Electronic device manufacturing system

Abstract

The invention discloses a manufacturing system of an electronic device. The manufacturing system comprises manufacturing equipment and a monitoring system, the manufacturing equipment is used for providing data strings. The monitoring system is used for receiving a data string and calculating a characteristic value to define a data string model. Wherein the manufacturing equipment receives the data string model to monitor the manufacturing program.

Description

Technical Field

[0001] The present application relates to a manufacturing system, and more particularly, to a manufacturing system for an electronic device. Background Art

[0002] As the workload of manufacturing equipment increases, the management requirements for manufacturing equipment also increase day by day, and various automated monitoring requirements gradually increase. For example, process parameters (such as temperature, pressure, etc.) of manufacturing equipment need to be automatically detected or managed.

[0003] However, the data strings of these process parameters (such as data distribution) often show an asymmetric normal distribution. The data of these process parameters are not suitable for being processed by the statistical methods of traditional normal assumptions. For example, it is difficult to set the control boundaries for normal state and abnormal state for these data. When the control boundaries are set too strictly, it will lead to too many misjudged abnormal states, and when the control boundaries are set too loosely, there will be a risk of missing abnormal states. Therefore, data strings of various data types must use specifically corresponding processing methods (such as corresponding control boundaries). Currently, only professionals can judge the data types of data strings. However, human power is unable to handle a huge amount of data. Therefore, how to automatically distinguish the data types of data strings is a problem that needs to be solved currently.

[0004] Therefore, it is necessary to provide a novel manufacturing system to improve the above problems. Summary of the Invention

[0005] The present application provides a manufacturing system for an electronic device, including: a manufacturing device and a monitoring system. The manufacturing device is used to provide a data string. The monitoring system is used to receive the data string and calculate a characteristic value to define a data string model. Among them, the manufacturing device receives the data string model to monitor the manufacturing process.

[0006] From the following detailed description and in combination with the accompanying drawings, other technical features of the present application will become clearer. Brief Description of the Drawings

[0007] Figure 1 A step flowchart showing the management method of the manufacturing system according to an embodiment of the present application.

[0008] Figure 2 A system architecture diagram showing the manufacturing system according to an embodiment of the present application.

[0009] Figure 3A A schematic diagram showing the data type of the data string according to an embodiment of the present application.

[0010] Figure 3B A schematic diagram showing the data type of the data string according to another embodiment of the present application.

[0011] Figure 4A Schematic diagram of data string corresponding to the detection program for the data value change points in the first form of the present application.

[0012] Figure 4B Flowchart showing the steps of the detection program for the data value change points in the first form of an embodiment of the present application.

[0013] Figure 4C Flowchart showing the steps of the detection program for the data value change points in the first form of another embodiment of the present application.

[0014] Figure 5A Schematic diagram of data string corresponding to the detection program for the data value change points in the second form of the present application.

[0015] Figure 5B Flowchart showing the steps of the detection program for the data value change points in the second form of an embodiment of the present application.

[0016] Figure 5C Flowchart showing the steps of the detection program for the data value change points in the second form of another embodiment of the present application.

[0017] Figure 6 Schematic diagram of the training process of the data type classification unit in an embodiment of the present application.

[0018] Figure 7 Schematic diagram of the actual operation of the change point detection unit, data string segmentation unit, eigenvalue calculation unit, and data type classification unit in an embodiment of the present application.

[0019] Reference numerals: Steps S1 to S6, S21 to S27, S61 to S66, S71 to S76

[0020] Manufacturing system 1000

[0021] Manufacturing equipment 200

[0022] Monitoring system 300

[0023] Detection unit 100

[0024] Change point detection unit 10

[0025] Data string segmentation unit 20

[0026] Eigenvalue calculation unit 30

[0027] Data type classification unit 40

[0028] Machine learning module 41

[0029] Monitoring end storage device 50

[0030] Data string model database 60

[0031] Manufacturing end storage device 210

[0032] User side 230 Specific implementation manners

[0033] In the following description, different embodiments of the present application are provided, which are used to explain the technical content of the present application and not to limit the protection scope of the present application. The features described in one embodiment can be applied to other embodiments through appropriate modification, replacement, combination or separation.

[0034] It should be noted that in this specification, when an element is described as "including", "having", or "containing" an element, it means that the element can include one or more elements, and the element can simultaneously include other elements, and it does not mean that the element only has one of the elements, unless otherwise specified.

[0035] Moreover, in this specification, ordinal numbers such as "first" or "second" are only used to distinguish multiple elements with the same name, and do not mean that there is an essential class, rank, execution order, or manufacturing order between the elements, unless otherwise specified. The serial numbers of the elements in the specification may be different from those in the claims. For example, the "second" element in the specification may be the "first" element in the claims.

[0036] In this specification, unless otherwise specified, feature A "or" or "and / or" feature B means the existence of only feature A, the existence of only feature B, or the existence of both feature A and feature B. Feature A "and" feature B means the existence of both feature A and B.

