A method for predicting the working hours of a blister processing procedure

Abstract

The application discloses a kind of blister processing process's work hour prediction method, specifically related to work hour prediction technical field;It is by obtaining the process state data corresponding to blister processing process, extracts the heat input change relationship between adjacent processes, and calculates the mapping relationship of temperature variation trajectory of sheet cooling stage and heat input change relationship of forming stage, generates the shrinkage difference representation data of workpiece;Combined with the actual time record data of demolding process and trimming process, analyze the time conduction relationship of workpiece shrinkage difference in demolding process and trimming process, generate inter-process conduction representation data;The work procedure time sequence in continuous process time sequence data set is sectioned and the section offset is calculated, and implicit work hour drift representation data is formed;Based on implicit work hour drift representation data and continuous process time sequence data set, construct work procedure work hour prediction model, realize the accurate work hour prediction of blister processing process.

Description

Technical Field

[0001] This invention relates to the field of time prediction technology, and more specifically, to a method for predicting the time of a thermoforming part processing step. Background Technology

[0002] In existing vacuum forming production lines, equipment temperature and process parameters are typically set based on experience or predetermined standard parameters, and unified production management is implemented for multiple vacuum-formed parts processed continuously. However, in actual production, due to the mutual influence of heat input and mold temperature between consecutive production steps, the thermal states of adjacent steps are often not entirely consistent. This leads to differences in the thermal states between each processing step, causing fluctuations in the dimensional stability of the vacuum-formed parts, ultimately resulting in significant differences in the actual processing time of each step.

[0003] Existing thermoforming processes cannot accurately consider the dynamic impact of thermal history differences between consecutive processes on the actual working time of the process, resulting in low accuracy of time prediction during production. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a method for predicting the working time of the thermoforming part processing steps to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for predicting the time required for thermoforming part processing steps includes the following steps:

[0007] S1: Obtain the process status data corresponding to the processing steps of the thermoformed parts, and align the time according to the process sequence to form a continuous process time sequence dataset;

[0008] S2: Based on the continuous process time sequence dataset, extract the relationship between the heat input changes in the forming stage between adjacent processes according to the continuous production sequence identification data, and construct process heat history inheritance characterization data;

[0009] S3: Based on the process thermal history inheritance characterization data, calculate the mapping relationship between the temperature change trajectory of the sheet cooling stage and the thermal input change relationship in the forming stage in the corresponding process, and generate part shrinkage difference characterization data;

[0010] S4: Based on the shrinkage difference characterization data of the part, combined with the time record data of the demolding process and the trimming process, analyze the time transmission relationship of the shrinkage difference of the part in the demolding process and the trimming process, and generate inter-process transmission characterization data;

[0011] S5: Based on the inter-process transmission characterization data, the process time series in the continuous process time series dataset is divided into segments and the segment offset is calculated to form implicit time drift characterization data.

[0012] S6: Based on the implicit time drift characterization data and the continuous process time series dataset, construct the process time prediction model and output the predicted time data of the corresponding thermoforming process.

[0013] In a preferred embodiment, the process status data includes sheet material parameter data, heating equipment operating status data, forming die temperature data, and continuous production sequence identification data.

[0014] In a preferred embodiment, S1 specifically refers to:

[0015] Extract sheet material parameter data corresponding to the thermoforming part processing steps from production records, extract heating equipment operating status data from equipment operation records, extract forming mold temperature data from temperature acquisition records, and extract continuous production sequence identification data from work order flow records;

[0016] The sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identifier data are correlated and time-aligned according to the occurrence time of the thermoforming part processing steps to form a continuous process time sequence dataset.

[0017] In a preferred embodiment, S2 specifically refers to:

[0018] The sequential relationship between adjacent thermoforming parts processing steps is determined based on the continuous production sequence identifier data in the continuous process time sequence dataset, and the operating status data of the heating equipment and the temperature data of the forming mold corresponding to the adjacent thermoforming parts processing steps are extracted.

[0019] Based on the sequential relationship of adjacent thermoforming parts processing steps, compare the changes in heating equipment operating status data and forming mold temperature data between the previous and subsequent thermoforming parts processing steps to generate the corresponding heat input change relationship of the forming stage of adjacent thermoforming parts processing steps.

[0020] The relationship between the heat input changes in the forming stages of each group of adjacent thermoforming parts processing steps is collected to form process thermal history inheritance characterization data.

[0021] In a preferred embodiment, S3 specifically refers to:

[0022] Based on the process thermal history inheritance characterization data, the thermal input variation relationship of each thermoforming part processing process is extracted for the forming stage corresponding to the forming stage.

[0023] Extract the forming mold temperature data for the corresponding thermoforming process from the continuous process time sequence dataset;

[0024] Based on the time of each thermoforming part processing step, the temperature data of the forming mold is continuously analyzed to form the temperature change trajectory of the sheet cooling stage corresponding to each thermoforming part processing step.

[0025] By correlating the temperature change trajectory during the sheet cooling stage with the heat input change during the forming stage, data characterizing the shrinkage difference of the part is generated.

[0026] In a preferred embodiment, S4 specifically refers to:

[0027] The correlation results of shrinkage difference for each thermoforming part processing step are determined based on the shrinkage difference characterization data.

[0028] Extract the demolding time record data and trimming time record data of the corresponding vacuum forming part processing process from the continuous process time sequence dataset;

[0029] Based on the time of each vacuum forming process, the shrinkage difference correlation results are matched with the demolding time record data and the trimming time record data to form the process time transfer record for each vacuum forming process.

[0030] Based on the process time transmission records, the time transmission relationship of part shrinkage difference between the demolding process and the trimming process is analyzed to form inter-process transmission characterization data.

