013] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by collecting structured interface information and constructing parameter mapping relationships, the hierarchical nature and accuracy of parameter organization are improved. Dominant and dependent parameters are set to form a control logic hierarchy, enhancing the clarity and controllability of process parameter responses, improving the stability and anomaly adaptability of path combinations, and forming interference decomposition segments by comparing node frequency and sequence differences. This strengthens the perception of structural anomalies and the efficiency of local reconstruction. The compression and reorganization of repeated path segments constructs a path tree structure, improving the reusability and compactness of path information. Under multi-process collaboration, parameter logic nesting, path change tracking, and interference analysis are achieved, strengthening the logical coupling recognition capability, structural adaptability, and expression accuracy of process modeling. Attached Figure Description

[0014] Figure 1 This is a system flowchart of the present invention; Figure 2 This is a flowchart of the parameter acquisition module of the present invention; Figure 3 This is a flowchart of the nested parsing module of the present invention; Figure 4 This is a flowchart of the path construction module of the present invention; Figure 5 This is a flowchart of the coupling deconstruction module of the present invention; Figure 6 This is a flowchart of the structural compression module of the present invention. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0016] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0017] Please see Figure 1 Intelligent process management systems based on process parameter modeling include: The parameter acquisition module acquires the structural information of the process interface in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, groups them by processing batch, establishes the mapping structure between parameters, and generates process input parameter structure blocks. The nested parsing module analyzes the control response sequence between the heat treatment temperature setpoint and the protective atmosphere flow rate based on the process input parameter structure block. It sets the heat treatment temperature setpoint as the dominant parameter and uses the protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code as dependent parameters. It arranges the parameter hierarchy according to the control sequence, constructs a nested relationship graph, and generates a nested logic combination model. The path construction module, based on the nested logic combination model, calls three main parameters in the injection molding process: mold cavity temperature, cooling channel opening and closing signal, and gate status identifier. It identifies nodes with discontinuous states in the path stages, and logically identifies the corresponding nodes as path branch points. By combining and mapping the path node order through state discrimination, it splices together multi-level path segments and generates multi-level path deduction structure fragments. The coupling deconstruction module calls multi-level path deduction structural segments, compares the frequency of node occurrence between the stamping frequency and the die closure amount in the stamping process, identifies the nodes with concentrated interference, verifies the structural interference points by the difference in the order of node positions in the path, extracts the area that needs to be reconstructed in the path, and generates the path interference deconstruction segment area. The structure compression module decomposes the segment area based on path interference, extracts welding current, welding gun position code and welding speed from the welding process, determines whether the main control parameters in continuous path segments are completely overlapping, marks repeated path structure segments, integrates them into the path tree structure and updates the index order, and generates a set of reconfigurable path segment structures.

[0018] The process input parameter structure block includes heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type, and start / stop time period code. The nested logic combination model includes dominant parameters, dependent parameters, parameter hierarchy, and nested relationship graph. The multi-level path deduction structure segment includes path stage, state discontinuity node, path branch point, path node sequence, and path segment. The path interference decomposition segment area includes node occurrence frequency, structural interference point, and area to be reconstructed. The reconstructable path segment structure set includes repeating path structure segment, path tree structure, and index order.

[0019] Please see Figure 2 The parameter acquisition module includes: The process interface acquisition submodule acquires the process interface structure information in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, classifies and groups the parameters according to the processing batch, and generates a classified and grouped data structure. To obtain the process interface structure information in the heat treatment process, it is first necessary to directly read the data information of each relevant interface from the industrial control system or data acquisition module. This includes, but is not limited to, the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type, and start / stop time period code. This information is generally distributed in different signal acquisition channels. For example, the temperature setpoint is obtained from the temperature control system interface, such as batch A001 set to 850℃, A002 to 880℃, and A003 to 890℃. The protective atmosphere flow rate is obtained from the gas flow controller, such as A001 to 20L / min, A002 to 25L / min, and A003 to 22L / min. The quality code is automatically matched by the production planning system or material label, such as C45, 20Cr, and 42CrMo respectively. The proportional-integral-derivative (PI) control type is recorded by the main control command input system, for example, A001 is PI control, A002 is PID control, and A003 is P control. The start / stop time period code is matched by the mapping relationship between the time period and the number table, such as T1 representing 08:00-10:00, T2 representing 10:00-12:00, and T3 representing 12:00-14:00. After the data is collected, these parameters need to be grouped and organized by batch, that is, the parameter values ​​from different sources are matched and integrated by batch fields to form structured grouped data, as shown below: As shown in Table 1, multiple parameter items within each batch have been integrated to form structured data blocks, providing a foundation for the next stage of extracting the structural combinations of each parameter, ultimately resulting in a classified and grouped data structure.