[0037] In addition, in this specification, terms such as "top", "upper", "bottom", "front", "rear", or "middle", and terms such as "above", "on the top", "on the upper side", "below", "underneath", or "between" are used to describe the relative positions between multiple elements, and the described relative positions can be interpreted to include their translation, rotation, or reflection.

[0038] In addition, the terms described in the specification and claims, such as "above", "on the top", "on the upper side", "below", or "underneath", are intended to mean that an element can not only directly contact other elements, but also indirectly contact another element.

[0039] In addition, terms recited in the specification and claims, such as "connected", mean that an element can be connected to other elements not only directly but also indirectly. On the other hand, terms such as "electrically connected" and "coupled" recited in the specification and claims mean that an element can be electrically connected to other elements not only directly but also indirectly.

[0040] In the specification and claims, the terms "about", "approximately", "substantially", and "substantially the same" generally mean within 10%, or 5%, or 3%, or 2%, or 1%, or 0.5% of a given value. The given quantity is an approximate quantity, that is, the meanings of "about", "approximately", "substantially", and "substantially the same" can still be implied even without specific recitation of "about", "approximately", "substantially", or "substantially the same". In addition, the terms "ranging from a first value to a second value" and "ranging between a first value and a second value" mean that the range includes the first value, the second value, and other values therebetween.

[0041] In this specification, unless otherwise specified, the terms used herein (including technical and scientific terms) have the meanings commonly known to those skilled in the art. It should be noted that, unless otherwise specified in the embodiments of the present application, these terms (such as those defined in a general dictionary) should have the same meaning as those skilled in the art, the background art of the present application, or the context of this specification, and should not be construed in an ideal or overly formal manner.

[0042] The management method of the manufacturing system of the present application can be used to manage a manufacturing system for manufacturing electronic devices and / or electronic components. The electronic device can include a display device, an imaging device, an assembling device, a backlight device, an antenna device, a splicing device, a touch display, a curved display, or a free shape display, but is not limited thereto. The electronic device can also include electronic components. The electronic components can include passive components and active components, such as capacitors, resistors, inductors, diodes, transistors, etc. The diodes can include light-emitting diodes or photodiodes. The light-emitting diodes can include, for example, organic light-emitting diodes (OLEDs), mini light-emitting diodes (mini LEDs), micro light-emitting diodes (micro LEDs), or quantum dot light-emitting diodes (quantum dot LEDs), but are not limited thereto.

[0043] Figure 1is a flowchart showing the steps of a method for managing a manufacturing system of an electronic device according to an embodiment of the present application. The method for managing the manufacturing system can be executed by a detection unit 100 (shown in Figure 2 ). First, step S1 is executed to receive a data string, where the data string contains multiple data values, such as a process parameter at different time points when manufacturing an electronic device by a manufacturing device (such as a machine tool), such as process temperature, flow rate of the gas introduced, pressure, concentration of the fluid introduced, pH value, etc., but not limited thereto. Then, step S2 is executed to perform a detection program for data value change points to determine whether the data string has at least one data value change point. When the data string has at least one data value change point, step S3 is executed to divide the data string into at least one first section and a second section according to the at least one data value change point. Then, step S4 is executed to calculate at least one characteristic value of the data values in the first section, where the first section has a first number of data values, the second section has a second number of data values, and the first number is greater than the second number, that is, the first section has a larger number of data values. When the data string does not have a data value change point, step S5 is executed to calculate at least one characteristic value of all the data values in the data string. Then, step S6 is executed to define the data type of the data string according to the at least one characteristic value.

[0044] In an embodiment, the method for managing the manufacturing system can be used to detect and determine the data type of the data string, so as to facilitate subsequent monitoring of the manufacturing process and / or process parameters of the manufacturing device. In an embodiment, the data string can be regarded as a collection of data values within a period, such as a collection of process parameters of a manufacturing device during a working period, and not limited thereto. The data string can be defined as different data types according to its data distribution, and various data types can be regarded as various data string models. In addition, in an embodiment, the data string may have multiple data value change points, so the data string may be divided into more than two sections. At this time, the detection unit 100 will find a section with the largest number of data values among these sections, calculate at least one characteristic value of this section, and use the at least one characteristic value of this section to define the data type (i.e., the data string model) of the data string.