[0031] In a preferred embodiment, S5 specifically refers to:

[0032] Extract the time transmission correlation results corresponding to each thermoforming part processing step based on the inter-process transmission characterization data;

[0033] Extract the process time sequence of the corresponding thermoforming part processing process from the continuous process time sequence dataset;

[0034] The process time series is divided into segments according to the consistency between the changing trends of adjacent times and the time transmission correlation results, forming a segment sequence of the corresponding thermoforming part processing process;

[0035] The segment offset is determined based on the time variation difference between adjacent segments in the segment sequence, forming implicit time drift characterization data.

[0036] In a preferred embodiment, S6 specifically refers to:

[0037] The drift state classification results corresponding to each thermoforming part processing step are determined based on the implicit time drift characterization data;

[0038] Extract sheet material parameter data, heating equipment operating status data, forming die temperature data, and process time series corresponding to the drift state classification results from the continuous process time series dataset;

[0039] Based on the drift state classification results, the process time series, sheet material parameter data, heating equipment operating status data, and forming die temperature data are combined to form the time prediction input data;

[0040] Regression analysis is performed based on the input data for time prediction and the time series of the process to construct a time prediction model for the process and output the predicted time data.

[0041] The technical effects and advantages of the method for predicting the processing time of a thermoforming part according to the present invention are as follows:

[0042] By associating and aligning sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identification data over time, a continuous process time-series dataset is formed. This transforms discrete process information in thermoforming parts processing into continuously analyzable time-series information. Further extraction of the thermal input variation relationship between adjacent processes during the forming stage and construction of process thermal history inheritance characterization data can represent the thermal state inheritance relationship between preceding and subsequent processes under continuous production conditions. Combining the temperature change trajectory during the sheet cooling stage generates part shrinkage difference characterization data, reflecting the impact of thermal state differences on part state changes. Furthermore, combining the time record data of the demolding and trimming processes generates inter-process transmission characterization data, representing the transmission relationship of part state changes to subsequent process time. Based on this, implicit time drift characterization data is formed, and a process time prediction model is constructed, improving the correspondence between predicted and actual time data in thermoforming parts processing. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of a method for predicting the time of a vacuum forming part processing step according to the present invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. Example 1

[0045] Figure 1 The present invention provides a method for predicting the time of processing steps in the vacuum forming process, which includes the following steps:

[0046] S1: Obtain the process status data corresponding to the processing steps of the thermoformed parts, and align the time according to the process sequence to form a continuous process time sequence dataset;

[0047] S2: Based on the continuous process time sequence dataset, extract the relationship between the heat input changes in the forming stage between adjacent processes according to the continuous production sequence identification data, and construct process heat history inheritance characterization data;

[0048] S3: Based on the process thermal history inheritance characterization data, calculate the mapping relationship between the temperature change trajectory of the sheet cooling stage and the thermal input change relationship in the forming stage in the corresponding process, and generate part shrinkage difference characterization data;

[0049] S4: Based on the shrinkage difference characterization data of the part, combined with the time record data of the demolding process and the trimming process, analyze the time transmission relationship of the shrinkage difference of the part in the demolding process and the trimming process, and generate inter-process transmission characterization data;

[0050] S5: Based on the inter-process transmission characterization data, the process time series in the continuous process time series dataset is divided into segments and the segment offset is calculated to form implicit time drift characterization data.

[0051] S6: Based on the implicit time drift characterization data and the continuous process time series dataset, construct the process time prediction model and output the predicted time data of the corresponding thermoforming process.

[0052] S1: Obtain the process status data corresponding to the processing steps of the thermoformed parts, and align the time according to the process sequence to form a continuous process time sequence dataset, including:

[0053] Extract sheet material parameter data corresponding to the thermoforming part processing steps from production records, extract heating equipment operating status data from equipment operation records, extract forming mold temperature data from temperature acquisition records, and extract continuous production sequence identification data from work order flow records;

[0054] The system reads production record data corresponding to each processing step of the thermoformed parts from the thermoforming parts production record system. This production record data includes, but is not limited to, information on the type of sheet material, sheet thickness, initial sheet dimensions, and batch information. For example, sheet material type information includes, but is not limited to, polypropylene, polyvinyl chloride, and polyethylene terephthalate. Sheet thickness information can be measured using a thickness gauge and recorded in the production record data. Initial sheet dimensions include the length and width of the sheet. Batch information is recorded using the material batch number or production date.

[0055] For each thermoforming part processing step, the system accesses the equipment operation records generated by the heating equipment used in that step. It then extracts the equipment operation status data of the heating equipment during the thermoforming part processing step from these records. This data includes, but is not limited to, real-time temperature data, heating power data, heater on / off status data, and running time data. For example, real-time temperature data is obtained through temperature sensors installed inside the heating equipment; heating power data is recorded in real-time by the heating equipment's power controller; heater on / off status data is recorded through relay on / off actions; and running time data is recorded using timestamps.

[0056] For each thermoforming part processing step, the temperature data of the forming mold throughout the entire processing cycle is obtained by accessing the temperature acquisition records corresponding to multiple temperature sensors installed on the inner surface of the thermoforming mold. The forming mold temperature data includes, but is not limited to, the temperature measurement values ​​at multiple temperature measurement points on the inner surface of the mold and the corresponding measurement time information. The temperature sensors on the inner surface of the mold are installed in a uniformly distributed manner, and the measurement point positions are determined according to the mold size and shape. The temperature measurement values ​​at each temperature measurement point are collected and recorded by the temperature sensors at regular intervals, and each measurement value corresponds to a collection timestamp to describe the temperature change trend as the processing step progresses.

