High-precision preparation method of flexible die-cutting circuit board

Abstract

The invention relates to the technical field of data processing, in particular to a high-precision preparation method of a flexible die cutting circuit board. S100, acquiring a pressure sequence of the flexible circuit board when the flexible circuit board is subjected to circular cutter die cutting to generate a space-time cubic pressure body; s200, calculating jitter characteristics on a time axis according to the space-time cubic pressure body; s300, abnormal flexible circuit boards are removed according to the jitter characteristics, and an abnormal report is generated. According to the method and the device, accurate tracing of line abnormity generation reasons can be realized, and the production quantity of unqualified products is reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, in particular to a high-precision preparation method of a flexible die-cut circuit board. BACKGROUND

[0002] The round knife die-cutting of a flexible circuit board (FPC) is a key process in its preparation process. The current quality control method has significant technical defects: such as the prior art CN116963395A, which mainly screens unqualified products by analyzing the pressure image of the single board after die-cutting, and cannot intervene in real time when defects occur, resulting in continuous generation of waste products. The existing method regards each FPC as an independent individual and analyzes its static pressure distribution. However, the round knife die-cutting is a continuous dynamic process, and the wear of the tool, the slight changes of the material or the vibration of the equipment will leave traces on the pressure data over time. When the quality problem of the circuit board is found, it is difficult to quickly locate whether the tool or the material batch is the problem, that is, it is difficult to accurately locate the cause of the accident, thereby reducing the production efficiency. SUMMARY

[0003] The purpose of the present application is to provide a high-precision preparation method of a flexible die-cut circuit board to solve one or more technical problems existing in the prior art and at least provide a beneficial choice or create conditions.

[0004] In order to achieve the above-mentioned purpose, a high-precision preparation method of a flexible die-cut circuit board is provided, which includes the following steps: S100, obtaining a spatiotemporal cubic pressure body of a flexible circuit board when the flexible circuit board is subjected to round knife die-cutting; S200, calculating a jitter feature on a time axis according to the spatiotemporal cubic pressure body; S300, removing abnormal flexible circuit boards according to the jitter feature and generating an abnormal report; Further, in S100, the specific method for obtaining the pressure sequence of the flexible circuit board during the circular die-cutting process to generate a spatiotemporal cubic pressure body is as follows: When the circular die-cutting processing device cuts the flexible circuit board, the pressure data of the contact area between the die-cutting roller and the flexible circuit board is continuously collected at a time resolution of 10 frames per second through the distributed pressure sensor array on the circular die-cutting machine to form a pressure sequence. All the pressure values ​​in the pressure sequence collected in each frame are arranged according to the position of the contact area between the die-cutting roller and the flexible circuit board to form a multi-frame pressure snapshot of the flexible circuit board. All the pressure snapshots are stacked according to the collection time order to generate a spatiotemporal cubic pressure body. Let i be the sequence number of the pressure snapshot, P(x, y, ti) represent the pressure snapshot of the flexible circuit board obtained at the i-th time, where ti represents the collection time of the sensor array, and P(xj, yk, ti) represents the pressure value in the j-th row and k-th column of the pressure snapshot of the flexible circuit board obtained at the i-th time. Furthermore, the sensor uses a TekScan A201 thin-film pressure sensor with a range of 0-100 psi, an accuracy of ±2.5%, and a resolution of 100 sensels / inch².