[0045] Figure 2 is a system architecture diagram of a manufacturing system 1000 according to an embodiment of the present application. Please also refer to Figure 1 . As shown in Figure 2As shown, the manufacturing system 1000 may include, for example, a manufacturing device 200 and a monitoring system 300. The monitoring system 300 may include a detection unit 100, and the detection unit 100 may include, for example, a change point detection unit 10, a data string segmentation unit 20, a feature value calculation unit 30, and a data type classification unit 40. The monitoring system 300 may obtain a data string from the manufacturing device 200 through wired or wireless transmission via various suitable signal transmission interfaces (such as step S1) for providing it to the detection unit 100 for calculation and / or analysis. For example, the manufacturing device 200 may transmit the process parameters during the operation of the manufacturing process to a manufacturing-side storage device 210, and the manufacturing-side storage device 210 may communicate with a monitoring-side storage device 50 of the monitoring system 300 to transmit the process parameters to the monitoring-side storage device 50, but not limited thereto. The change point detection unit 10 may execute a detection program for data value change points to detect whether the data string has data value change points (such as step S2). The data string segmentation unit 20 may segment the data string into multiple sections according to the data value change points (such as step S3). The feature value calculation unit 30 may select one of the segmented sections and calculate the feature value of this section (such as step S4), or when the data string does not have data value change points, the feature value calculation unit 30 may calculate the feature value of the entire data string (such as step S5). The data type classification unit 40 may be used to compare the feature value with the feature value of the model to define the data type of the data string (such as step S6, i.e., the data string model). For example, the monitoring system 300 may include a data string model database 60, and the data string model database 60 stores data of various data string models for the detection unit 100 to compare and define what kind of data string model the data string provided by the manufacturing device 200 is. Then, the detection unit 100 may transmit the data type of the data string (i.e., the data string model) to the manufacturing device 200 through the monitoring system 300, and the manufacturing device 200 may perform corresponding processing according to the data string model, such as monitoring the manufacturing process performed by the manufacturing device 200 in a monitoring manner that conforms to a specific data string model. In addition, the detection unit 100 may also transmit the data type of the data string to the client 230, such as transmitting it to a display or to a communication device of the client 230, but not limited thereto.

[0046] In one embodiment, the data type classification unit 40 may include a machine learning model 41. After being trained, the machine learning model 41 can have the ability to define the data type of a data string. For example, the trained machine learning model 41 can generate an analysis result based on feature values and compare the analysis result with the data string models in the data string model database 60 to find the closest data string model, but not limited thereto. More specifically, the machine learning model 41 can have a "training stage" and an "actual operation stage" after training is completed. In addition, the operation result of the machine learning model 41 in the "actual operation stage" can also be fed back to the machine learning model 41 itself, so that the ability of the machine learning model 41 can be continuously improved. In one embodiment, the type of the machine learning model 41 may include a convolutional neural network (CNN), a deep neural network (DNN), a recurrent neural network (RNN), a random forest, a decision tree, or a model with similar functions, etc., and not limited thereto. In one embodiment, the machine learning model 41 in the training stage can be trained with a large amount of training data, and the number of training data can be, for example, at least 150, but not limited thereto.

[0047] In one embodiment, the detection unit 100 may be, for example, an entity device with a processor, or it may also be the processor itself, or it may also be a computer program product executed by the processor. This application is not limited thereto. In one embodiment, the change point detection unit 10, the data string segmentation unit 20, the feature value calculation unit 30, and the data type classification unit 40 may be, for example, functional modules, which can be implemented by the processor executing the instructions of the computer program product. The computer program product can be stored in a non-transitory computer-readable medium and may include one or more instructions for causing the processor to perform special operations. The non-transitory computer-readable medium may be, for example, a storage element, such as a memory, a hard disk, a USB flash drive, or a cloud drive, etc., and not limited thereto. In one embodiment, the processor and the storage element may be, for example, arranged in a computing electronic device, and the computing electronic device may be, for example, but not limited to, a desktop computer, a laptop computer, a smart mobile device, a server, or a cloud host, etc. In addition, the change point detection unit 10, the data string segmentation unit 20, the feature value calculation unit 30, and the data type classification unit 40 can be implemented by the same or different processors and can be arranged on the same or different computing electronic devices.

[0048] To make the purpose of this application easier to understand, the data type of the data string will be described next. Figure 3A It is a schematic diagram of a data string in an embodiment of this application. Figure 3BSchematic diagram of a data string according to another embodiment of the present application. Please also refer to Figure 1 and Figure 2 . In addition, in the graphs shown in Figure 3A and Figure 3B , the horizontal axis is the sequence of data values in the data string, and the vertical axis is the measured value of the data value.

[0049] As Figure 3A shown, during a period, the data string has multiple data values and can be divided into multiple regions. Among them, the measured values of the data values in the first region (the sequence of data values is approximately between 0 and 2000) are approximately between 40000 and 50000, while the measured values of the data values in the second region (the sequence of data values is approximately between 2000 and 4500) are approximately between 5000 and 15000. The average value of the data values in the first region and the average value of the second region have obvious differences. Therefore, it can be regarded as a change in the average values of the first region and the second region. Therefore, there is a data value change point between the first region and the second region. By analogy, it can be known that Figure 3A the data string of

[0050] has changed 6 times during this period, so it has 6 data value change points. Figure 3B shown, during a period, the data string can be divided into multiple regions. The measured values of the data values in the first region (the sequence is approximately between 0 and 100) are approximately between 240 and 320, while the measured values of the data values in the second region (the sequence exceeds 100) are approximately between 240 and 280. The standard deviation of the data values in the first region and the standard deviation of the second region have obvious differences. Therefore, it can also be regarded as a change in the standard deviations of the first region and the second region. Therefore, there is a data value change point between these two regions. By analogy, Figure 3B the data string of