[0057] For each thermoforming part processing step, continuous production sequence identifier data is extracted from the work order flow records stored in the production management database. This data represents the sequential production order of the thermoforming part processing steps and includes, but is not limited to, work order number, production batch number, and production line sequence number. For example, the work order number uses a standard production management coding method, the production batch number is pre-generated by the production scheduling system, and the production line sequence number is recorded by workshop operators during production according to the actual processing order and entered into the production management database. This represents the actual sequential production order of each thermoforming part processing step.

[0058] Based on the time of occurrence of the thermoforming part processing steps, the sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identification data are correlated and time-aligned to form a continuous process time sequence dataset.

[0059] Based on the actual occurrence time information corresponding to each thermoforming part processing step, time correlation and alignment are performed. The actual production sequence and start and end timestamps for each thermoforming part processing step are determined according to the continuous production sequence identifier data. Using these start and end timestamps as a benchmark, sheet material parameter data, heating equipment operating status data, and forming mold temperature data are correlated to ensure that all types of data remain synchronized on the same thermoforming part processing step's time scale. For example, for heating equipment operating status data, the timestamps of real-time temperature data, heating power data, and heater on / off status data are matched with the corresponding thermoforming part processing step's occurrence timestamp; similarly, for forming mold temperature data, a timestamp matching method is used to align the temperature measurement values ​​at the temperature measuring points on the inner surface of the mold with the corresponding thermoforming part processing step's occurrence timestamp.

[0060] Based on the associated and aligned sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identification data, a continuous process time-series dataset is formed through data fusion processing. Data fusion processing involves standardizing and integrating all data associated with each thermoforming part processing step using a unified data format structure. Each processing step corresponds to a complete set of time-series data records in the continuous process time-series dataset. Each set of time-series data records is expressed using a unified timestamp sequence, covering the entire time period from the start to the end of the thermoforming part processing step. Each timestamp corresponds to sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identification data. For example, the time-series data records can be stored in a unified time-series format, with each data field including, in sequence, timestamp, sheet type, sheet thickness, sheet size, real-time heating equipment temperature, heating power, heater on / off status, mold temperature measurement value, work order number, and production batch number, thus enabling the continuous process time-series dataset to express the actual state of the entire thermoforming part processing step.

[0061] S2: Based on the continuous process time-series dataset, extract the relationship between the heat input changes in the forming stages between adjacent processes according to the continuous production sequence identifier data, and construct process thermal history inheritance characterization data, including:

[0062] The sequential relationship between adjacent thermoforming parts processing steps is determined based on the continuous production sequence identifier data in the continuous process time sequence dataset, and the operating status data of the heating equipment and the temperature data of the forming mold corresponding to the adjacent thermoforming parts processing steps are extracted.

[0063] The sequential relationship between adjacent thermoforming part processing steps is determined based on the sequential production sequence identifier data contained in the continuous process time sequence dataset. Specifically, the sequential production sequence identifier data corresponding to each thermoforming part processing step is read sequentially from the continuous process time sequence dataset, including but not limited to work order number, production batch number, and production line sequence number. The sequential production sequence identifier data is arranged in order of the numerical value of the production line sequence number to determine the sequential relationship between adjacent thermoforming part processing steps. For example, if the production line sequence number is an Arabic numeral, the sequential relationship between each thermoforming part processing step and adjacent thermoforming part processing steps can be determined by sorting the production line sequence numbers in ascending order. Finally, the sequential relationship information of each pair of adjacent thermoforming part processing steps is recorded.

[0064] After determining the sequential relationship between adjacent thermoforming part processing steps, the operating status data of the heating equipment corresponding to the adjacent thermoforming part processing steps are extracted. Specifically, for each pair of adjacent thermoforming part processing steps, based on the actual timestamps of the processing steps recorded in the continuous process time sequence data, the operating status data of the heating equipment corresponding to the previous and subsequent thermoforming part processing steps are read respectively. The operating status data of the heating equipment includes, but is not limited to, real-time temperature data, heating power data, heater on / off status data, and running time data of the heating equipment. Taking the start to end timestamp of the processing step as the starting and ending range, all real-time temperature data, heating power data, heater on / off status data, and running time data of the heating equipment within the corresponding interval are extracted to obtain complete operating status data of the heating equipment for each processing step.

[0065] Extracting the forming mold temperature data corresponding to adjacent thermoforming part processing steps involves the following steps: For each pair of adjacent thermoforming part processing steps with a defined sequence, based on the actual timestamp of each thermoforming part processing step recorded in the continuous process time sequence dataset, the forming mold temperature data corresponding to the previous and subsequent thermoforming part processing steps are read from the continuous process time sequence dataset. The forming mold temperature data includes, but is not limited to, the temperature measurement values ​​and corresponding measurement time information at multiple temperature measuring points on the inner surface of the mold. Using the start and end timestamps of each thermoforming part processing step as a limit, all temperature measurement values ​​and corresponding timestamps recorded at each temperature measuring point on the inner surface of the mold within the corresponding time period are extracted to obtain a complete set of forming mold temperature data for each thermoforming part processing step.

[0066] Based on the sequential relationship of adjacent thermoforming parts processing steps, compare the changes in heating equipment operating status data and forming mold temperature data between the previous and subsequent thermoforming parts processing steps to generate the corresponding heat input change relationship of the forming stage of adjacent thermoforming parts processing steps.

[0067] For each pair of adjacent thermoforming part processing steps, the operating status data of the heating equipment in the subsequent thermoforming part processing step is compared with that in the preceding thermoforming part processing step. The real-time temperature data sequence of the heating equipment is numerically compared to determine the temperature change trend and magnitude between the two adjacent processing steps. The heating power data is also numerically compared to determine the differences in the power output state of the heating equipment between the two adjacent processing steps. The heater's on / off state data is compared moment-by-moment to determine whether the heater's operating state changes between the two adjacent processing steps. For example, by subtracting the temperature data sequence of the corresponding moment in the preceding processing step from the temperature data sequence measured in the subsequent processing step, a real-time temperature change trend curve can be obtained to characterize the changing pattern of the heating equipment's state.