[0005] Furthermore, in S200, the method for calculating the spatiotemporal jitter coefficient on the time axis based on the spatiotemporal cubic pressure volume is as follows: within the range of i, the pressure values ​​contained in the pressure snapshots collected by the flexible circuit board at each acquisition time are iterated, and the formula is used to calculate the pressure values. Calculate the average pressure u(t) of the pressure snapshot at time t using the formula. Calculate the pressure distribution value of the pressure snapshot at time t. (t), where N represents the total number of rows of pressure values ​​in each frame of pressure snapshot, and M represents the total number of columns of pressure values ​​in each frame of pressure snapshot; The pressure values ​​of each point in the pressure snapshots collected at each acquisition time in the spacetime cubic pressure volume are traversed in chronological order. The pressure values ​​corresponding to each point are arranged into a timeline sequence according to the acquisition time. The maximum pressure value is marked as max_P(x,y), and the minimum pressure value is marked as min_P(x,y). The time interval between the maximum and minimum pressure values ​​is calculated as TPi, which serves as the limiting window for the pressure snapshot collected at acquisition time i. Limiting windows for all pressure snapshots in the spacetime cubic pressure volume are obtained, and the minimum value of all limiting windows is selected as the detection window. The average pressure and pressure distribution values ​​of all pressure snapshots within the detection window are traversed within the pressure cube according to the formula... Calculate the spatial jitter coefficient Scff within the detection window; The spatial jitter coefficient (Scff) quantifies the spatial uniformity of the entire pressure field over time. It effectively detects localized pressure anomalies caused by uneven tool wear or installation deviations. The spatial jitter coefficient consists of two parameters: the average pressure and the pressure distribution within a unit pressure snapshot. The average pressure reflects the magnitude of the average pressure on the flexible circuit board, while the pressure distribution reflects the uniformity of the pressure distribution. If tool wear or installation deviations occur during operation, these will be reflected in the average and distribution pressure values ​​during circuit board cutting. Since tool wear is a dynamic process, manifesting as notching and dulling, these factors will cause regular changes in the pressure snapshot's values ​​within a localized or range-wide area, such as gradually increasing or decreasing. Because the exact moment of change cannot be determined due to variations in processing time or the different materials of the flexible circuit board, the detection window is determined by the distribution of the extreme values ​​of pressure over time. The detection window changes with the processing process. By continuously adjusting the detection window based on the range of pressure extreme values ​​at the same historical point, the quality of the flexible circuit boards produced at different times can be adaptively monitored. When tool wear occurs, the jitter coefficient Scff increases due to the increased pressure distribution value, and the jitter coefficient Scff also increases with the degree of wear. At the same time, the jitter coefficient Scff amplifies the small changes in data caused by tool dulling through an exponential function. In addition, to avoid the instantaneous increase in data caused by abrupt changes in local data when the tool is notched, the results are normalized by extreme values ​​to reduce the impact of data on the calculation speed.

[0006] In the spatiotemporal cubic pressure volume, the recording time of all pressure values ​​in any column of the sensors in the first pressure snapshot is obtained. The time difference from the first pressure value to the last pressure value is calculated as Δt using a cross-correlation algorithm. The distribution distance between any two adjacent sensors in the same column of the sensor array is obtained as Δx. (Velocity) is used as the pressure feedback coefficient T_(1) for the first pressure snapshot. Within the range of t, the pressure feedback coefficients T_(t) for all pressure snapshots are obtained, according to the formula... Calculate the time jitter coefficient Tcff within the detection window; The max(·) function is the maximum value function, and △T_(t) represents the difference between all pairs of adjacent pressure feedback coefficients; The time jitter coefficient (Tcff) is used to analyze the dynamic characteristics of pressure propagation on the material surface. Its changes can sensitively reflect changes in material properties or drifts in process parameters. The pressure feedback coefficient is the pressure feedback from the sensor to the flexible circuit board. More specifically, the pressure feedback coefficient can reflect the increase in pressure feedback coefficient due to the difference in force propagation between the flexible circuit board and normal materials when there are material problems. The time jitter coefficient traverses the changes of all pressure feedback coefficients in the spatiotemporal cubic pressure body through the detection window, and calculates the standard deviation through the extreme values ​​of the pressure feedback coefficient changes. When material problems occur, they will be quickly reflected, improving the sensitivity of detection.

