Method for controlling a production process and production process system

Abstract

The embodiment of the present disclosure provides a product production process control method and a product production process system, wherein the product production process control method comprises the following steps: acquiring a plurality of different dimension process information corresponding to each process station in one or more process stations within a preset time period; generating an influence weight matrix according to the plurality of different dimension process information; the influence weight matrix comprises an importance degree corresponding to each correction operation to be performed by the product when the plurality of different dimension process information meets different preset conditions; generating an operation mapping table according to the influence weight matrix; the operation mapping table comprises a correction operation with an importance degree exceeding a preset degree and an operation process corresponding to the correction operation with the importance degree exceeding the preset degree; and the operation mapping table is used for guiding the product to perform the corresponding correction operation.

Description

Technical Field

[0001] This disclosure relates to the field of semiconductor technology, and in particular to a product manufacturing process control method and product manufacturing process system. Background Technology

[0002] With the significant improvement in the technological level and manufacturing processes of the semiconductor manufacturing industry, the production scale of the semiconductor industry has also gradually expanded. In the semiconductor manufacturing process, wafers must go through a series of hundreds or even thousands of processes to become the final product. Different processes are carried out on different machines, and one or more machines that complete the same process form a machine group. Only after a wafer completes the current process will it be dispatched to the machine to perform the next process according to the process flow.

[0003] The ability to accurately and in real-time determine the production process status of each batch of wafers between the current and next production processes, and to dynamically allocate the optimal control logic to each batch of wafers, plays a crucial role in improving the production efficiency of each batch of wafers. Summary of the Invention

[0004] In view of the above, this disclosure provides a product manufacturing process control method and a product manufacturing process system to solve at least one technical problem existing in the prior art.

[0005] According to a first aspect of the present disclosure, a method for controlling a product manufacturing process is provided, comprising:

[0006] Obtain multiple different dimensions of process information for the product at one or more process stations within a preset time period;

[0007] Based on the process information of multiple different dimensions, an influence weight matrix is ​​generated; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions.

[0008] An operation mapping table is generated based on the influence weight matrix; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes for correction operations whose importance exceeds a preset level; the operation mapping table is used to guide the product to perform the corresponding correction operations.

[0009] In the above scheme, an influence weight matrix is ​​generated based on the process information of multiple different dimensions, including:

[0010] A criterion matrix is ​​generated based on the process information from multiple different dimensions; the criterion matrix includes the relationship between the process information satisfying different preset conditions and the various correction operations to be performed; and

[0011] The influence weight matrix is ​​generated based on the criterion matrix.

[0012] In the above scheme, a criterion matrix is ​​generated based on the process information of multiple different dimensions, including:

[0013] Obtain the threshold values ​​of different preset conditions in the multiple different dimensions of process information;

[0014] At the end of the preset time period, process information that exceeds the corresponding threshold is filtered out from the multiple different dimensions of process information;

[0015] Based on the process information that exceeds the corresponding threshold, determine the corrective operations to be performed on the product.

[0016] In the above scheme, generating the influence weight matrix based on the criterion matrix includes:

[0017] Based on the criterion matrix, the scale of each correction operation to be performed is calculated, whereby the scale represents the priority and trigger count of the correction operation.

[0018] A judgment matrix is ​​formed based on the scale of each correction operation to be performed. The judgment matrix includes a comparison between any two scales among the multiple scales corresponding to each correction operation to be performed.

[0019] The influence weight matrix is ​​generated based on the judgment matrix.

[0020] In the above scheme, generating the influence weight matrix based on the judgment matrix includes:

[0021] The column vectors of the judgment matrix are normalized.

[0022] The average value of the row vectors of the judgment matrix after normalization is calculated to obtain the influence weight matrix.

[0023] In the above scheme,

[0024] The acquisition of multiple different dimensions of process information corresponding to each of the multiple process stations within a preset time period includes:

[0025] When the product travels to the start station, the timing is started. The start station is the first process station among the plurality of process stations. When the product travels to the end station, the timing is stopped. The end station is the process station after the start station among the plurality of process stations.