[0068] Simultaneously, the temperature data of the forming mold in adjacent thermoforming parts processing steps are analyzed and compared. Specifically, for each pair of adjacent thermoforming parts processing steps, the temperature data of the forming mold corresponding to the subsequent thermoforming part processing step is compared point by point with the temperature data of the forming mold corresponding to the previous thermoforming part processing step. The temperature measurement values ​​of multiple temperature measuring points on the inner surface of the mold corresponding to the subsequent processing step and the previous processing step are compared to obtain the temperature change of each temperature measuring point between the two processes. Then, the overall temperature trend of the mold is judged based on the temperature change of each temperature measuring point. For example, by subtracting the temperature value of the corresponding temperature measuring point in the previous processing step from the temperature value of each temperature measuring point in the subsequent processing step, a set of mold temperature difference data is obtained to reflect the dynamic change trend of the forming mold temperature state in the continuous production process.

[0069] After completing the analysis and comparison of the heating equipment operating status data and the forming mold temperature data, the heat input change relationship of adjacent thermoforming parts processing steps is generated. Specifically, based on the real-time temperature change trend curve of the heating equipment, the heating power change value, the heater opening and closing status change information and the mold temperature difference data set obtained from the comparative analysis, a data set of heat input change relationship of the forming stage is generated. The integration method includes, but is not limited to, combining the heating equipment operating status data change results and the forming mold temperature data change results into a unified data set to form complete data that characterizes the dynamic differences in heat input between adjacent processing steps.

[0070] The relationship between the heat input changes in the forming stages corresponding to the processing steps of adjacent thermoformed parts is collected to form process thermal history inheritance characterization data.

[0071] The data collection and processing methods include, but are not limited to, orderly splicing together the data sets of heat input change relationships in all forming stages according to the continuous production sequence to generate a unified data structure that represents the heat input change patterns between each thermoforming part processing step; for example, using a timestamp sequence sorting method, arranging the heat input change relationship data of each pair of adjacent thermoforming part processing steps according to the processing sequence to form heat history inheritance characterization data covering all processing steps in the entire continuous production cycle.

[0072] S3: Based on the process thermal history inheritance characterization data, calculate the mapping relationship between the temperature change trajectory of the sheet cooling stage and the thermal input change relationship in the forming stage in the corresponding process, and generate part shrinkage difference characterization data, including:

[0073] Based on the process thermal history inheritance characterization data, the thermal input variation relationship of each thermoforming part processing process is extracted for the forming stage corresponding to the forming stage.

[0074] The thermal input variation data for each thermoforming part processing step is retrieved from the process thermal history inheritance characterization data according to the processing sequence corresponding to the continuous production sequence identifier data. The thermal input variation data for each forming stage includes the real-time temperature change trend curve of the heating equipment, the heating power change value, the heater opening and closing status change information, and the temperature difference data set of the forming mold. For each thermoforming part processing step, based on the timestamp information in the thermal input variation data for each forming stage, the real-time temperature change trend curve of the heating equipment is discretized at a uniform time interval to obtain the temperature change value sequence corresponding to continuous time points. At the same time, the heating power change values ​​are aligned at the same time interval. To form a power change sequence on a unified time scale; for heater on / off state change information, the on / off state is numerically encoded, with the heater on state encoded as 1 and the heater off state encoded as 0, thus forming a state sequence consistent with the time of the temperature change sequence and the power change sequence; for the temperature difference data set of the forming mold, the temperature difference values ​​of each temperature measuring point are arranged on the same time scale according to the order of the temperature measuring point location to form a multidimensional temperature difference data sequence; the temperature change sequence, power change sequence, state sequence and multidimensional temperature difference data sequence are combined to form a standardized forming stage heat input change relationship data structure corresponding to each thermoforming part processing step.

[0075] Extract the forming mold temperature data for the corresponding thermoforming process from the continuous process time sequence dataset;

[0076] After determining the time interval for each thermoforming part processing step based on the continuous production sequence identifier data, the forming mold temperature data within the corresponding time interval is read from the continuous process time sequence dataset. The forming mold temperature data includes temperature measurement values ​​at multiple temperature measuring points on the inner surface of the mold and corresponding timestamps. The read forming mold temperature data is then sorted by time, arranging the forming mold temperature data in ascending order of timestamps. Missing values ​​in the forming mold temperature data are then filled in. For timestamps where no temperature measurement values ​​are recorded, linear interpolation of temperature measurement values ​​from nearby time points is used to fill in the missing values, ensuring the continuity of the forming mold temperature data in the time series. Simultaneously, the temperature measurement values ​​at multiple temperature measuring points are spatially integrated, arranging the temperature measurement values ​​at multiple temperature measuring points in order of their positions to form a multi-dimensional temperature vector corresponding to each time point, thus obtaining a complete forming mold temperature data sequence.

[0077] Based on the time of each thermoforming part processing step, the temperature data of the forming mold is continuously analyzed to form the temperature change trajectory of the sheet cooling stage corresponding to each thermoforming part processing step.

[0078] Based on the temperature data sequence of the forming mold, the timestamps of the forming end time and the demolding start time in each thermoforming part processing step are determined. The time interval between the forming end time and the demolding start time is defined as the sheet cooling stage time interval. Within the sheet cooling stage time interval, the forming mold temperature data sequence is extracted to obtain the temperature data within the corresponding time period. The extracted temperature data is arranged continuously in chronological order to form an initial cooling temperature sequence. The initial cooling temperature sequence is smoothed using a moving average method, that is, the average of the temperature measurement values ​​at the current time point and the adjacent time points before and after is taken as the smoothed temperature value. For example, the temperature measurement values ​​of the two time points before and after the current time point are selected for the average calculation to eliminate random fluctuations in temperature measurement. Through the above processing, the smoothed sheet cooling stage temperature change trajectory is obtained. The temperature change trajectory is plotted with time on the horizontal axis and temperature value on the vertical axis, which fully describes the temperature change process of the sheet during the cooling stage.