[0007] Furthermore, in S300, the method for eliminating flexible circuit boards with abnormalities based on the jitter coefficient and generating an anomaly report is as follows: The spatial jitter coefficients Scff and Tcff calculated from all acquisition windows are used to construct the spatial sequence SC and the time sequence TC, respectively. On the spatial sequence, the first spatial jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time as the adjacent coefficient. The acquisition time of the adjacent coefficient and the time length of the most recent acquisition time within 24 hours are defined as Time1. The spatial sequence is divided into multiple time periods according to the length of Time1. Any time period is selected as the target time period Period. The average value of all spatial jitter coefficients in the target time period is calculated as the reference coefficient of the target time period. The absolute value of the difference between all spatial jitter coefficients in the target time period and the reference coefficient is calculated in turn. The maximum value of the absolute value of the difference is taken as the spatial offset O of the target coefficient. Let the set of target time periods be {Period1, Period2, ..., Periodr}, and the spatial offset of each target time period be denoted as {O1, O2, ..., Or}, where r represents the index of the target time period; traverse the spatial offsets of all target time periods in chronological order, and starting from the second target time period reaching the spatial offset, subtract the spatial offset of the current target time period from the spatial offset of the next target time period in turn, and take the ratio of the absolute value of the difference result to the absolute value of the sum of squares from the first spatial offset to the current spatial offset as the spatial offset difference of the current target. Starting from the second target time period, iterate through the spatial offset differences of all target time periods in the set. If the spatial offset difference of the current target time period is greater than the spatial offset difference of the previous target time period, and also greater than the spatial offset difference of the next target time period, and greater than the average spatial offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level I. Spatial offset difference represents the difference in spatial offset within a certain target time period compared to the overall target time period across all target time periods. It can effectively eliminate interference caused by normal process fluctuations. Since the offset difference is a dimensionless parameter calculated based on adjacent spatial offsets, it can provide timely warnings of minute changes occurring in a short period of time. When the die-cutting blade wears, it manifests as a slight and continuous change in the pressure value. Spatial offset difference can achieve millisecond-level response, providing maintenance personnel with a timely processing window and avoiding the continuous generation of defective circuit boards due to the difficulty in detecting the initial and middle stages of blade wear. The above methods can respond and warn of changes in the external production environment in a timely manner. However, if there are defects in the internal materials of the flexible circuit board, especially when there are local differences in the internal materials of the flexible circuit board, it is difficult to directly reflect the pressure value obtained when facing the cutting of the die-cutting blade. However, due to the pressure transmission properties of different materials, there will be a slight offset on the time axis when the sensor responds to the pressure value. The degree of offset is proportional to the degree of structural difference.

[0008] In the time series, the time jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time and taken as the adjacent coefficient. The time length between the acquisition time of the adjacent coefficient and the most recent acquisition time within 24 hours is defined as Time2. The time series is divided into multiple time periods according to the length of Time2, and the mean time jitter coefficient in each time period is calculated in turn. The standard deviation of the jitter coefficient is calculated as u1 based on all jitter coefficients in each time period. The absolute value of the difference between jitter coefficients between adjacent time periods within each time period is calculated as the neighbor difference value. The standard deviation of all neighbor differences within each time period is calculated as u2. Each time period contains one jitter coefficient mean and multiple neighbor differences Diffs, where s represents the index of the neighbor difference value, and Diffs represents the s-th neighbor difference value. This is achieved using the formula... Calculate the time offset difference for each time period. Num2 is the number of time jitter coefficients in the time period, and Num1 represents the number of times in the time period that the current time jitter coefficient is greater than the average time jitter coefficient. The time offset difference calculated using the above method can capture the slight shift in response time caused by changes in circuit board materials during sensor data acquisition. When changes in circuit board materials cause a shift in sensor response time, the time jitter coefficient will undergo a small transition. If the frequency of material changes increases, the number of transition points will also increase. This can occur when there are one or more transition points within a defined time period. In the above method, the standard deviation of the time offset value will increase, and the size of Num1 will also increase, leading to a greater time offset difference. The above method is simple and effective and can be applied on a large scale in the production process.