[0020] The parameter structure extraction submodule is based on a classification and grouping data structure. It classifies and pairs the heat treatment temperature setpoint, protective atmosphere flow rate and start-stop time period code in each batch. It filters the differences in proportional-integral-derivative control type and workpiece material code within the group, combines the correspondence between control parameters and materials, and establishes a parameter mapping structure. Based on the constructed classification and grouping data structure, the parameter structure extraction stage begins. First, the temperature setpoint, atmosphere flow rate, and start / stop time interval codes within each processing batch are combined and paired. For example, in batch A002, this is <880, 25, T2>. Next, the PID type field is used to determine if the group has proportional-integral-derivative control characteristics. If the field is PID type, and the corresponding material code field is not empty and is in the valid material code list (e.g., C45, 20Cr, 42CrMo), then this group of data is valid and continues to participate in the mapping establishment process. For example, in group A002, the PID... The data set containing the 20Cr material code is valid and included in the valid list, so it is retained. However, batch A003, despite having complete other parameters, has a PID type of P, which does not meet the control characteristic requirements, and therefore is not included in the mapping construction. In the above filtering process, the judgment logic is to simultaneously satisfy the control type field being PID and the material field being in the valid set; if either condition is not met, the data is discarded. The filtered data items are used to construct the parameter mapping structure. In specific operations, temperature value, atmosphere flow rate, and start / stop time period are used as key fields, paired with the control type and material code. For example, A002 is <880, 25, T2> →<PID,20Cr> A mapping relationship is established, and all the selected valid combinations are finally constructed into a key-value mapping structure to facilitate subsequent data retrieval and structure reconstruction, ultimately resulting in a parameter mapping structure.

[0021] The input structure generation submodule calls the parameter mapping structure to integrate the heat treatment temperature setting, protective atmosphere flow rate, workpiece material code and start / stop time period code corresponding to each batch, and combines them to construct a unified input parameter structure form to generate the process input parameter structure block. The aforementioned parameter mapping structure is invoked to extract the temperature setpoint, atmosphere flow rate, material code, and start / stop time period code from the data set corresponding to each batch of workpieces. A structured input block is then constructed according to the field order. This input block must ensure that the fields are complete and the format is consistent. The structure reconstruction process first calls the mapping relationship stored in the parameter mapping structure to match the fields. For example, if the input for batch A002 is <880, 25, 20Cr, T2>, if this combination already exists in the parameter mapping structure, its structure is confirmed to be valid. Then, a unified data matrix is ​​generated based on the structure template. This matrix is ​​concatenated with the batch as the index and the four parameter fields as column values. In the case of missing fields, such as a batch lacking a time period code due to abnormal data acquisition, a default code value T0 is set, or it is filled in by inferring from the time periods of the preceding and following batches. The integrated structured input is uniformly encapsulated into an array or matrix form for use by subsequent modules, and finally, the process input parameter structure block is generated.

[0022] Please see Figure 3 The nested parsing module includes: The parameter identification submodule extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code based on the process input parameter structure block, arranges the parameters in the order of their appearance, identifies the hierarchical relationship between control parameters and dependent parameters, and obtains the parameter hierarchy identification value. Based on the process input parameter structure block, the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type, and start / stop time period code are extracted. Standard heat treatment process numbers HT101 and HT102 are set to correspond to heat treatment operations of different batches. The heat treatment temperature setpoint is read from the control system database and recorded according to the standard: HT101 is set to 920℃, and HT102 is set to 950℃, with corresponding protective atmosphere flow rates recorded as 15.2L / min and 18.5L / min. The workpiece material code is extracted, the former being 45# steel and the latter being GCr15 bearing steel. PI or PID control types are set according to different control strategies, and time period codes T1-A and T2-B are assigned to them in the timing logic to form the original parameter structure block. Then, the parameters are judged hierarchically. Based on their direction of action in the control logic, the heat treatment temperature setpoint is identified as the dominant parameter, and the other four items are identified as dependent parameters based on their subordinate response process, completing the parameter master-slave division. The results are shown in Table 2, obtaining the parameter hierarchy identification value. As shown in Table 2, by classifying and extracting various process parameters, standard parameter combination blocks are formed and hierarchical structures are determined, laying a structural foundation for subsequent control response sequence determination.

[0023] The control response analysis submodule calls the parameter hierarchy identification value, takes the heat treatment temperature setpoint as the dominant parameter, analyzes the difference between the protective atmosphere flow response time period and the temperature change time node, calculates the response ratio and determines whether it meets the control sequence requirements, and obtains the response sequence judgment coefficient. The parameter hierarchy identification value was called, with the heat treatment temperature setpoint as the dominant parameter. The relationship between the protective atmosphere flow response time period and the temperature change time node was analyzed. The heat treatment setpoint change time was set to 300 seconds and 350 seconds. The protective atmosphere flow response start time was collected in the response control system at 310 seconds and 362 seconds, and ended at 360 seconds and 400 seconds. Through difference calculation, it can be seen that the first response has a 10-second delay and lasts for 50 seconds, and the second response has a 12-second delay and lasts for 38 seconds. According to the atmosphere change control value record, the two adjustments are 3.2L / min and 4.5L / min respectively. The atmosphere adjustment value is divided by the set temperature adjustment section length and the average flow response rate is calculated. By judging whether it is within the standard judgment threshold of 0.15-0.35, the final two ratios are 0.25 and 0.28 respectively, both of which meet the effective response standard. The response order judgment coefficient is obtained. See Table 3 for details. As shown in Table 3, the response lag and adjustment range meet the judgment rules. The offset values ​​during the response process do not exceed the baseline range of 0.5 seconds or the flow rate adjustment ratio limit range, indicating that the atmosphere adjustment is dominated by temperature control and forms an effective feedback chain.