[0079] By correlating the temperature change trajectory during the sheet cooling stage with the heat input change during the forming stage, data characterizing the shrinkage difference of the part is generated.

[0080] For each thermoforming part processing step, the temperature change trajectory of the sheet cooling stage is matched with the corresponding heat input change relationship data structure of the forming stage according to a time alignment method, so that the temperature change trajectory of the sheet cooling stage and the corresponding heat input change relationship data structure of the forming stage have consistent time coordinates on the same time scale. After time alignment, a joint analysis is performed on the temperature change trajectory of the sheet cooling stage and the heat input change relationship data structure of the forming stage. The joint analysis process includes calculating the temperature change rate of the sheet cooling stage and the heat input change rate of the forming stage, and constructing a shrinkage change index based on the difference between the temperature change rate and the heat input change rate. The temperature change rate during the cooling stage is calculated by the ratio of the temperature difference between adjacent time points to the time difference. The heat input change rate during the forming stage is calculated by the ratio of the heat input change value difference between adjacent time points to the time difference. The shrinkage change index is defined as the absolute value of the difference between the temperature change rate and the heat input change rate. By accumulating and summing the shrinkage change index at each time point, the shrinkage difference measurement value of the corresponding thermoforming part processing step is obtained. The shrinkage difference measurement values ​​corresponding to each thermoforming part processing step are arranged in the continuous production sequence to form part shrinkage difference characterization data, which is used to characterize the shrinkage changes caused by thermal history differences between different thermoforming part processing steps.

[0081] S4: Based on the shrinkage difference characterization data of the part, combined with the time record data of the demolding and trimming processes, analyze the time transmission relationship of the shrinkage difference of the part in the demolding and trimming processes, and generate inter-process transmission characterization data, including:

[0082] The correlation results of shrinkage difference for each thermoforming part processing step are determined based on the shrinkage difference characterization data.

[0083] The shrinkage difference metric value corresponding to each thermoforming part processing step is extracted from the shrinkage difference characterization data. The shrinkage difference metric value is obtained by accumulating the absolute value of the difference between the temperature change rate of the sheet cooling stage and the heat input change rate of the forming stage corresponding to each thermoforming part processing step.

[0084] The shrinkage difference correlation result is determined based on the shrinkage difference metric value corresponding to each thermoforming part processing step. This includes classifying the shrinkage difference metric value, that is, dividing it according to the range of the shrinkage difference metric value to form a shrinkage difference level. The method of dividing the shrinkage difference level is, for example, using the mean segmentation method, that is, calculating the average value of the shrinkage difference metric value corresponding to all thermoforming part processing steps, and dividing the shrinkage difference metric value into multiple intervals, including but not limited to low difference level, medium difference level, and high difference level. For example, when the shrinkage difference metric value of the thermoforming part processing step is less than 0.8 times the average value, it is defined as low difference level; when the shrinkage difference metric value is between 0.8 times and 1.2 times the average value, it is defined as medium difference level; and when the shrinkage difference metric value exceeds 1.2 times the average value, it is defined as high difference level.

[0085] Extract the demolding time record data and trimming time record data of the corresponding vacuum forming part processing process from the continuous process time sequence dataset;

[0086] Based on the continuous production sequence identification data, the timestamps for the demolding and trimming processes corresponding to each thermoforming part processing step are determined. Specifically, the start and end timestamps for the demolding and trimming processes are extracted from the actual start and end timestamps of each thermoforming part processing step recorded in the continuous process time sequence dataset. The demolding process time record data includes the time length between the actual start and end times of the demolding process, and the trimming process time record data includes the time length between the actual start and end times of the trimming process. For example, the time length record between the actual start and end times of the demolding process is accurate to the second, and this time length record is generated and stored in the continuous process time sequence dataset by the real-time automatic timing system on the production site. Similarly, the time length record between the actual start and end times of the trimming process is also generated by the on-site automatic timing system.

[0087] Based on the time of each vacuum forming process, the shrinkage difference correlation results are matched with the demolding time record data and the trimming time record data to form the process time transfer record for each vacuum forming process.

[0088] Based on the production sequence of each thermoforming part processing step determined by the continuous production sequence identification data, the level corresponding to the shrinkage difference correlation result of each thermoforming part processing step is determined. Simultaneously, the corresponding demolding and trimming time record data are extracted from the continuous process time sequence dataset. The shrinkage difference correlation result is matched step-by-step with the demolding and trimming time record data, meaning that the shrinkage difference correlation result of each thermoforming part processing step corresponds to the demolding and trimming time record data under a unified processing step number. Through matching, a process time transfer record is formed, including the thermoforming part processing step number, shrinkage difference correlation level, actual demolding time, and actual trimming time.

[0089] Based on the process time transmission records, the time transmission relationship of part shrinkage difference between the demolding process and the trimming process is analyzed to form inter-process transmission characterization data;

[0090] A joint analysis was conducted on the shrinkage difference correlation level, the actual time spent in the demolding process, and the actual time spent in the trimming process in the process time transfer record. The joint analysis used regression analysis to determine the statistical correlation between the shrinkage difference correlation level and the actual time spent in the demolding process and the actual time spent in the trimming process.