[0009] Starting from the second target time period, iterate through the time offset differences of all target time periods in the set. If the time offset difference of the current target time period is greater than the time offset difference of the previous target time period, and also greater than the time offset difference of the next target time period, and greater than the average time offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level II. The system iterates through all abnormal circuit board markers, counts the number of circuit boards marked as Level I, and when the number of Level I circuit boards exceeds a preset threshold, the control console issues a Level I warning, quickly locates the abnormal period, and generates an abnormal feedback report. It also counts the number of circuit boards marked as Level II, and when the number of Level II circuit boards exceeds a preset threshold, the control console issues a Level II warning, quickly locates the abnormal period, and generates an abnormal feedback report. All abnormal reports are sent to the control terminal, where staff take corresponding maintenance measures based on the reports.

[0010] Beneficial effects: This invention uses sensors to collect the pressure values ​​of flexible circuit boards in a model in real time, forming a spatiotemporal cubic pressure body. Based on the spatiotemporal cubic pressure body, it calculates the spatial and temporal jitter characteristics of the flexible circuit board. By using the spatial and temporal jitter characteristics of the flexible circuit board, abnormal circuit boards can be marked and located. This not only enables real-time monitoring of circuit board quality but also generates feedback reports, overcoming the current inability to accurately locate the cause of accidents. It achieves accurate tracing of the causes of circuit abnormalities, improves the production efficiency of circuit boards, and reduces the number of defective products. Attached Figure Description

[0011] Figure 1 The diagram shows a flowchart of a high-precision fabrication method for flexible die-cut circuit boards. Detailed Implementation

[0012] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0013] Example 1 Figure 1 The diagram shows a flowchart of a high-precision fabrication method for flexible die-cut circuit boards.

[0014] Reference Figure 1 This invention proposes a high-precision manufacturing method for flexible die-cut circuit boards, the method comprising the following steps: S100, obtain the pressure sequence of the flexible circuit board when it is being die-cut by a circular cutter to generate a spacetime cubic pressure body; S200, calculates the jitter characteristics on the time axis based on the spatiotemporal cubic pressure body; S300, based on jitter characteristics, rejects abnormal flexible circuit boards and generates an anomaly report; Further, in S100, the specific method for obtaining the pressure sequence of the flexible circuit board during the circular die-cutting process to generate a spatiotemporal cubic pressure body is as follows: When the circular die-cutting processing device cuts the flexible circuit board, the pressure data of the contact area between the die-cutting roller and the flexible circuit board is continuously collected at a time resolution of 10 frames per second through the distributed pressure sensor array on the circular die-cutting machine to form a pressure sequence. All the pressure values ​​in the pressure sequence collected in each frame are arranged according to the position of the contact area between the die-cutting roller and the flexible circuit board to form a multi-frame pressure snapshot of the flexible circuit board. All the pressure snapshots are stacked according to the collection time order to generate a spatiotemporal cubic pressure body. Let i be the sequence number of the pressure snapshot, P(x, y, ti) represent the pressure snapshot of the flexible circuit board obtained at the i-th time, where ti represents the collection time of the sensor array, and P(xj, yk, ti) represents the pressure value in the j-th row and k-th column of the pressure snapshot of the flexible circuit board obtained at the i-th time. Furthermore, the sensor uses a TekScan A201 thin-film pressure sensor with a range of 0-100 psi, an accuracy of ±2.5%, and a resolution of 100 sensels / inch².