[0024] The combined modeling submodule calls the response sequence judgment coefficients, filters data segments that meet the control sequence requirements, establishes a parameter nesting mapping matrix, combines proportional-integral-derivative control types with start-stop time period codes, and constructs a multi-dimensional combined structure using the following formula: ; The nested topological responsivity of the parameter combination relationship is obtained by calculation, and a nested logical combination model is established; in, Indicates the duration of the heat treatment set temperature; This represents the weighting coefficient used to sum the time weights of each control priority level; Indicates the priority of the j-th control type; Indicates the corresponding integration time constant; Normalized time parameter; used to keep the formula results dimensionless. Indicates the estimated time for atmosphere adjustment response; This indicates the delay value caused by the influence of the metamaterial. This represents the threshold for the predicted response delay caused by temperature changes. Indicates nested topology responsiveness, with a control type index; The workpiece thickness, as recorded in the heat treatment process log, is 25mm. Industry standard empirical rules dictate that the holding time after heating should be 30 minutes per inch (approximately 25mm). The real-time temperature curve monitored by the equipment and the results of comparison with multiple thermocouples confirm that the holding time is 30 minutes (1800 seconds). This value is assigned... ; Normalized parameters The average integral time constant of the system, obtained from response time monitoring data, is 60 seconds. This value was extracted and analyzed from historical process data. ; The number of control priority settings is set to Experts analyzed the process flow and established a control type priority system, resulting in two types of control strategies. Priority values ​​were quantified using an exponential weighted average method based on historical big data characteristics. and These values ​​are derived from the output of the control strategy impact assessment model and are not assumptions. Corresponding integration time constant and The results were obtained by fitting the system response curves, respectively. , This step is calculated using the curve fitting module; Control priority time weighting coefficient The value was 1.2, obtained through training with an engineering regression model. This coefficient was calculated using data-driven methods, and surveys show that the value range is usually between 1.0 and 1.5. Estimated time for atmosphere adjustment The phased estimated time is obtained from the real-time response monitoring of the control system. The monitoring value is 300s, and this value is assigned to it. Material affects delay Temperature changes cause predictive response delay Both were obtained from system statistics and monitoring, and were 20s and 15s respectively, output through the delay analysis module; The calculation process is as follows: Substitute the above values ​​into the formula: Calculate the normalized weighted sum of time constants: ; Multiply by the weighting coefficient and take the square root: ; ; Molecular calculations: ; Denominator calculation: ; Substitute the values ​​to calculate the R value: ; The results show that the nested topology responsivity is 8.5076, which is a numerical index value. Subsequently, it was compared with the control sequence threshold system. If the standard control sequence threshold is set between 5 and 10, then this value meets the effective response range, indicating that the response structure is stable and effective.

[0025] Explanation of the meaning of the results: The calculated 2595.96s (numerator) is the quantified response index obtained by multiplying the weighted normalized differential time constant by the holding time, representing the "cumulative influence" of the holding process on the system response; 305s (denominator) is the sum of the direct response times of atmosphere adjustment, reflecting the impact of the time delay adjustment magnitude on the response. Finally, R=8.5076 represents the level of nested topology responsibility, indicating the system's response efficiency and the availability of sequence judgment under the current process conditions. A threshold judgment mechanism is used to determine whether the response is in the optimal control state, which can then be used for subsequent sequence judgment coefficient mapping. The formula's operational logic reflects the comprehensive influence of the timeliness, priority weights, and control coupling strength of the heat treatment process response. The numerator uses the set temperature holding time as the basic time scale for response strength, multiplied by the weighted square root of the priority time constant combination term, to reflect the coupling enhancement effect of control type response capability on the overall heat treatment process. The weighting term uses a linear superposition of the priority coefficient and the normalized integral time constant to reflect the relative influence of each control type on response timeliness. This is then corrected by multiplying by a weighting coefficient, followed by square root processing to normalize multiple time-domain influencing factors into a nonlinear growth curve of unified response strength. The denominator is the basic response time required for system atmosphere regulation plus the delay difference caused by process material properties and temperature disturbances, reflecting the weakening effect of operating condition complexity on system responsiveness. The overall result is a standardized topological response ratio formed by dividing the response enhancement term by the response consumption term, outputting a dimensionless nested responsibility index for the control model to identify whether the optimal execution order condition is met. The nested topology response degree of parameter combinations is a metric that measures the degree of mutual influence, temporal coupling strength, and execution sequence coordination of multiple control parameters in a process system. It quantifies the matching of response capabilities among different control types in heat treatment processes. This metric comprehensively considers temperature holding time, control priority, integral time constant, system atmosphere conditioning time, and delays caused by material and environmental changes, reflecting whether the control strategy forms effective linkage and sequential response under specific process conditions. Higher values ​​indicate a more compact response path topology and stronger response coordination among parameters, resulting in more proactive and efficient system control. Lower values ​​indicate problems such as response delays, misalignment, or sequence mismatch, suggesting the need to optimize control logic and parameter configuration. This metric provides a basis for judging the dynamic monitoring of the process flow and adjusting control strategies.