[0026] The duration is determined when the product travels to any process station after the starting station, or when the product completes each process station after the starting station.

[0027] If the timing duration has not reached the preset duration, continue to acquire process information of multiple different dimensions corresponding to each process station;

[0028] The step of generating an influence weight matrix based on the process information from multiple different dimensions includes:

[0029] When the timing duration reaches or exceeds the preset duration, the influence weight matrix is ​​generated based on the acquired process information from multiple different dimensions.

[0030] In the above scheme, after generating the operation mapping table, the method further includes:

[0031] The location of the product is determined to determine whether it has reached the target site; the target site is the last of the plurality of process sites.

[0032] If the product does not reach the target site, continue to acquire process information of multiple different dimensions corresponding to each operation process;

[0033] When the product arrives at the target site, the corresponding correction operation of the target site is executed according to the information in the operation mapping table, or the product production process of the target site is controlled normally.

[0034] In the above scheme, performing corresponding correction operations or normally controlling the product manufacturing process of the target site according to the information in the operation mapping table includes:

[0035] When the duration of the timekeeping is greater than or equal to the preset duration, the corresponding correction operation of the target station is performed according to the information in the operation mapping table;

[0036] When the duration of the timing is less than the preset duration, the product manufacturing process at the target site is controlled normally.

[0037] In the above scheme, generating an operation mapping table based on the influence weight matrix includes:

[0038] Obtain the first correction operation corresponding to the maximum importance level and the first operation process corresponding to the correction operation corresponding to the maximum importance level.

[0039] The operation mapping table is generated based on the first correction operation and the first operation process.

[0040] According to a second aspect of the present disclosure, a product manufacturing process system is provided, the product manufacturing process system being configured to perform the following operations:

[0041] Obtain multiple different dimensions of process information for the product at one or more process stations within a preset time period;

[0042] Based on the process information of multiple different dimensions, an influence weight matrix is ​​generated; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions.

[0043] An operation mapping table is generated based on the influence weight matrix; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes for correction operations whose importance exceeds a preset level; the operation mapping table is used to guide the product to perform the corresponding correction operations.

[0044] In the various embodiments of this disclosure, by recording multiple dimensions of process information of the product at each process station within a preset time period, an influence weight matrix is ​​generated based on the multiple dimensions of process information; and an operation mapping table is generated based on the influence weight matrix to guide the product to perform corresponding correction operations when the product enters the subsequent related process station. In this way, the production process status of each batch of products can be accurately and in real time determined between the current process station and the next process station, and the optimal control logic can be dynamically allocated to the production process of each batch of products, thereby improving the production efficiency of the products. Attached Figure Description

[0045] Figure 1 A flowchart illustrating a method for controlling a product manufacturing process, as provided in this disclosure.

[0046] Figure 2 A flowchart is provided for another product manufacturing process control method according to embodiments of this disclosure;

[0047] Figure 3 This disclosure provides a flowchart of a script mounting process for a product manufacturing process system. Detailed Implementation

[0048] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0049] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of this disclosure. However, it will be apparent to those skilled in the art that this disclosure may be practiced without one or more of these details. In other instances, to avoid confusion with this disclosure, certain technical features well-known in the art have not been described; that is, not all features of actual embodiments are described herein, nor are well-known functions and structures described in detail.

[0050] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the terms “comprise” and / or “comprising,” when used in this specification, identify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0051] To fully understand this disclosure, detailed steps and structures will be presented in the following description to illustrate the technical solutions of this disclosure. Preferred embodiments of this disclosure are described in detail below; however, other embodiments may also be implemented in addition to these detailed descriptions.