[0015] Furthermore, in S200, the method for calculating the spatiotemporal jitter coefficient on the time axis based on the spatiotemporal cubic pressure volume is as follows: within the range of i, the pressure values ​​contained in the pressure snapshots collected by the flexible circuit board at each acquisition time are iterated, and the formula is used to calculate the pressure values. Calculate the average pressure u(t) of the pressure snapshot at time t using the formula. Calculate the pressure distribution value of the pressure snapshot at time t. (t), where N represents the total number of rows of pressure values ​​in each frame of pressure snapshot, and M represents the total number of columns of pressure values ​​in each frame of pressure snapshot; The pressure values ​​of each point in the pressure snapshots collected at each acquisition time in the spacetime cubic pressure volume are traversed in chronological order. The pressure values ​​corresponding to each point are arranged into a timeline sequence according to the acquisition time. The maximum pressure value is marked as max_P(x,y), and the minimum pressure value is marked as min_P(x,y). The time interval between the maximum and minimum pressure values ​​is calculated as TPi, which serves as the limiting window for the pressure snapshot collected at acquisition time i. Limiting windows for all pressure snapshots in the spacetime cubic pressure volume are obtained, and the minimum value of all limiting windows is selected as the detection window. The average pressure and pressure distribution values ​​of all pressure snapshots within the detection window are traversed within the pressure cube according to the formula... Calculate the spatial jitter coefficient Scff within the detection window; In the spatiotemporal cubic pressure volume, the recording time of all pressure values ​​in any column of the sensors in the first pressure snapshot is obtained. The time difference from the first pressure value to the last pressure value is calculated as Δt using a cross-correlation algorithm. The distribution distance between any two adjacent sensors in the same column of the sensor array is obtained as Δx. (Velocity) is used as the pressure feedback coefficient T_(1) for the first pressure snapshot. Within the range of t, the pressure feedback coefficients T_(t) for all pressure snapshots are obtained, according to the formula... Calculate the time jitter coefficient Tcff within the detection window; The max(·) function is the maximum value function, and △T_(t) represents the difference between all pairs of adjacent pressure feedback coefficients; Furthermore, in S300, the method for eliminating flexible circuit boards with abnormalities based on the jitter coefficient and generating an anomaly report is as follows: The spatial jitter coefficients Scff and Tcff calculated from all acquisition windows are used to construct the spatial sequence SC and the time sequence TC, respectively. On the spatial sequence, the first spatial jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time as the adjacent coefficient. The acquisition time of the adjacent coefficient and the time length of the most recent acquisition time within 24 hours are defined as Time1. The spatial sequence is divided into multiple time periods according to the length of Time1. Any time period is selected as the target time period Period. The average value of all spatial jitter coefficients in the target time period is calculated as the reference coefficient of the target time period. The absolute value of the difference between all spatial jitter coefficients in the target time period and the reference coefficient is calculated in turn. The maximum value of the absolute value of the difference is taken as the spatial offset O of the target coefficient. Let the set of target time periods be {Period1, Period2, ..., Periodr}, and the spatial offset of each target time period be denoted as {O1, O2, ..., Or}, where r represents the index of the target time period; traverse the spatial offsets of all target time periods in chronological order, and starting from the second target time period reaching the spatial offset, subtract the spatial offset of the current target time period from the spatial offset of the next target time period in turn, and take the ratio of the absolute value of the difference result to the absolute value of the sum of squares from the first spatial offset to the current spatial offset as the spatial offset difference of the current target. Starting from the second target time period, iterate through the spatial offset differences of all target time periods in the set. If the spatial offset difference of the current target time period is greater than the spatial offset difference of the previous target time period, and also greater than the spatial offset difference of the next target time period, and greater than the average spatial offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level I. In the time series, the time jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time and taken as the adjacent coefficient. The time length between the acquisition time of the adjacent coefficient and the most recent acquisition time within 24 hours is defined as Time2. The time series is divided into multiple time periods according to the length of Time2, and the mean time jitter coefficient in each time period is calculated in turn. The standard deviation of the jitter coefficient is calculated as u1 based on all jitter coefficients in each time period. The absolute value of the difference between jitter coefficients between adjacent time periods within each time period is calculated as the neighbor difference value. The standard deviation of all neighbor differences within each time period is calculated as u2. Each time period contains one jitter coefficient mean and multiple neighbor differences Diffs, where s represents the index of the neighbor difference value, and Diffs represents the s-th neighbor difference value. This is achieved using the formula... Calculate the time offset difference for each time period. Num2 is the number of time jitter coefficients in the time period, and Num1 represents the number of times in the time period that the current time jitter coefficient is greater than the average time jitter coefficient. Starting from the second target time period, iterate through the time offset differences of all target time periods in the set. If the time offset difference of the current target time period is greater than the time offset difference of the previous target time period, and also greater than the time offset difference of the next target time period, and greater than the average time offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level II. The system iterates through all abnormal circuit board markers, counts the number of circuit boards marked as Level I, and when the number of Level I circuit boards exceeds a preset threshold, the control console issues a Level I warning, quickly locates the abnormal period, and generates an abnormal feedback report. It also counts the number of circuit boards marked as Level II, and when the number of Level II circuit boards exceeds a preset threshold, the control console issues a Level II warning, quickly locates the abnormal period, and generates an abnormal feedback report. All abnormal reports are sent to the control terminal, where staff take corresponding maintenance measures based on the reports.