[0091] The shrinkage difference correlation level is converted into numerical data for analysis. For example, low, medium, and high difference levels are assigned numerical values ​​of 1, 2, and 3, respectively. Using the shrinkage difference correlation level as the input variable and the actual time spent in the demolding and trimming processes as the output variables, a linear regression model is constructed to analyze the impact of the shrinkage difference level on the time spent in the demolding and trimming processes. To accurately characterize the time transmission relationship of shrinkage difference from the demolding to the trimming process, inter-process transmission characterization data is constructed, including determining the parameters of the linear regression model for the impact of shrinkage difference on the time spent in the demolding and trimming processes. For example, the linear regression model can be expressed as:

[0092] T_Demolding process = α × grade + β; T_Trimming process = γ × grade + δ.

[0093] Wherein, T_demolding process represents the actual time spent on demolding, T_trimming process represents the actual time spent on trimming, and the level is numerical data representing the correlation level of shrinkage difference. α, β, γ, and δ are model regression parameters. The regression parameters α, β, γ, and δ are obtained through the least squares method, that is, by using the time transmission records of all processes and determining the optimal values ​​through mathematical optimization. For example, parameter α can characterize the increase in the actual time spent on demolding when the shrinkage difference level increases by one unit, and parameter γ can characterize the increase in the actual time spent on trimming when the shrinkage difference level increases by one unit. The inter-process transmission characterization data obtained through the above method can reflect the actual time influence law of shrinkage difference in continuous processes.

[0094] S5: Based on inter-process transmission characterization data, the process time series in the continuous process time series dataset is segmented and the segment offset is calculated to form implicit time drift characterization data, including:

[0095] Extract the time transmission correlation results corresponding to each thermoforming part processing step based on the inter-process transmission characterization data;

[0096] Parameters for linear regression models constructed for the demolding and trimming processes are extracted from the inter-process transmission characterization data, including but not limited to the linear regression model parameters α and β for the demolding process, and the linear regression model parameters γ and δ for the trimming process. Using the extracted linear regression model parameters, the predicted values ​​of the actual time consumption of the demolding and trimming processes in each thermoforming process are calculated. The calculation formula is as follows: the predicted time consumption of the demolding process is equal to the product of the linear regression model parameter α and the correlation level of the part shrinkage difference plus the linear regression model parameter β; the predicted time consumption of the trimming process is equal to the product of the linear regression model parameter γ and the correlation level of the part shrinkage difference plus the linear regression model parameter δ. For each thermoforming process, calculations are performed based on the numerical data of the correlation level of the part shrinkage difference (e.g., low difference level is 1, medium difference level is 2, high difference level is 3). The calculation results form the time transmission correlation results corresponding to each thermoforming process. The time transmission correlation results are a set of values, corresponding to the predicted time consumption of the demolding process and the predicted time consumption of the trimming process in each thermoforming process.

[0097] Extract the process time sequence of the corresponding thermoforming part processing process from the continuous process time sequence dataset;

[0098] For each thermoforming part processing step, the occurrence time range of each processing step is determined based on the continuous production sequence identifier data. Process time series data within this time range is extracted from the continuous process time series dataset. The process time series includes key time nodes such as the processing start time stamp, forming stage end time stamp, cooling stage start and end time stamp, demolding stage start and end time stamp, and trimming stage start and end time stamp for each thermoforming part processing step. The method for extracting the process time series data is as follows: based on the production batch number and production line sequence number of the thermoforming part processing step, the corresponding process time node data is filtered from the continuous process time series dataset according to the continuous production sequence. Accurate extraction is achieved through timestamp matching and truncation. After extraction, a process time series data covering each process stage is formed. For example, the process time series data corresponding to a single thermoforming part processing step is a series of timestamp data, expressing the exact start and end times of each stage with a precision of seconds.

[0099] The process time series is divided into segments according to the consistency between the changing trends of adjacent times and the time transmission correlation results, forming a segment sequence of the corresponding thermoforming part processing process;

[0100] For each thermoforming part processing step's time sequence, the time variation trend of adjacent stages in the time sequence is determined. The method for determining the time variation trend includes calculating the duration of each stage based on its start and end timestamps to determine the actual time consumption of each stage, and then calculating the difference in actual time consumption between adjacent stages to determine the time variation trend between adjacent stages. Based on the time propagation correlation results, the predicted time consumption of the demolding and trimming processes is used as standard values. By calculating the difference between the actual time consumption and the predicted time consumption, the specific trend of the deviation between the actual time consumption and the predicted time consumption is determined. The actual time consumption is then compared with the predicted time consumption. The actual time is divided into segments based on the consistency of the trend between the actual and predicted time. For example, when the difference between the actual and predicted time remains within a set threshold range (e.g., ±10 seconds), they are divided into the same segment. When the difference exceeds the set threshold range, they are divided into new segments. The threshold is determined by statistical analysis. That is, by analyzing the overall distribution pattern of the difference between the actual and predicted time in all historical data, a multiple of the standard deviation of the difference in all historical data (e.g., 1 standard deviation) is taken as the segment division threshold to ensure the consistency and stability of segment division.

[0101] The segment offset is determined based on the time variation difference between adjacent segments in the segment sequence, forming implicit time drift characterization data;

[0102] From the sequence of segments corresponding to each thermoforming part processing step, for each thermoforming part processing step, calculate the actual time difference between adjacent segments, that is, the actual total time of the later segment minus the actual total time of the earlier segment, to determine the time variation difference between adjacent segments; convert the time variation difference between each adjacent segment of each thermoforming part processing step into the corresponding segment offset. The method for determining the segment offset includes taking the first segment as the reference segment and calculating the offset of the actual time of each subsequent segment relative to the actual time of the reference segment, that is, the difference between the actual time of each segment and the actual time of the reference segment; for example, if the actual time of the reference segment is T_reference, and the actual time of the later segment is T_later segment, then the segment offset of the later segment is T_later segment minus T_reference; then all thermoforming part processing steps... The corresponding segment offsets are sorted according to the continuous production sequence identifier data to generate a complete segment offset sequence. By analyzing the variation pattern of the segment offset sequence throughout the entire continuous production cycle, the implicit time drift characteristics exhibited by each thermoforming part processing step are determined, ultimately forming implicit time drift characterization data. The implicit time drift characterization data is expressed in numerical form, with each data point recording the actual time offset state of each thermoforming part processing step relative to the reference segment, fully reflecting the true situation of implicit time drift of each processing step in the entire production process. The implicit time drift characterization data is organized through a standardized data structure, with each data record including the number of the thermoforming part processing step, the segment sequence number, the actual time consumed, and the corresponding segment offset, forming an accurate record of the time drift pattern in the actual production process.