[0026] Please see Figure 4 The path building module includes: The path recognition submodule uses a nested logic combination model to call three main parameters: mold cavity temperature, cooling channel opening and closing signal, and gate status identifier to extract the node status in the path stage. By judging whether there are jump differences between states, comparing the combination range of change rate and state amplitude, it filters out nodes with discontinuous states and generates a sequence of discontinuous state nodes. Based on three main parameters—mold cavity temperature, cooling channel opening / closing signal, and gate status indicator—the mold cavity status is monitored during injection molding to identify potential abrupt state change points. Data is collected at 0.5-second sampling intervals, and the obtained data includes the mold cavity temperature change rate, cooling channel opening / closing response time, and gate status change indicator values. Taking the monitoring data listed in Table 1 as an example, the mold cavity temperature change rates are 2.5°C / s, 3.0°C / s, 1.8°C / s, and 4.2°C / s, respectively; the cooling channel opening / closing response times are 0.8s, 1.2s, 0.9s, and 1.1s, respectively; and the gate status change indicator values ​​are 1, 0, 1, and 1, respectively. The node position numbers are 12, 27, 33, and 45, respectively. Based on the above data, firstly, it is determined whether the rate of change of the mold cavity temperature exceeds the specified abrupt change judgment interval. Here, 3.5°C / s is used as the upper limit threshold for abrupt change, and node number 45 exceeds this threshold. Secondly, if the cooling channel switching response time is less than 1.0s, it is considered a response-sensitive node, and nodes numbered 12 and 33 meet this condition. At the same time, if the gate state jumps from 0 to 1, it indicates a state transition, and nodes numbered 12 and 33 also meet this condition. Combining the three judgment criteria, nodes numbered 12 and 33 simultaneously meet the conditions of a cooling response time of less than 1.0s, a gate state jump, and a mold cavity temperature change exceeding the basic fluctuation range. Therefore, they are determined to be state abrupt change nodes, and their corresponding numbers are recorded in the abrupt change node sequence. Such nodes can be used for subsequent path structure construction. As shown in Table 4, the indicators at nodes numbered 12 and 33 meet the threshold conditions for mutation determination. Therefore, they are selected and uniformly recorded as a sequence of discontinuous nodes.

[0027] The path splicing submodule calls the sequence of discontinuous state nodes, logically identifies nodes with abrupt state changes according to time order, and compares the combination order between nodes based on the temporal relationship of the gate state identifiers, splicing them to form a hierarchical path segment structure, thus obtaining the path segment splicing sequence. After calling nodes 12 and 33 from the discontinuous node sequence, they are first sorted on the overall timeline according to their node numbers, and the interval between them is calculated. Here, the interval between nodes 12 and 33 is 10.5 seconds, which meets the allowable interval range (less than 20 seconds) set in the segment splicing condition. Therefore, it is considered a splicable segment. Then, the logical path structure is analyzed according to the changes in the gate state between the segments. The state change flag values ​​of nodes 12 and 33 are both 1, indicating that their state logically represents the "continuously open" type, which is defined as flag type A. Then, node 27 is analyzed. The gate state change value of this node is 0, and the nodes before and after it are in state 1, forming a 0-1-0 jump pattern. Therefore, the logical types of path segments 12 to 27 and then to 33 are ABA splicing segment groups in sequence. According to the splicing logic, they can be combined into a first-level segment group, which is set as segment group P1. Meanwhile, considering node number 45, although its mold cavity temperature change rate exceeds the mutation judgment threshold, its cooling response time is 1.1s, which exceeds the response critical threshold, and its gate state change value is 1, which does not meet the three conditions. Therefore, this node is not included in the splicing segment group. Finally, the segment type and number corresponding to segment group P1 are recorded and used for the next stage of path deduction processing to obtain the path segment splicing sequence.

[0028] The deduction generation submodule compares and classifies the structural mapping combinations between path segments based on the path segment splicing sequence and the mold cavity temperature and cooling channel opening and closing signals. It identifies segment nodes with significant differences in path branches, filters structural combinations with strong path extensibility, and obtains multi-level path deduction structural segments. Based on the constructed path segment splicing sequence, the mapping relationship and structural deduction were further constructed by combining its three parameters. First, it was confirmed that the cavity temperature change rates corresponding to nodes 12 and 33 in segment group P1 were 2.5°C / s and 1.8°C / s, respectively, with an average value of 2.15°C / s, which is far below the structural abrupt change judgment threshold of 4.0°C / s. Therefore, it was determined that this segment group does not have the characteristics of high temperature response abrupt change. Subsequently, the response time of the cooling channel was analyzed. The response times of nodes 12 and 33 were 0.8s and 0.9s, respectively, with an average value of 0.85s, which is less than the set cooling response judgment threshold of 1.0s, thus meeting the cooling sensitivity index. Finally, the total number of nodes in this segment group is 2, which is less than the threshold of 5 for mapping nodes, thus failing to meet the complex mapping conditions. However, if node number 27 is introduced, although its cooling response time is 1.2s, which exceeds the limit, its mold cavity temperature change rate is 3.0°C / s. In some schemes, the structural standard can be adjusted, thereby expanding segment group P1 into P2. Nodes 12, 27, and 33 are summarized as structural combinations for deriving the topology network diagram. Finally, all segments in the segment group that satisfy a certain order of magnitude and state combination logic are used as the basis for generating node branches in the path structure diagram. After unified organization, multi-level path deduction structure fragments are generated.