[0052] In semiconductor manufacturing, different products are typically produced using different process flows. A batch of wafers produced using the same process flow is called a wafer lot. Each lot is processed according to its specific process flow, entering different machines to perform different processes. For each machine, the lot to be processed by that machine is called the Lot to be processed. Due to the repetitive nature of processes in semiconductor manufacturing, different process flows may include some identical processes, such as coating, etching, and photolithography.

[0053] In some embodiments, a preset duration is used to control the waiting time (Queue Time, hereinafter referred to as Q_Time) during critical process steps after the completion of the previous step. The mechanism for controlling Q_Time involves selecting two stations along the Lot's process flow as the execution interval of Q_Time, setting the preceding station as the start station and the following station as the target station, setting the duration of Q_Time, and defining the operation for ending Q_Time. Q_Time is triggered when the Lot reaches the start station. If the duration of Q_Time ends at any point before the Lot reaches the target station, the operation for ending Q_Time is executed.

[0054] However, in the above embodiments, a Q_Time can only set a set of stations (one starting station and one target station) and correspond to a duration. Therefore, the execution interval of a Q_Time only has a single judgment condition, namely whether the duration has ended. It cannot meet the comprehensive judgment of multiple situations that occur in the execution interval of a Q_Time (such as the duration of Q_Time ending multiple times at any time before the Lot reaches the target station).

[0055] Furthermore, only one correction action can correspond to the end of Q_Time, and this correction action is a specific correction operation, such as hold or rework. The above embodiment cannot implement combined control logic containing a set of correction operations when Q_Time ends. Usually, after Q_Time ends, relevant personnel need to control the Lot to execute one or more related correction operations based on their experience before dispatching the Lot to the machine in the next process for processing.

[0056] Furthermore, in a multi-Q_Time composite design, at any given moment, the system only determines whether the Q_Time with the shortest duration has ended. If it has ended, only the operation for the Q_Time at the end is executed, and the end flag for that Q_Time is cleared. Multiple Q_Times will not act in concert.

[0057] The ability to accurately and in real-time determine the production process status of each batch of Lots between the machine in the current process and the machine in the next process, and to dynamically allocate the optimal control logic to the production process of each batch of Lots, plays a crucial role in improving the production efficiency of each batch of Lots.

[0058] Figure 1 This disclosure provides a flowchart of a method for controlling a product manufacturing process.

[0059] like Figure 1 As shown, according to a first aspect of the present disclosure, a method for controlling a product manufacturing process is provided, comprising the following steps:

[0060] S101. Obtain multiple different dimensions of process information corresponding to each of the process stations in one or more process stations within a preset time period;

[0061] S102. Generate an influence weight matrix based on the process information of multiple different dimensions; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions.

[0062] S103. Generate an operation mapping table based on the influence weight matrix; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes of correction operations whose importance exceeds a preset level; the operation mapping table is used to guide the product to perform the corresponding correction operations.

[0063] Here, the process can also be understood in the industry as running goods or dispatching goods, and the process control method can also be understood as control logic or running goods logic.

[0064] In semiconductor manufacturing, different products typically employ different production processes (or workflows). A batch of wafers using the same production process is called a wafer set. Each lot is processed according to its specific production process, entering different process stations (a set of one or more machines used to perform that process) for further processing. For each process station, the lot to be processed by that station is called the Lot to be processed. Since semiconductor manufacturing processes may include some identical steps, such as deposition, etching, and photolithography, etc.

[0065] Here, a preset time is set to control the waiting time (Queue Time, hereinafter referred to as Q_Time) of key process steps during manufacturing. The Q_Time execution interval refers to the time required for the product (or Lot) to travel from the previous process station to the next process station. Within the Q_Time execution interval, process information of the Lot at each process station is recorded in different dimensions. The process information in different dimensions includes multi-dimensional data that affects the end of Q_Time.