[0103] S6: Based on implicit time drift characterization data and continuous process time series datasets, construct a process time prediction model and output the predicted time data for the corresponding thermoforming part processing process, including:

[0104] The drift state classification results corresponding to each thermoforming part processing step are determined based on the implicit time drift characterization data;

[0105] The segment offset sequence for each processing step of a thermoforming part is extracted from the implicit time drift characterization data. This segment offset sequence represents the actual time offset of each processing step relative to the baseline segment. Cluster analysis is performed on all segment offset data to classify the drift state. Cluster analysis methods include, but are not limited to, K-means clustering. An initial number of clusters, i.e., the number of cluster centers, is set. The silhouette coefficient evaluation method is used to determine the number of cluster centers. For example, before cluster analysis, preliminary clustering is performed for different numbers of clusters, and the silhouette coefficient is calculated under each cluster number condition. The number of clusters corresponding to the highest silhouette coefficient value is selected as the optimal number of clusters. The optimal clustering is then determined. After counting the clusters, the segment offset sequence is used as the input variable for cluster analysis, and the similarity is calculated according to the Euclidean distance between the segment offsets. The segment offset sequence is assigned to the nearest cluster center. After each assignment, the center position of each cluster is recalculated. This process is repeated until the change in the cluster center position is lower than a preset threshold, such as 0.001. After the cluster analysis, multiple drift states are formed, and the segment offset data within each drift state exhibits consistent drift characteristics. The drift state classification results of each thermoforming part processing step are recorded to form corresponding classification labels. For example, the drift state classification results can be labeled as slight drift state, moderate drift state, and severe drift state.

[0106] Extract sheet material parameter data, heating equipment operating status data, forming die temperature data, and process time series corresponding to the drift state classification results from the continuous process time series dataset;

[0107] Based on the drift state classification results, the data for each thermoforming part processing step is identified according to the continuous production sequence. Sheet material parameter data, heating equipment operating status data, forming mold temperature data, and process time series are extracted from the continuous process time series dataset. The data index in the continuous process time series dataset is located based on the thermoforming part processing step number. The corresponding sheet material parameter data is read according to the data index, including but not limited to sheet material type information, sheet material thickness information, initial sheet material size information, and sheet material batch information. Heating equipment operating status data is extracted according to the data index, including real-time temperature data, heating power data, heater on / off status data, and running time data. Forming mold temperature data is read according to the data index, including temperature measurements at multiple temperature measurement points on the inner surface of the mold and the corresponding measurement time information. The process time series is extracted one by one according to the data index, including processing start timestamp, forming stage end timestamp, cooling stage start and end timestamp, demolding stage start and end timestamp, and trimming stage start and end timestamp. Each data extraction operation is based on the actual start and end timestamps of the processing step to ensure that the extracted data are accurately matched and synchronized on the time scale.

[0108] Based on the drift state classification results, the process time series, sheet material parameter data, heating equipment operating status data, and forming die temperature data are combined to form the time prediction input data;

[0109] Using the drift state classification results as the basis for data combination, a complete dataset for each drift state is constructed. The corresponding process time series, sheet material parameter data, heating equipment operating status data, and forming mold temperature data under each drift state classification are aligned according to the same processing process number. During data combination, data format standardization methods include, but are not limited to: discretizing and interpolating the real-time temperature data of the heating equipment and the forming mold temperature data at uniform time intervals to form a fixed-length time series; for sheet material parameter data, the type information is processed using numerical encoding, such as using unique thermal encoding to mark different sheet types to ensure data consistency; based on the unified time axis after data alignment, the process time series information is transformed into a data structure of the duration of specific stages, recording the duration of each stage in seconds, thereby achieving data dimension unification; after unification, the dataset for each thermoforming part processing process forms the time prediction input data. Each dataset contains the drift state classification results, sheet material encoding information, real-time temperature time series of the heating equipment, heating power time series, heater on / off status sequence, mold temperature multidimensional sequence, and actual time consumption data for each stage.

[0110] Regression analysis is performed based on the input data for time prediction and the time series of the process to construct a process time prediction model and output the predicted time data.

[0111] Using the time prediction input data under each drift state category as the input variable and the actual time consumption of the corresponding processing step as the output variable, a process time prediction model is established using regression analysis. The regression analysis method can be multiple linear regression or nonlinear regression, such as random forest regression model or gradient boosting tree regression model. For example, when using gradient boosting tree regression model, multiple basic decision tree models are constructed through step-by-step iteration. Each iteration optimizes the residual of the previous model. During model training, parameter optimization methods are set, such as determining the optimal number of trees (e.g., the number of trees is set between 50 and 300, and the step size is set to 10) and the maximum tree depth (e.g., the maximum depth is set between 3 and 10) through cross-validation. At the same time, the loss function is defined as mean squared error (MSE) to ensure the optimization of prediction accuracy during model training. After the model parameters are determined, The training process employs a random partitioning of the training and test sets, with a clearly defined ratio of 8:2. This means 80% of the predicted time input data is used for model training, and the remaining 20% ​​is used for model testing, ensuring the model's generalization ability. The resulting gradient boosting tree regression model is the process time prediction model. This model can output the predicted time duration for each stage of the processing steps under different drift state classifications. The final output of the process time prediction model is the predicted time data for each processing step, including the predicted time for the forming stage, cooling stage, demolding stage, and trimming stage. Each predicted time is output in seconds. The output predicted time data fully reflects the actual time drift characteristics of the processing steps under each drift state classification and can serve as guidance data for actual production scheduling and efficiency improvement.