[0051] has changed 1 time during this period and has 1 data value change point. Figure 3A and Figure 3B Because the data values have changed, the data strings of the Figure 3A embodiments cannot be controlled by conventional methods. Among them, Figure 3A the embodiment is similar to the quasi-normal data type with a change in the average value and is not suitable to use the same average value as the control standard; while Figure 3BThe embodiments are similar to the partial single-sided data patterns with varying standard deviations, and it is not suitable to use the same standard deviation value as the control standard. Therefore, the two need to be controlled through corresponding control methods respectively. In one embodiment, in actual use, the data patterns of the data string may include, for example, data approximately following a normal distribution (hereinafter referred to as a data string model with approximately normal data), data highly concentrated near the central value (hereinafter referred to as a data string model with a high and narrow peak), data asymmetrically distributed on both sides of the central value (hereinafter referred to as a partial single-sided data string model), data values in the data string gradually increasing or decreasing over time (hereinafter referred to as a trend-type data string model), or other data patterns, and are not limited thereto. It can be seen that for the purpose of process automation, it is necessary to find the data value change points in the data string and determine the data pattern (i.e., the data string model) of the data string for appropriate control.

[0052] Next, the details of the Figure 1 system management method will be described. Hereinafter, the description will directly start from the detection procedure of the data value change point in step S2. Figure 4A Show the data string schematic diagram corresponding to the detection procedure of the data value change point of the first form of the present application, Figure 4B Show the step flowchart of the detection procedure of the data value change point of the first form of an embodiment of the present application, and please refer to Figures 1 to 3B .

[0053] As Figure 4A shown, by executing the detection procedure of the data value change point, the change point detection unit 10 can detect whether one of the data values in the data string is a data value change point. Among them, the currently detected one of the data values (hereinafter referred to as the detection point) is marked with the symbol b1, the data value of the first sequence before the detection point b1 is marked with the symbol a0 (for convenience of description, hereinafter, "the data value of the nth sequence" will be abbreviated as "the nth data value"), the second data value before the detection point b1 is marked with the symbol a1, the third data value before the detection point b1 is marked with the symbol a2, the first data value after the detection point b1 is marked with the symbol b2, and the second data value after the detection point b1 is marked with the symbol b3. In the process of determining whether the detection point b1 is a data value change point, at least a part of the data values a0 to a3 and b1 to b3 can be used as the basis for judgment, but it is not limited thereto. In addition, Figure 4A Show the situation where the average value of the data string changes.

[0054] The detection procedure of the data value change point (step S2) can be executed by the change point detection unit 10. As Figure 4BAs shown, first, step S21 is executed to set one of the data values in the data string as a detection point b1. Then, step S22 is executed to set a first detection section A and a second detection section B of the data string according to the detection point b1. Next, step S23 is executed to calculate the average value and the standard deviation value of the data values in the first detection section A (hereinafter referred to as the first average value and the first standard deviation value), and to calculate the average value and the standard deviation value of the data values in the second detection section B (hereinafter referred to as the second average value and the second standard deviation value). Then, step S24 is executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 meet a first preset condition. When the first preset condition is not met, step S25 is executed to set the current detection point b1 as not a data value change point, and then step S21 can be restarted and a new detection point b1 can be set. When the first preset condition is met, step S26 is executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 meet a second preset condition. When the second preset condition is not met, step S25 is executed, that is, to set the current detection point b1 as not a data value change point, and then step S21 can be restarted and a new detection point b1 can be set. When the second preset condition is met, step S27 is executed to set the current detection point b1 as a data value change point, and then step S21 is restarted and a new detection point b1 is set, and the judgment is performed in this way in a loop. Accordingly, when the first detection section A and the second detection section B corresponding to the detection point b1 simultaneously meet the first preset condition and the second preset condition, the detection point b1 can be determined and set as a data value change point.

[0055] Regarding step S21, in one embodiment, when the detection program for the data value change point starts to execute, the change point detection unit 10 is set to use the (N + 1)-th data value in the data string as the first detection point b1, and after the current detection point b1 is detected, the subsequent data values are detected in sequence, where N can be between 30 and 70 (30 ≤ N ≤ 70), such as 50, and is not limited thereto. By using the (N + 1)-th data value as the first detection point b1, it can be ensured that the data amounts of the first detection section A and the second detection section B are sufficient for analysis, and is not limited thereto.

[0056] Regarding step S22, in one embodiment, the first detection section A can be set to include N data values before the detection point b1 (excluding the detection point b1), and is not limited thereto. The second detection section B can be set to include the detection point b1 and N data values after the detection point b1 (a total of N + 1 data values), or the second detection section B can be set to include the detection point b1 and N - 1 data values after the detection point b1 (a total of N data values), and is not limited thereto.