[0029] Please see Figure 5 The coupling deconstruction module includes: The frequency comparison submodule is based on the multi-level path deduction structure segment, obtains the die closure amount data and node occurrence frequency in the stamping process, and compares the difference between the stamping frequency value and the node occurrence frequency based on the frequency distribution corresponding to the node. It filters the node set whose frequency deviation exceeds the stamping reference frequency difference value, establishes the occurrence frequency offset record of the corresponding node, and generates the occurrence frequency offset sequence. To acquire the die closure amount data and node occurrence frequency for each stamping operation, the die closure amount during the stamping process is first recorded using stroke counting. The die position code and closure stroke value are recorded for each closure. For example, in 20 stamping operations, the die closure amount sequence is recorded as [8.2mm, 8.1mm, 8.4mm…]. Simultaneously, a node detection counter is set at the die closure point to record the frequency of each node within a specified stamping cycle. For example, node A appears 18 times, node B appears 12 times, and node C appears 20 times. Based on the collected data, a node frequency distribution vector is constructed. The frequency of each node is compared with the stamping frequency to determine if there is a significant difference. The frequency difference comparison is performed using… Frequency deviation limits are used as screening criteria. For example, if the stamping reference frequency difference is set to ±2 times, in the frequency difference analysis, if node B appears 12 times while the total number of stampings is 20, the frequency difference is 8 times, which exceeds the set ±2 times range. This node is judged as a frequency deviation node and is included in the frequency offset set. The correspondence between the node and its frequency offset value is established to obtain the node offset value vector. For example, the offset value of node B is -8, the offset value of node A is -2, and the offset value of node C is 0. Nodes with an absolute offset value greater than 2 are selected through this vector to obtain the node group with significant frequency offset. The frequency fluctuation trend of the node group in the stamping sequence and the corresponding node number and stamping sequence position index are recorded. Finally, the frequency offset sequence is generated.

[0030] The path difference extraction submodule extracts the index position of the involved nodes in the path structure based on the frequency offset sequence, compares the index difference of corresponding nodes between paths, determines whether it exceeds the tolerance range of the interference path, extracts the path node combination with the difference exceeding the limit, obtains the path segment in the structure, and obtains the path interference segment index interval. Based on the frequency offset sequence, the index position of the involved node in the path structure is extracted. First, the node set contained in the frequency offset sequence is used to find the occurrence position of each node in the process path structure and the position is recorded in the form of path number and node index. For example, if node B is at index position 3 in path 1 and index position 5 in path 2, the index difference between the paths of node B is 2. After extracting the node index difference, it is determined whether the difference exceeds the tolerance range of the interference path. The tolerance range is preset to ±1. If the difference is 2, it is determined to be a path interference node. The corresponding path structure segment is extracted and the start index and end index are recorded. For example, the interference segment of node B is index 24 in path 1 and index 46 in path 2. All path segments determined to be interference nodes are merged and sorted by path number to finally obtain the path interference segment index range.

[0031] The interference zone generation submodule calls the structural node order and spacing of the corresponding path segment according to the path interference segment index interval, identifies the concentration degree and minimum interval of interference nodes in the segment, filters path segments whose concentration exceeds the interference baseline interval, obtains the node range, and generates the path interference decomposition segment area. Based on the path interference segment index interval, the structural node order and spacing of the corresponding path segment are called to identify the concentration and minimum interval of interference nodes in the segment. First, the actual structural order between nodes in the extracted path segment is analyzed. The node arrangement density is determined by the segment structure diagram. The spacing between nodes is measured in millimeters and the minimum interval value is calculated. For example, in the path 1 segment, the spacing between nodes B, D, and E is 2.1mm, 1.8mm, and 3.4mm respectively, so the minimum interval is 1.8mm. The minimum interval value is compared with the set interference tolerance between nodes. For example, the interference baseline spacing is 2.0mm. It is determined whether there are node pairs in the segment that are less than this threshold. If they exist, it is determined to be a node concentration segment. The node concentration parameter of each path segment is counted and compared with the interference baseline interval. If the concentration exceeds the set range, such as exceeding the node density standard of 3 nodes / 10mm, the segment is considered to be an interference segment. Its path number and node index number are extracted and marked as interference structure segment. Continuous interference segments under the same path number are integrated. The start and end node position numbers and interference status labels of each path segment are recorded. Finally, the path interference decomposition segment area is generated. As shown in Table 5, the node density in both Path 1 and Path 2 exceeds the set interference reference density (3 nodes / 10mm), and there are node pairs smaller than the interference reference spacing (2.0mm). Therefore, this segment of Path 1 and Path 2 can be identified as an interference segment and enter the disassembly area.