[0112] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0113] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0114] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0116] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0117] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0118] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0119] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0120] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for predicting the time required for a vacuum forming process, characterized in that, Includes the following steps: S1: Obtain the process status data corresponding to the processing steps of the thermoformed parts, and align the time according to the process sequence to form a continuous process time sequence dataset; S2: Based on the continuous process time sequence dataset, extract the relationship between the heat input changes in the forming stage between adjacent processes according to the continuous production sequence identification data, and construct process heat history inheritance characterization data; S3: Based on the process thermal history inheritance characterization data, calculate the mapping relationship between the temperature change trajectory of the sheet cooling stage and the thermal input change relationship in the forming stage in the corresponding process, and generate part shrinkage difference characterization data; S4: Based on the shrinkage difference characterization data of the part, combined with the time record data of the demolding process and the trimming process, analyze the time transmission relationship of the shrinkage difference of the part in the demolding process and the trimming process, and generate inter-process transmission characterization data; S5: Based on the inter-process transmission characterization data, the process time series in the continuous process time series dataset is divided into segments and the segment offset is calculated to form implicit time drift characterization data. S6: Based on implicit time drift characterization data and continuous process time series dataset, construct process time prediction model and output the predicted time data of corresponding thermoforming part processing process.

2. The method for predicting the time of a thermoforming part processing step according to claim 1, characterized in that, Process status data includes sheet material parameter data, heating equipment operating status data, forming die temperature data, and continuous production sequence identification data.

3. The method for predicting the time of a thermoforming part processing step according to claim 2, characterized in that, S1, specifically: Extract sheet material parameter data corresponding to the thermoforming part processing steps from production records, extract heating equipment operating status data from equipment operation records, extract forming mold temperature data from temperature acquisition records, and extract continuous production sequence identification data from work order flow records; The sheet material parameter data, heating equipment operating status data, forming mold temperature data, and continuous production sequence identifier data are correlated and time-aligned according to the occurrence time of the thermoforming part processing steps to form a continuous process time sequence dataset.

4. The method for predicting the time of a thermoforming part processing step according to claim 3, characterized in that, S2, specifically: The sequential relationship between adjacent thermoforming parts processing steps is determined based on the continuous production sequence identifier data in the continuous process time sequence dataset, and the operating status data of the heating equipment and the temperature data of the forming mold corresponding to the adjacent thermoforming parts processing steps are extracted. Based on the sequential relationship of adjacent thermoforming parts processing steps, compare the changes in heating equipment operating status data and forming mold temperature data between the previous and subsequent thermoforming parts processing steps to generate the corresponding heat input change relationship of the forming stage of adjacent thermoforming parts processing steps. The relationship between the heat input changes in the forming stages of each group of adjacent thermoforming parts processing steps is collected to form process thermal history inheritance characterization data.

5. The method for predicting the time of a vacuum forming part processing step according to claim 4, characterized in that, S3, specifically: Based on the process thermal history inheritance characterization data, the thermal input variation relationship of each thermoforming part processing process is extracted for the forming stage corresponding to the forming stage. Extract the forming mold temperature data for the corresponding thermoforming process from the continuous process time sequence dataset; Based on the time of each thermoforming part processing step, the temperature data of the forming mold is continuously analyzed to form the temperature change trajectory of the sheet cooling stage corresponding to each thermoforming part processing step. By correlating the temperature change trajectory during the sheet cooling stage with the heat input change during the forming stage, data characterizing the shrinkage difference of the part is generated.

6. The method for predicting the time of a thermoforming part processing step according to claim 5, characterized in that, S4, specifically: The correlation results of shrinkage difference for each thermoforming part processing step are determined based on the shrinkage difference characterization data. Extract the demolding time record data and trimming time record data of the corresponding vacuum forming part processing process from the continuous process time sequence dataset; Based on the time of each vacuum forming process, the shrinkage difference correlation results are matched with the demolding time record data and the trimming time record data to form the process time transfer record for each vacuum forming process. Based on the process time transmission records, the time transmission relationship of part shrinkage differences between the demolding and trimming processes is analyzed to form inter-process transmission characterization data.

7. The method for predicting the time of a vacuum forming part processing step according to claim 6, characterized in that, S5, specifically: Extract the time transmission correlation results corresponding to each thermoforming part processing step based on the inter-process transmission characterization data; Extract the process time sequence of the corresponding thermoforming part processing process from the continuous process time sequence dataset; The process time series is divided into segments according to the consistency between the changing trends of adjacent times and the time transmission correlation results, forming a segment sequence of the corresponding thermoforming part processing process; The segment offset is determined based on the time variation difference between adjacent segments in the segment sequence, forming implicit time drift characterization data.

8. The method for predicting the time of processing steps in a thermoforming part according to claim 7, characterized in that, S6, specifically: The drift state classification results corresponding to each thermoforming part processing step are determined based on the implicit time drift characterization data; Extract sheet material parameter data, heating equipment operating status data, forming die temperature data, and process time series corresponding to the drift state classification results from the continuous process time series dataset; Based on the drift state classification results, the process time series, sheet material parameter data, heating equipment operating status data, and forming die temperature data are combined to form the time prediction input data; Regression analysis is performed based on the input data for time prediction and the time series of the process to construct a time prediction model for the process and output the predicted time data.

Images