[0057] Regarding step S23, in one embodiment, before calculating the first average value, the first standard deviation value, the second average value, and the second standard deviation value, the change point detection unit 10 may first perform an operation of data cleaning on the first detection section A and the second detection section B. For example, outliers (noise) in the first detection section A and the second detection section B are removed. Therefore, the calculated first average value, first standard deviation value, second average value, and second standard deviation value can make subsequent analysis more accurate. However, the operation of data cleaning can be selectively executed according to requirements.

[0058] Regarding step S24, in one embodiment, the first preset condition at least includes: the absolute value of the difference between the detection point b1 and the first data value a0 before it is greater than m times the first standard deviation value (|b1 - a0| > m * the first standard deviation value), and the absolute value of the difference between the first data value b2 after the detection point b1 and the second data value a1 before the detection point b1 is greater than m times the first standard deviation value (|b2 - a1| > m * the first standard deviation value), or the absolute value of the difference between the detection point b1 and the first data value a0 before it is greater than m times the second standard deviation value (|b1 - a0| > m * the second standard deviation value), and the absolute value of the difference between the first data value b2 after the detection point b1 and the second data value a1 before the detection point b1 is greater than m times the second standard deviation value (|b2 - a1| > m * the second standard deviation value), where m is a positive integer greater than 0. In addition, in one embodiment, m can be regarded as the gap parameter of the average value change between the first detection section A and the second detection section B. In one embodiment, m can be preset to 3 and can also be adjusted according to requirements.

[0059] Furthermore, in one embodiment, in addition to the foregoing content (|b1 - a0| > m * the first standard deviation value, and |b2 - a1| > m * the first standard deviation value, or |b1 - a0| > m * the second standard deviation value, and |b2 - a1| > m * the second standard deviation value), the first preset condition may further include: the absolute value of the difference between the second data value b3 after the detection point b1 and the third data value a2 before the detection point b1 is greater than m times the first standard deviation value (|b3 - a2| > m * the first standard deviation value), or the absolute value of the difference between the second data value b3 after the detection point b1 and the third data value a2 before the detection point b1 is greater than m times the first standard deviation value (|b3 - a2| > m * the second standard deviation value), and is not limited thereto. Therefore, the accuracy of judgment can be improved and the processing time will not be too long. In addition, in some other embodiments, the first preset condition may further include more content.

[0060] Regarding step S26, in one embodiment, the second preset condition at least includes: the absolute value of the difference between the second average value and the first average value is greater than m times the first standard deviation value (|second average value - first average value| > m * first standard deviation value), or the absolute value of the difference between the second average value and the first average value is greater than m times the second standard deviation value (|second average value - first average value| > m * second standard deviation value), and is not limited thereto.

[0061] Regarding steps S25 and S27, after the detection of the current detection point b1 is completed, the change point detection unit 10 can perform steps S21 to S27 again for a new detection point. In one embodiment, the new detection point can be the first data value after the current detection point (such as data value b2), and is not limited thereto. In addition, regarding step S27, in one embodiment, after the detection of the first detection point to the last detection point is completed, the change point detection unit 10 may find multiple data value change points in the data string.

[0062] Therefore, the detection unit 100 can automatically determine whether there are data value change points in the data string and can find the data value change points.

[0063] Figure 4C Show the step flowchart of the detection program for the data value change points in the first form of another embodiment of the present application. Figure 4C The detection program for the data value change points can include steps S21 to S25, S26a, and S27. Among them, steps S21 to S25 and S27 can be applicable Figure 4B to the description of the embodiment, so only step S26a will be described below.

[0064] When the first preset condition is satisfied, step S26a can be executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 satisfy a third preset condition.

[0065] In one embodiment, the third preset condition at least includes: after calculation by a specific algorithm, the difference between the first average value and the second average value is less than a threshold value. In one embodiment, the specific algorithm is, for example, the Mann-Whitney two-sample test method, but is not limited thereto. In one embodiment, the threshold value can be between 0.01 and 0.1 (0.01 ≤ threshold value ≤ 0.1), and is not limited thereto. For example, assuming the threshold value is 0.05, if the difference between the first average value and the second average value is less than 0.05 at this time, it means that there is a significant difference between the first average value and the second average value. Therefore, the probability that the detection point is a data value change point is high. Therefore, when the first detection section A and the second detection section B corresponding to the current detection point b1 simultaneously meet the first preset condition and the third preset condition, the change point detection unit 10 can determine and set the current detection point b1 as the data value change point.

[0066] It should be noted that Figure 4B and Figure 4C the detection procedure for the data value change point is suitable for detecting the data string form of the average value change to which the data value change point belongs, but can still be used to detect other forms of data value change points.

[0067] Figure 5A showing a data string schematic diagram corresponding to the detection procedure for the data value change point of the second form of the present application, Figure 5B showing a step flowchart of the detection procedure for the data value change point of the second form in one embodiment of the present application.