[0032] Please see Figure 6 The structural compression module includes: The path decomposition submodule decomposes the segment area based on path interference, collects the start and end position codes of the path segments and the distribution information of the number of segments, indexes and marks the segments according to the segment intervals, extracts the welding process dataset in the path segments, including welding current sequence, welding torch spatial position coordinates and welding speed trajectory, forms a path segmentation parameter table, and obtains the segment path parameter set; The path interference resolution operation treats segmented areas as the processing object, dividing the path trajectory into zones according to the start and end points of the interference. For example, if a jump in the welding torch position signal is detected in the 80mm to 175mm interval in the automatic welding path, segment numbers 1 to 3 can be set as start and end boundaries. Segment 1 is labeled as the 80mm to 110mm interval, segment 2 as the 110mm to 145mm interval, and segment 3 as the 145mm to 175mm interval, respectively. Subsequently, welding current, welding speed, and welding torch spatial position trajectory data in each segment need to be collected and synchronously recorded on the path acquisition platform. The welding current is collected through the current curve sequence from the power monitoring terminal at a sampling frequency of 100Hz, and the average current value of each segment is calculated by segment integration. For example, the average current value after integration is 120A for segment 1, 135A for segment 2, and 110A for segment 3. The welding speed is recorded by the trajectory following system. The gun translation speed was sampled at a frequency of 10Hz. The speed value was obtained by averaging the values ​​for each segment: 5.2mm / s for segment 1, 5.5mm / s for segment 2, and 4.9mm / s for segment 3. The path lengths were directly calculated from the coordinate trajectory, resulting in distances between segments of 80mm, 95mm, and 78mm, respectively. The welding gun position errors were obtained by comparing the end-effector encoder with the theoretical trajectory, recorded using the maximum error value method, and were 3.5mm, 2.8mm, and 3.1mm, respectively. The welding energy density was the energy value per unit area of ​​cladding welding, sampled from the measured infrared thermographic sequence. It was calculated by calibrating and solving the ratio of the input heat to the effective weld width for each weld segment, expressed in J / mm². The values ​​for segments 1-3 were 0.18, 0.22, and 0.19 J / mm², respectively. All parameter data were summarized in a table for subsequent identification steps, as shown below: As shown in Table 6, the path parameters have been fully collected and labeled by segment number. The data has been archived and packaged to generate a segment path parameter set.

[0033] The main control parameter identification submodule calls the segment path parameter set, constructs the segment main control parameter group based on welding current, welding torch position code, and welding speed, obtains the parameter arrangement order and position interval boundaries in the continuous path segment, and performs interval comparison and sequence matching of the segment path parameter group using the following formula: ; The algorithm calculates the overlap value of the path segment parameters, determines whether the overlap reaches the overlap threshold, filters the segment numbers that meet the overlap conditions, obtains the path segment parameter overlap discrimination results, and generates a set of indexes of repeated structure segments. Where Uᵢ represents the voltage value (V) of the i-th path segment, Iᵢ represents the current value (A) of the i-th path segment, Vᵢ represents the welding speed (mm / s) of the i-th path segment, Lᵢ represents the length (mm) of the i-th path segment, Δxᵢ represents the welding torch position coding error (mm) of the i-th segment, Pᵢ represents the welding energy density (J / mm²) of the i-th segment, Aᵢ represents the cross-sectional area of ​​the i-th welding path segment, α represents the correction coefficient (used to adjust the energy disturbance introduced by the position error), and R s Indicates the degree of overlap of the main control parameters (dimensionless). The overlap of the main control parameters of the path segment is used to measure the degree of overlap between the energy distribution and the welding process of each path segment. All parameters in the core formula need to be obtained through actual acquisition and sensor system calculation. The parameter acquisition and calculation are as follows: Path segment number: Path segment 1; U1: Path voltage value; The voltage is obtained by the power monitoring system using an industrial-grade sampling module, and the output voltage is stably controlled within the range of 24V to 28V.