[0016] Although the invention has been described in considerable detail and particularly with regard to several of the described embodiments, it is not intended to limit itself to any of these details or embodiments or any particular embodiment, thereby effectively covering the intended scope of the invention. Furthermore, the invention has been described above with respect to embodiments foreseeable by the inventors in order to provide a useful description, and non-substantial modifications to the invention that have not yet been foreseen may still represent equivalent modifications.

Claims

1. A high-precision manufacturing method for flexible die-cut circuit boards, characterized in that, The method includes the following steps: S100, obtain the pressure sequence of the flexible circuit board when it is being die-cut by a circular cutter to generate a spacetime cubic pressure body; S200, calculates the jitter characteristics on the time axis based on the spatiotemporal cubic pressure body; The S300 eliminates abnormal flexible circuit boards based on jitter characteristics and generates an anomaly report.

2. The high-precision manufacturing method for a flexible die-cut circuit board according to claim 1, characterized in that, In S100, the method for generating a spatiotemporal cubic pressure body by obtaining the pressure sequence of the flexible circuit board during rotary die cutting is as follows: When the rotary die cutting processing device cuts the flexible circuit board, the pressure data of the contact area between the die cutting roller and the flexible circuit board is continuously collected by the distributed pressure sensor array on the rotary die cutting machine at a time resolution of 10 frames per second to form a pressure sequence. The pressure values ​​of all the pressure values ​​collected in each frame are arranged according to the position of the contact area between the die cutting roller and the flexible circuit board to form a multi-frame pressure snapshot of the flexible circuit board. All the pressure snapshots are stacked according to the collection time order to generate a spatiotemporal cubic pressure body. Let i be the sequence number of the pressure snapshot, P(x, y, ti) represent the pressure snapshot of the flexible circuit board acquired at time i, where ti represents the collection time of the sensor array, and P(xj, yk, ti) represents the pressure value of the j-th row and k-th column in the pressure snapshot of the flexible circuit board acquired at time i.

3. The high-precision manufacturing method for a flexible die-cut circuit board according to claim 1, characterized in that, In S200, the method for calculating the jitter characteristics on the time axis based on the spatiotemporal cubic pressure volume is as follows: Within the range of values ​​for i, the pressure values ​​included in the pressure snapshots collected by the flexible circuit board at each acquisition time are iterated. The average pressure value u(t) of the pressure snapshot at acquisition time t is calculated according to the formula, and the pressure distribution value of the pressure snapshot at acquisition time t is calculated according to the formula. (t); The pressure values ​​of each point in the pressure snapshots collected at each acquisition time in the spatiotemporal cubic pressure volume are traversed according to the acquisition time sequence. The pressure values ​​corresponding to each point are arranged into a time axis sequence according to the acquisition time sequence. The maximum pressure value is marked as max_P(x,y) and the minimum pressure value is marked as min_P(x,y) on the time axis. The time interval between the maximum and minimum pressure values ​​is calculated as TPi, which serves as the limit window of the pressure snapshot collected at acquisition time i. The limit windows of all pressure snapshots in the spatiotemporal cubic pressure volume are obtained. The minimum value of all limit windows is selected as the detection window. The average pressure value and pressure distribution value of all pressure snapshots within the window are traversed in the pressure cube with a detection window length w. The spatial jitter coefficient Scff within the detection window is calculated according to the formula. In the spatiotemporal cubic pressure volume, the recording time of all pressure values ​​in any column of the sensors in the first pressure snapshot is obtained. The time difference from the first pressure value to the last pressure value is calculated as Δt using a cross-correlation algorithm. The distribution distance between any two adjacent sensors in the same column of the sensor array is obtained as Δx. The pressure feedback coefficient T_(1) of the first pressure snapshot is used as the pressure feedback coefficient T_(t) of all pressure snapshots within the range of t. The time jitter coefficient Tcff within the detection window is calculated according to the formula.