[0068] As Figure 5A shown, the current detection point of the detection procedure for the data value change point is marked with the symbol b1, the first data value before the detection point b1 is marked with the symbol a0, the second data value before the detection point b1 is marked with the symbol a1, the third data value before the detection point b1 is marked with the symbol a2, the first data value after the detection point b1 is marked with the symbol b2, and the second data value after the detection point b1 is marked with the symbol b3. The above detection point b1 and the data values a0 to a3, b2 to b3 before and after it can be used in the detection procedure for the data value change point. In addition, Figure 5A shows the situation where the standard deviation value of the data string changes.

[0069] In Figure 5B the embodiment, the detection procedure for the data value change point may include steps S21 to S23, S24b, S25, S26b, and S27, where steps S21 to S23, S25, and S27 are applicable to Figure 4B the description of the embodiment, so the following mainly describes steps S24b and S26b.

[0070] After step S23 is executed, step S24b can be executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 satisfy a fourth preset condition. When the fourth preset condition is not satisfied, step S25 can be executed. When the fourth preset condition is satisfied, step S26b can be executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 satisfy a fifth preset condition.

[0071] Regarding step S24b, in one embodiment, the fourth preset condition at least includes: the absolute value of the difference between the detection point b1 and the first data value a0 before it is greater than w times the first standard deviation value (|b1 - a0| > w * the first standard deviation value), and the absolute value of the difference between the first data value b2 after the detection point b1 and the second data value a1 before the detection point b1 is greater than w times the first standard deviation value (|b2 - a1| > w * the first standard deviation value), or the absolute value of the difference between the detection point b1 and the first data value a0 before it is greater than w times the second standard deviation value (|b1 - a0| > w * the second standard deviation value), and the absolute value of the difference between the first data value b2 after the detection point b1 and the second data value a1 before the detection point b1 is greater than w times the second standard deviation value (|b2 - a1| > w * the second standard deviation value), where w is a positive integer greater than 0. In addition, in one embodiment, w can be regarded as a gap parameter for the variation of the standard deviation values of the first detection section A and the second detection section B. In one embodiment, w can be preset to 3 or can be adjusted according to requirements. Additionally, in one embodiment, w and m are not necessarily the same.

[0072] Furthermore, in one embodiment, in addition to the foregoing (|b1 - a0| > w * the first standard deviation value, and |b2 - a1| > w * the first standard deviation value, or |b1 - a0| > w * the second standard deviation value, and |b2 - a1| > w * the second standard deviation value), the fourth preset condition may further include: the absolute value of the difference between the second data value b3 after the detection point b1 and the third data value a2 before the detection point b1 is greater than w times the first standard deviation value (|b3 - a2| > w * the first standard deviation value), or the absolute value of the difference between the second data value b3 after the detection point b1 and the third data value a2 before the detection point b1 is greater than w times the first standard deviation value (|b3 - a2| > w * the second standard deviation value), and is not limited thereto. Therefore, the accuracy of the judgment can be improved, and the processing time will not be too long. In addition, in some other embodiments, the fourth preset condition may further include more content.

[0073] Regarding step S26b, in one embodiment, the fifth preset condition at least includes: the second standard deviation value is greater than w times the first standard deviation value, or the first standard deviation value is greater than w times the second standard deviation value. When the first detection section A and the second detection section B corresponding to the current detection point b1 simultaneously satisfy the fourth preset condition and the fifth preset condition, the current detection point b1 can be determined and set as a data value change point.

[0074] Figure 5C The flowchart of the steps of the detection program for the data value change point in the second form of another embodiment of the present application is shown. In Figure 5C one embodiment, the detection program for the data value change point may include steps S21 to S23, S24b, S25, S26c, and S27. Among them, steps S21 to S23, S24b, S25, and S27 are applicable Figure 5B to the description of the embodiment, so only step S26c will be described below.

[0075] When the first preset condition is satisfied, step S26c can be executed to determine whether the first detection section A and the second detection section B corresponding to the detection point b1 satisfy a sixth preset condition.

[0076] In one embodiment, the sixth preset condition at least includes: after being calculated by a specific algorithm, the difference between the first standard deviation value and the second standard deviation value is less than a threshold value (|first standard deviation value - second standard deviation value| < threshold value). In one embodiment, the specific algorithm is, for example, the Mann-Whitney two-sample test method, but is not limited thereto. In one embodiment, the threshold value can be between 0.01 and 0.1 (0.01 ≤ threshold value ≤ 0.1), and is not limited thereto. For example, assuming the threshold value is 0.05, if the difference between the first standard deviation value and the second standard deviation value is less than 0.05 at this time, it means that there is a significant difference between the first standard deviation value and the second standard deviation value, so the probability that the detection point is a data value change point is high. Therefore, when the first detection section A and the second detection section B corresponding to the current detection point b1 simultaneously satisfy the fifth preset condition and the sixth preset condition, the current detection point b1 can be determined and set as a data value change point.