[0034] The sampling frequency is 100Hz, and the effective values ​​within the sampling time period are processed by the RMS method to obtain the average voltage: ; I1: Path current value; The current value is acquired in real time through the power acquisition module and the Hall closed-loop current sensor. The average value within the sampling period is obtained by integrating the current sampling curve. ; L1: Path length; The start and end points of the path segment are calculated based on the coordinate differences in the trajectory encoding. The coordinate trajectory is located by a vision-location fusion navigation system, and the trajectory difference corresponds to the path length. ; V1: Welding speed; The welding torch movement speed is obtained by a trajectory tracking system at a sampling frequency of 10Hz, and the average speed is obtained by combining the actual travel distance with the time series. ; P1: Welding energy density; The infrared thermal imaging system acquires the thermal input spectrum of the weld, and estimates the energy input value per unit weld area by calculating the heat flux density of the image and reconstructing the weld morphology. ; A1: Welding cross-sectional area; The weld width and weld penetration depth are calculated using a weld contour recognition algorithm. Segment 1 has a width of 3.5 mm, a penetration depth of 2.7 mm, and a cross-sectional area of: ; Δx1: Position error; The error value is obtained by comparing the measured coordinates at the end of the welding torch sensor with the theoretical trajectory coordinates. ; α: Adjustment coefficient; This coefficient is derived from fitting a sample set of welding offset errors and represents the average effect of unit position offset on energy deviation. The empirical regression model gives a range of 0.06–0.09, as set in segment 1: ; Substitute each of the above parameters into the formula and calculate step by step as follows: Calculate energy inputs: ; Calculate the standard energy term: ; Since the path segment length is 95mm, it needs to be multiplied by the length to convert it to the total energy: ; Calculate the error adjustment energy term: ; Molecular merging: ; Denominator calculation: ; Overlap calculation of path segment 1: ; The results show that the actual energy input of path segment 1 during the welding process is much higher than its theoretical standard energy input value, with an overlap deviation of 309.39. This value deviates significantly from the theoretical target (overlap close to 1), indicating a serious anomaly in the energy distribution of the path segment. This numerical result is used as an overlap judgment index in path segment screening, classifying the segment as a non-overlapping or abnormal segment, marking it for entry into the path structure elimination or analysis module, and can be used for subsequent statistical analysis and process traceability. This formula evaluates the degree of overlap of the main control parameters by calculating the relative deviation between the actual welding energy and the standard energy of the welding path segment. The product of voltage, current, and path length, divided by the welding speed, represents the actual energy input during welding. The standard energy component, composed of welding energy density and weld cross-sectional area, represents the energy input per unit length under ideal conditions, further reflecting the process objective. Energy disturbances caused by positional deviations are represented by multiplying the error distance by energy density and area, and then multiplied by an adjustment coefficient to control its influence weight. Since the three energy terms are dimensionlessly uniform, they can be added or subtracted to obtain the deviation. Finally, the standard energy is used as the denominator for normalization, forming a dimensionless ratio. Absolute value calculations ensure that all deviation terms are non-negative, making the result stable and cumulative. The entire summation structure reflects the comprehensive overlap deviation of the welding path segment in terms of space and energy consumption; the closer to zero, the higher the overlap; the larger the deviation, the more mismatched the energy state of the welding path segment during welding. The path segment parameter overlap value measures the degree of matching between the actual energy input and the theoretical standard energy of a certain segment in the welding path. This value reflects the consistency between the parameter execution status and the process planning target during the execution of the welding task. The closer the value is to 1, the higher the consistency between the actual welding energy and the expected input, indicating that the process execution of the path segment is stable and reliable. The greater the deviation, the more abnormal fluctuations or execution errors there are in key parameters such as voltage, current, welding speed, or welding torch positioning in that path segment, which may lead to a decrease in weld quality or failure to meet structural performance standards. Therefore, the path segment parameter overlap value can serve as an important quantitative basis for screening path anomalies, identifying welding deviations, and evaluating the stability of the trajectory process, and is widely used in path segment reconstruction, interference identification, and quality control.

[0035] The structural segment integration submodule, based on the repeated structural segment index set, rearranges the indexes of repeated structural segments according to the original path order, adjusts the paragraph hierarchy and index mapping order of the path tree structure, reconstructs the path tree structure, integrates all paragraph numbers and main control parameter order information, and generates a reconstructable path paragraph structure set. Based on the segment numbers 1 to 3 marked in the repeated structural segment index set, the original path tree structure needs to be adjusted by index mapping. In the path structure tree construction module, the repeated segments are first rearranged in order of segment number, and the segment numbers are mapped to structural level nodes. The numerical feature identifiers of the path segment nodes are formed by the current, speed and length in the main control parameter group. Segment 1 is mapped to node ID_01, segment 2 to ID_02, and so on. The connection relationship between nodes is established according to the order of the segments to establish a directed path relationship. The relationship is stored in the path segment database using JSON structure encoding. At the same time, the segment level mapping table is updated, the main control parameter sequence is added to the structural nodes, and a complete mapping chain of segment-parameter-structural node is established by sequential numbering to form a structured path segment index set. Finally, a reconfigurable path segment structure set is generated.

[0036] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An intelligent process management system based on process parameter modeling, characterized in that, The system includes: The parameter acquisition module acquires the structural information of the process interface in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, establishes the mapping structure between parameters, and generates process input parameter structure blocks. The nested parsing module sets the heat treatment temperature setpoint as the dominant parameter based on the process input parameter structure block, and sets the protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code as dependent parameters. It arranges the hierarchy according to the control order, constructs a nested relationship graph, and generates a nested logic combination model. Based on the nested logic combination model, the path construction module calls the mold cavity temperature, cooling channel opening and closing signals and gate status identifiers in the injection molding process, identifies discontinuous nodes and marks them as path branch points, sorts the nodes according to the state relationship, and generates multi-level path deduction structure segments. The coupling deconstruction module calls the multi-level path deduction structure segment, compares the node frequency of stamping frequency with the die closure amount, identifies interference points based on position differences, extracts and reconstructs path segments, and generates path interference deconstruction segment areas.

2. The intelligent process management system based on process parameter modeling according to claim 1, characterized in that, The process input parameter structure block includes heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type, and start / stop time period code. The nested logic combination model includes dominant parameters, dependent parameters, parameter hierarchy, and nested relationship graph. The multi-level path deduction structure segment includes path stage, state discontinuity node, path branch point, path node sequence, and path segment. The path interference decomposition segment area includes node occurrence frequency, structural interference point, and area to be reconstructed.

3. The intelligent process management system based on process parameter modeling according to claim 1, characterized in that, The parameter acquisition module includes: The process interface acquisition submodule acquires the process interface structure information in the heat treatment process, extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code, classifies and groups the parameters according to the processing batch, and generates a classified and grouped data structure. Based on the classification and grouping data structure, the parameter structure extraction submodule classifies and pairs the heat treatment temperature setpoint, protective atmosphere flow rate and start-stop time period code in each batch, filters the differences in proportional integral derivative control type and workpiece material code within the group, combines the correspondence between control parameters and materials, and establishes a parameter mapping structure. The input structure generation submodule calls the parameter mapping structure to integrate the heat treatment temperature setting, protective atmosphere flow rate, workpiece material code and start / stop time period code corresponding to each batch, and combines them into a unified input parameter structure form to generate the process input parameter structure block.