[0066] Here, the influence weight matrix includes: multiple process information of different dimensions that meet different preset conditions trigger various corrective operations to be performed on the product, and the weights corresponding to each corrective operation. Multiple process information of different dimensions that meet different preset conditions may have various effects on the production process; the occurrence of each type of process information may have the same or different effects on the production process, one type of process information may have multiple effects on the production process, and one effect may also be caused by multiple types of process information. The weights of the multiple effects are calculated; and the operation corresponding to each effect is predefined to obtain the influence weight matrix.

[0067] Here, the correction operation includes at least one of the following: Rework; Branch; Lot Hold; Skip Mea. The weights corresponding to each correction operation indicate the importance of each correction operation.

[0068] Here, the operation mapping table includes: corrective operations whose importance exceeds a preset level, and the corresponding operation processes (subsequent related process stations) for corrective operations whose importance exceeds the preset level. The operation mapping table is used to guide the product to perform the corresponding corrective operations at subsequent related process stations.

[0069] In this embodiment, within the Q_Time execution interval, process information of the Lot at each process station is recorded in different dimensions. This process information includes multi-dimensional data affecting the end of Q_Time. Any node where Q_Time ends within the Q_Time execution interval can trigger a judgment mechanism. Upon the end of Q_Time, an influence weight matrix is ​​generated based on the process information in multiple different dimensions. An operation mapping table is also generated based on the influence weight matrix. This operation mapping table is used to guide the product to perform corresponding corrective operations when the Lot arrives at subsequent related process stations.

[0070] Perform step S101 to obtain process information from multiple different dimensions.

[0071] Figure 2 A flowchart is provided for another product manufacturing process control method according to embodiments of this disclosure.

[0072] like Figure 2 As shown, in some embodiments,

[0073] The acquisition of multiple different dimensions of process information corresponding to each of the multiple process stations within a preset time period includes:

[0074] When the product travels to the start station, the timing is started. The start station is the first process station among the plurality of process stations. When the product travels to the end station, the timing is stopped. The end station is the process station after the start station among the plurality of process stations.

[0075] The duration is determined when the product travels to any process station after the starting station, or when the product completes each process station after the starting station.

[0076] If the timing duration has not reached the preset duration, continue to acquire process information of multiple different dimensions corresponding to each process station;

[0077] The step of generating an influence weight matrix based on the multiple different dimensions of process information includes:

[0078] When the timing duration reaches or exceeds the preset duration, the influence weight matrix is ​​generated based on the acquired process information from multiple different dimensions.

[0079] The Q_Time operation mechanism selects the two stations before and after the Lot's process flow as the execution interval of Q_Time, sets the previous station as the start station of Q_Time, and sets the next station as the end station of Q_Time (one of the process stations after the start station among multiple process stations, which can be the target station), sets the conditions for Q_Time to end (including the duration of the timer when Q_Time ends), and sets the operation when Q_Time ends.

[0080] When the Lot travels to the Q_Time start station, it triggers the Q_Time timing. If the Q_Time ends at any time before the Lot reaches the target station, the following operations are performed: if the timing duration has not reached the preset duration, continue to acquire multiple different dimensions of process information corresponding to each process station; if the timing duration reaches or exceeds the preset duration, generate the influence weight matrix based on the acquired multiple different dimensions of process information.

[0081] As shown in Table 1, process information with multiple different dimensions is recorded: within the Q_Time execution interval, process information of each process station (one of Ope.1 to Ope.N) is recorded (at least one of Condition1 to ConditionN).

[0082] Table 1

[0083] Condition1 Condition2 Condition 3 …… ConditionN Ope.1 Ope.2 Ope.3 Ope.4 …… Ope.N-1 Ope.N

[0084] The process information in different dimensions (at least one of Condition1 to ConditionN) includes at least one of the following: waiting time, process time, measurement result alarm status, and whether a timeout occurred. The process information in different dimensions contains multi-dimensional data that affects the end of Q_Time, including at least one of the following: waiting time timeout; process timeout; measurement result deviation (OOS, Out Of Specification); and judgment node timeout.