[0077] Figure 5B and Figure 5C The detection program for the data value change point of the embodiment is suitable for detecting the data string form of the average value change to which the data value change point belongs, but can still be used to detect other forms of data value change points.

[0078] Therefore, the detection program for the data value change point (step S2) can already be understood.

[0079] Next, regarding Figure 1Step S3 will be described. In one embodiment, when the data string has multiple data value change points, the data string splitting unit 20 can split the data string into multiple segments according to the multiple data value change points, so as to Figure 3A Taking the data string of Figure 3A as an example, it has 6 data value change points. Therefore, the data string splitting unit 20 can split the data string into 7 segments according to these data value change points. In addition, when the data string does not have data value change points, the data string splitting unit 20 may not operate.

[0080] Next, Figure 1 Step S4 will be described. In one embodiment, after the data string is split into multiple segments, the eigenvalue calculation unit 30 can compare the number of data values in each segment, and select the segment with the largest number of data values, and then calculate the eigenvalue of this segment. It should be noted that in one embodiment, since in step S2, the first detection point b1 is the (N + 1)-th data value in the data string, however, there may also be undetected data value change points among the data values before the first detection point b1. To avoid these undetected data value change points from affecting the calculation of the eigenvalue, the eigenvalue calculation unit 30 can exclude the first N data values and the last N data values in the data string from the calculation of the eigenvalue. For example, when the segment selected by the eigenvalue calculation unit 30 contains the first N data values and the last N data values in the data string, the eigenvalue calculation unit 30 can delete the first N data values and the last N data values in the data string, and will not use these data values for calculating the eigenvalue, but it is not limited thereto.

[0081] In one embodiment, the eigenvalue may include the slope, skewness, kurtosis, or normal distribution value of the data values in the selected segment, or any combination of these, and it is not limited thereto. In one embodiment, the normal distribution value can be obtained through the Jarque - Bera Test (J - B Test).

[0082] In one embodiment, when the eigenvalue is the slope, it can be presented by the following formula:

[0083]

[0084] where xi is the order of the data values in the selected segment (starting from 1 and up to n), yi is the value of the data value corresponding to xi in the selected segment, n is the number of data values in the selected segment, x(bar) is the average value of xi, and y(bar) is the average value of yi.

[0085] In one embodiment, when the eigenvalue is the skewness, it can be presented by the following formula:

[0086]

[0087] In one embodiment, when the eigenvalue is kurtosis, it can be presented by the following formula:

[0088]

[0089] In one embodiment, when the eigenvalue is a normal distribution value, it can be presented by the following formula (Jarque - Bera test):

[0090]

[0091] where S is skewness and K is kurtosis.

[0092] Therefore, the eigenvalue calculation unit 30 can calculate the eigenvalue of the selected section and use this eigenvalue as the eigenvalue of the data string to be judged for subsequent processing.

[0093] Next, step S5 will be described. In one embodiment, when there is no data value change point in the data string, the eigenvalue calculation unit 30 can calculate the eigenvalues of all data values in the data string, and this is not limited thereto.

[0094] Next, step S6 will be described. In one embodiment, after the eigenvalue calculation unit 30 calculates the eigenvalue of a data string, the eigenvalue of this data string can be input into the trained machine learning model 41 of the data type classification unit 40. The machine learning model 41 can analyze the eigenvalue of this data string and define the data type of this data string. However, in order for the machine learning model 41 to have the above - mentioned analysis ability, the machine learning model 41 must first be trained with a large amount of training data. Next, the training process of the machine learning model 41 will be described.

[0095] Figure 6 A flowchart showing the training process of the data type classification unit 40 according to an embodiment of the present application is shown, and please also refer to Figures 1 to 5C . Among them, steps S61 to S64 can be regarded as pre - processing steps and can be executed by components other than the data type classification unit 40, while steps S65 and S66 are executed by the data type classification unit 40.

[0096] As Figure 6 shown, first, step S61 is executed. A large amount of training data is input into the detection unit 100. Each piece of training data can represent a data string, and each piece of training data has a label, and the label marks the data type of each piece of training data (i.e., the data string model). In one embodiment, the data type of each piece of training data can be marked manually, but this is not limited thereto.

[0097] Next, step S62 is executed. The change point detection unit 10 can detect the data value change points in these training data. This step is the same asFigure 1 Step S2, so it will not be elaborated further.

[0098] Next, step S63 is executed. The data string splitting unit 20 can segment the training data according to the data value change points of the training data. This step is the same as Figure 1 Step S3, so it will not be elaborated further.

[0099] Next, step S64 is executed. For each training data, the eigenvalue calculation unit 30 can select the segment with the most data values for calculation to calculate one or more eigenvalues of each training data. In addition, if a training data does not have a data value change point, the eigenvalue calculation unit 30 will directly calculate one or more eigenvalues of all the data values of this training data. This step is the same as Figure 1 Step S4 or S5, so it will not be elaborated further.