4. The intelligent process management system based on process parameter modeling according to claim 3, characterized in that, The nested parsing module includes: The parameter identification submodule extracts the heat treatment temperature setpoint, protective atmosphere flow rate, workpiece material code, proportional-integral-derivative control type and start-stop time period code based on the process input parameter structure block, arranges the parameters in the order of their appearance, identifies the hierarchical relationship between control parameters and dependent parameters, and obtains the parameter hierarchical identification value. The control response parsing submodule calls the parameter level identification value, takes the heat treatment temperature setpoint as the dominant parameter, analyzes the difference between the protective atmosphere flow response time period and the temperature change time node, calculates the response ratio and determines whether the control sequence requirements are met, and obtains the response sequence judgment coefficient. The combined modeling submodule calls the response order judgment coefficient, filters data segments that meet the control order requirements, establishes a parameter nesting mapping matrix, combines proportional-integral-derivative control types and start-stop time period codes, constructs a multi-dimensional combined structure, obtains the nested topological responsivity of parameter combination relationships, and establishes a nested logic combination model.

5. The intelligent process management system based on process parameter modeling according to claim 4, characterized in that, The path construction module includes: The path recognition submodule, based on the nested logic combination model, calls three main parameters: mold cavity temperature, cooling channel opening and closing signal, and gate status identifier, to extract the node status in the path stage. By judging whether there are jump differences between states, comparing the change rate and state amplitude combination range, it filters out nodes with discontinuous states and generates a sequence of discontinuous state nodes. The path splicing submodule calls the state discontinuous node sequence, logically identifies the state change nodes according to the time sequence, and compares the combination order between nodes according to the temporal relationship of the gate state identifier, splicing them to form a hierarchical path segment structure, and obtaining the path segment splicing sequence. The deduction generation submodule compares and classifies the structural mapping combinations between path segments based on the path segment splicing sequence and the mold cavity temperature and cooling channel opening and closing signals. It identifies segment nodes with significant differences in path branches, filters structural combinations with strong path extensibility, and obtains multi-level path deduction structural segments.

6. The intelligent process management system based on process parameter modeling according to claim 5, characterized in that, The coupling deconstruction module includes: The frequency comparison submodule obtains the die closure amount data and node occurrence frequency in the stamping process based on the multi-level path deduction structure segment. Based on the frequency distribution corresponding to the node, it calls the stamping frequency value and the node occurrence frequency for difference comparison, filters the node set whose frequency deviation exceeds the stamping reference frequency difference value, establishes the occurrence frequency offset record of the corresponding node, and generates the occurrence frequency offset sequence. The path difference extraction submodule extracts the index position of the involved node in the path structure according to the occurrence frequency offset sequence, compares the index difference of corresponding nodes between paths, determines whether it exceeds the tolerance range of the interference path, extracts the path node combination with the difference exceeding the limit, obtains the path segment in the structure, and obtains the path interference segment index interval. The interference zone generation submodule, based on the path interference segment index interval, calls the structural node order and spacing of the corresponding path segment, identifies the concentration degree and minimum interval of interference nodes in the segment, filters path segments whose concentration exceeds the interference baseline interval, obtains their node range, and generates path interference decomposition segment area.

7. The intelligent process management system based on process parameter modeling according to claim 1, characterized in that, The system also includes: The structure compression module decomposes the segment area according to the path interference, extracts the welding current, welding gun position code and welding speed in the welding process, determines whether the main control parameters in the continuous path segment are completely overlapping, marks the repeated path structure segment, integrates it into the path tree structure and updates the index order, and generates a set of reconfigurable path segment structures. The reconfigurable path segment structure set includes repeating path segment structures, path tree structures, and index order.

8. The intelligent process management system based on process parameter modeling according to claim 7, characterized in that, The structure compression module includes: The path decomposition submodule decomposes the path interference segment area, collects the start and end position codes of the path segments and the distribution information of the number of segments, indexes and marks the segments according to the segment intervals, extracts the welding process dataset in the path segments, including the welding current sequence, the spatial position coordinates of the welding torch and the welding speed trajectory, forms the path segmentation parameter table, and obtains the segment path parameter set. The main control parameter identification submodule calls the segment path parameter set, constructs the segment main control parameter group based on welding current, welding gun position code and welding speed, obtains the parameter arrangement order and position interval boundary in the continuous path segment, compares the parameter values ​​within the interval and matches the sequence of the segment path parameter group, obtains the path segment parameter overlap value, determines whether the overlap reaches the overlap threshold, filters the segment number that meets the overlap condition, obtains the path segment parameter overlap discrimination result, and generates a repeating structure segment index set. The structural segment integration submodule, based on the repeated structural segment index set, rearranges the indexes of the repeated structural segments according to the original path order, adjusts the paragraph hierarchy and index mapping order of the path tree structure, reconstructs the path tree structure, integrates all paragraph numbers and main control parameter order information, and generates a reconstructable path paragraph structure set.

Images