[0085] In some embodiments, at least one process station from Ope.1 to Ope.N is designated as a decision node. This decision node is used to determine whether a timeout has occurred after Q_Time is completed and to classify the timeout situation when generating the influence weight matrix. In practical applications, the selection of which process stations from Ope.1 to Ope.N to serve as decision nodes can be made based on the production process content and conditions, combined with experience.

[0086] For example, process station Ope.N is the target station, while process station Ope.4 is the decision node node1 and process station Ope.7 is the decision node node2. Decision nodes node1 and node2 are used to determine whether Q_Time has timed out after the generation of the influence weight matrix and to classify the timeout situation.

[0087] In some specific embodiments, process information for each process station is recorded within the Q_Time execution interval.

[0088] Step S102: Generate an influence weight matrix based on process information from multiple different dimensions.

[0089] In some embodiments, an influence weight matrix is ​​generated based on the process information of multiple different dimensions, including:

[0090] A criterion matrix is ​​generated based on the process information from multiple different dimensions; the criterion matrix includes the relationship between the process information satisfying different preset conditions and the various correction operations to be performed; and

[0091] The influence weight matrix is ​​generated based on the criterion matrix.

[0092] Multiple layers of process information can have various impacts on the production process. A single piece of process information can have multiple impacts, and a single impact can be caused by multiple pieces of process information. Therefore, after predefining the correction operations corresponding to each impact, process information that satisfies different preset conditions is selected. The impacts corresponding to process information that satisfies different preset conditions, and the corresponding correction operations, can then be used to obtain the relationship between process information that satisfies different preset conditions and the correction operations to be performed, thus obtaining the criterion matrix.

[0093] After the criterion matrix is ​​determined, the judgment matrix determines the final result: by calculating the weight of the influence, that is, calculating the importance of each influence corresponding to the process information that meets different preset conditions, we obtain the relationship between the process information of multiple different dimensions that meets different preset conditions to trigger the product to perform various correction operations and the weights corresponding to each correction operation, thus obtaining the influence weight matrix.

[0094] In some embodiments, a criterion matrix is ​​generated based on the process information of the plurality of different dimensions, including:

[0095] Obtain the threshold values ​​of different preset conditions in the multiple different dimensions of process information;

[0096] At the end of the preset time period, process information that exceeds the corresponding threshold is filtered out from the multiple different dimensions of process information;

[0097] Based on the process information that exceeds the corresponding threshold, determine the corrective operations to be performed on the product.

[0098] As shown in Table 2, in some specific embodiments, when Q_Time ends, the judgment condition for the end of Q_Time of the corresponding process station is checked, and the threshold of each different preset condition is obtained. The judgment condition that exceeds the threshold is updated according to the correction operation corresponding to its impact on the production process.

[0099] As shown in Table 2, in some specific embodiments, the process information of multiple dimensions includes: whether the judgment nodes (Node1 to NodeK) before this process station have experienced Q_Time termination, and the correction operations corresponding to the impact of the Q_Time termination of the judgment nodes (Node1 to NodeK) before this process station on the production process. It can be understood that, as shown in Table 2, the process information of multiple dimensions here includes: waiting time, process time, measurement result alarm status, timeout status, etc. (at least one of Condition1 to ConditionN), and whether the judgment nodes (Node1 to NodeK) before this process station have experienced Q_Time termination (at least one of Node1 Over Time to NodeK Over Time).

[0100] As shown in Table 2, the impact of the system initialization generating process information of each different dimension (at least one of Condition1 to ConditionN, at least one of Node1 Over Time to NodeK Over Time) (at least one of Impact1 to ImpactM-1, No_Impact).

[0101] Table 2

[0102] Condition1 Condition2 Condition 3 …… ConditionN Node1 Over Time Node2 Over Time …… NodeK Over Time Impact1 Impact2 Impact3 …… Impact M-1 No_Impact

[0103] As shown in Table 2, in some specific embodiments, multiple judgment nodes (Node1 to NodeK) can be set within the Q_Time execution interval. When each judgment node experiences the end of Q_Time, the process information of that process station is continuously recorded. When the Q_Time interval ends, that is, when the end station is reached (a process station after the start station among multiple process stations, which can be the target station), the judgment mechanism is activated to generate an operation mapping table and then the process information of that process station is cleared.