[0100] Next, step S65 is executed. The detection unit 100 can map the eigenvalues of the training data to the labels of the training data (for example, through an instruction input by the user or according to the default instruction), and input the training data with the corresponding label into the machine learning model 41 in the training stage to enable the machine learning model 41 to be trained.

[0101] Next, step S66 is executed. After the machine learning model 41 is trained, the machine learning model 41 can have the ability to analyze the eigenvalues of the data string and enter the actual operation stage. After that, the unknown data string and its eigenvalues can be input into the machine learning model 41 of the data type classification unit 40, and the machine learning model 41 of the data type classification unit 40 can automatically define the data type of the unknown data string (i.e., the data string model).

[0102] Next, the process of the actual operation stage of the machine learning model 41 of the data type classification unit 40 will be described. Figure 7 Show the schematic flow diagram of the actual operation of the change point detection unit 10, data string splitting unit 20, eigenvalue calculation unit 30, and data type classification unit 40 in an embodiment of the present application, and please refer to Figures 1 to 6 .

[0103] As Figure 7 shown, first, step S71 is executed, and a data string of unknown type is input into the detection unit 100.

[0104] Next, step S72 is executed. The change point detection unit 10 can detect the data value change points in the data string of the unknown type. This step is the same as Figure 1 Step S2.

[0105] Next, step S73 is executed. The data string segmentation unit 20 can segment the data string of the unknown type according to the data value change points. This step is like Figure 1 step S3 in

[0106] Next, step S74 is executed. The eigenvalue calculation unit 30 can select the section with the most data values for calculation to calculate one or more eigenvalues of the data string of the unknown type. In addition, if the data string of the unknown type does not have data value change points, the eigenvalue calculation unit 30 directly calculates one or more eigenvalues of all data values of the data string of the unknown type. This step is like Figure 1 step S4 or S5 in

[0107] Next, step S75 is executed. The detection unit 100 (such as the data type classification unit 40) can analyze the eigenvalues found in step S74 to define the data type (i.e., the data string model) of the data string of the unknown type.

[0108] Next, step S76 is executed. The manufacturing device 300 can receive the data type (i.e., the data string model) of the data string and perform corresponding control methods on the data string.

[0109] Therefore, the actual operation stage of the machine learning model 41 can be understood.

[0110] From the above description, it can be seen that the system management method executed by the detection system of the present application can automatically detect the data value change points of the data string and can automatically analyze the data type of the data string, which can greatly reduce the labor cost or shorten the time cost, or can improve the accuracy of analysis, or can improve the management efficiency.

[0111] The present application can at least determine whether it falls within the protection scope of the present application through the operation mode of the disputed product, or can determine whether it falls within the protection scope of the present application through the algorithm of the disputed product, and is not limited thereto. In an embodiment, the algorithm of the disputed product can be obtained, for example, through reverse engineering, and is not limited thereto.

[0112] The features between the embodiments of the present application can be arbitrarily combined and used as long as they do not violate the inventive spirit or conflict with each other.

[0113] The above embodiments are only examples for convenience of description. The scope of rights claimed in the present application should be subject to what is described in the claims, rather than being limited to the above embodiments.

Claims

1. A manufacturing system for an electronic device, characterized in that, Comprising: A manufacturing device for providing a data string; and A monitoring system for receiving the data string and calculating a characteristic value to define a data string model; Wherein, the manufacturing device receives the data string model to monitor a manufacturing process.

2. The manufacturing system according to claim 1, wherein, The data string includes a plurality of data values, and the monitoring system divides the data string into a first section and a second section according to a data value change point among the data values.

3. The manufacturing system according to claim 2, wherein: The first section includes a first quantity of data values, the second section includes a second quantity of data values, and the first quantity is greater than the second quantity.

4. The manufacturing system according to claim 3, characterized in that, The monitoring system calculates the data values of the first section to obtain the characteristic value.

5. The manufacturing system according to claim 1, characterized in that, The characteristic value includes a slope value.

6. The manufacturing system according to claim 1, characterized in that, The characteristic value includes a skewness value.

7. The manufacturing system according to claim 1, characterized in that, The characteristic value includes a kurtosis value.

8. The manufacturing system according to claim 1, wherein The characteristic value includes a normal distribution value.

9. The manufacturing system according to claim 8, wherein, The normal distribution value is obtained through Jarque - Bera test.

10. The manufacturing system according to claim 1, wherein, The monitoring system further includes a detection unit for performing a detection procedure of a data value change point, and the detection procedure further includes the steps of: Setting one of the data values of the data string as a detection point; Setting a first detection section and a second detection section of the data string according to the detection point; Calculating a first average value and a first standard deviation value of the data values in the first detection section, and calculating a second average value and a second standard deviation value of the data values in the second detection section; Judging whether the first detection section and the second detection section corresponding to the detection point satisfy a first preset condition; When the first preset condition is satisfied, judging whether the first detection section and the second detection section corresponding to the detection point satisfy a second preset condition; When the second preset condition is satisfied, setting the detection point as the data value change point.

Images