4. The high-precision manufacturing method for a flexible die-cut circuit board according to claim 1, characterized in that, In S300, the method for rejecting abnormal flexible circuit boards based on jitter characteristics and generating an anomaly report is as follows: The spatial jitter coefficients Scff and Tcff calculated from all acquisition windows are used to construct the spatial sequence SC and the time sequence TC, respectively. On the spatial sequence, the first spatial jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time as the adjacent coefficient. The acquisition time of the adjacent coefficient and the time length of the most recent acquisition time within 24 hours are defined as Time1. The spatial sequence is divided into multiple time periods according to the length of Time1. Any time period is selected as the target time period Period. The average value of all spatial jitter coefficients in the target time period is calculated as the reference coefficient of the target time period. The absolute value of the difference between all spatial jitter coefficients in the target time period and the reference coefficient is calculated in turn. The maximum value of the absolute value of the difference is taken as the spatial offset O of the target coefficient. Let the set of target time periods be {Period1, Period2, ..., Periodr}, and the spatial offset of each target time period be denoted as {O1, O2, ..., Or}, where r represents the index of the target time period; traverse the spatial offsets of all target time periods in chronological order, and starting from the second target time period reaching the spatial offset, subtract the spatial offset of the current target time period from the spatial offset of the next target time period in turn, and take the ratio of the absolute value of the difference result to the absolute value of the sum of squares from the first spatial offset to the current spatial offset as the spatial offset difference of the current target. Starting from the second target time period, iterate through the spatial offset differences of all target time periods in the set. If the spatial offset difference of the current target time period is greater than the spatial offset difference of the previous target time period, and also greater than the spatial offset difference of the next target time period, and greater than the average spatial offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level I. In the time series, the time jitter coefficient with the smallest difference from the most recent acquisition time within 24 hours is searched backwards from the most recent acquisition time and taken as the adjacent coefficient. The time length between the acquisition time of the adjacent coefficient and the most recent acquisition time within 24 hours is defined as Time2. The time series is divided into multiple time periods according to the length of Time2, and the mean time jitter coefficient in each time period is calculated in turn. The standard deviation of the jitter coefficient is calculated as u1 based on all jitter coefficients in each time period; the absolute value of the difference between adjacent jitter coefficients in each time period is calculated as the neighbor difference value, and the standard deviation of all neighbor difference values ​​in each time period is calculated as u2. Each time period contains a jitter coefficient mean and multiple neighbor difference values ​​Diffs, where s represents the sequence number of the neighbor difference value and Diffs represents the s-th neighbor difference value. The time offset difference of each time period is calculated using the formula. Starting from the second target time period, iterate through the time offset differences of all target time periods in the set. If the time offset difference of the current target time period is greater than the time offset difference of the previous target time period, and also greater than the time offset difference of the next target time period, and greater than the average time offset difference of all target time periods, mark the current target time period as an abnormal time period. All circuit boards produced in the abnormal time period are abnormal circuit boards. Remove all abnormal circuit boards and mark them as Level II. Traverse all abnormal circuit board markers, count the number of circuit boards marked as Level I, and when the number of Level I circuit boards exceeds the preset threshold, the console issues a Level I warning, quickly locates the abnormal period, and generates an abnormal feedback report. The system counts the number of circuit boards marked as Level II. When the number of Level II circuit boards exceeds a preset threshold, the control console issues a Level II warning, quickly locates the abnormal period, and generates an anomaly feedback report. All anomaly reports are sent to the control terminal, and staff take corresponding maintenance measures based on the anomaly reports.

Images