[0104] In some embodiments, generating the influence weight matrix based on the criterion matrix includes:

[0105] Based on the criterion matrix, the scale of each correction operation to be performed is calculated, whereby the scale represents the priority and trigger count of the correction operation.

[0106] A judgment matrix is ​​formed based on the scale of each correction operation to be performed. The judgment matrix includes a comparison between any two scales among the multiple scales corresponding to each correction operation to be performed.

[0107] The influence weight matrix is ​​generated based on the judgment matrix.

[0108] There are many ways to generate the influence weight matrix based on the criterion matrix. In some specific embodiments, the influence weight matrix is ​​generated by combining the criterion matrix with the Analytic Hierarchy Process (AHP). In this embodiment, the concept of the Analytic Hierarchy Process is introduced. By establishing a Q_Time matrix model judgment mechanism, the multiple conditions and multiple impacts that cause the Q_Time interval to end are qualitatively and quantitatively analyzed to dynamically generate an operation mapping table, so as to accurately execute the corresponding correction operations of the target site.

[0109] Specifically, as shown in Table 3, the scale of each correction operation to be performed is calculated according to the criterion matrix. The scale represents the priority and trigger count of the correction operation. During system initialization, the priority of each influence is defined, and the priority values ​​of influence 1 to influence M range from Lm to L1. Based on the data provided by the criterion matrix, the scale value is the product of the trigger count and the priority value of the correction operation. For example, the scale value of ImpactM is the product of the priority Lm of the correction operation and the number of conditions that cause Impact1.

[0110] Table 3

[0111] Scale Calculation method Impact1 D1 L1*(number of conditions that cause Impact1) Impact2 D2 L2*(number of conditions that cause Impact1) Impact3 D3 L3*(number of conditions that cause Impact1) …… …… …… ImpactM Dm Lm*(number of conditions that cause Impact1)

[0112] As shown in Tables 3 and 4, a judgment matrix is ​​constructed based on the scale corresponding to the influence. However, comparing multiple influences at the same time can lead to strong interference. Therefore, a pairwise comparison method is used to improve the reliability of the influence judgment.

[0113] Table 4

[0114] Impact 1 Impact 2 Impact 3 …… Impact M Impact 1 1 D1-D2 D1-D3 …… D1-Dm Impact 2 1 / D1-D2 1 D2-D3 …… D2-Dm Impact 3 1 / D1-D3 1 / D2-D3 1 …… D3-Dm …… …… …… …… 1 …… Impact M 1 / D1-Dm 1 / D2-Dm 1 / D3-Dm …… 1

[0115] The rows and columns of the judgment matrix represent the priority values ​​of influence 1 to influence M (referred to as Impact1 to ImpactM in the table and below), ranging from Lm to L1 (Lm>……>L1), forming pairwise comparisons. This matrix is ​​a square matrix. Dij represents the importance of the comparison between Impacti and Impactj, calculated as follows: if Di>Dj, then Dij=Di-Dj; if Di≤Dj, then Dij=1 / (Dj–Di). It can be seen that when Dij is greater than 1, it means that Impact i is more important than Impact j, and the value of the influence Dji of Impact j on Impact i is defined as the reciprocal of Dij; therefore, the values ​​of the diagonals are all 1; where Di represents the priority value of each influence in the row of the judgment matrix (the value of one of the scales of Impact1 to ImpactM), and Dj represents the priority value of each influence in the column of the judgment matrix (the value of one of the scales of Impact1 to ImpactM). For ease of representation, it is assumed that D1>D2>D3……>Dm, and the corresponding influences can be represented as shown in Table 4 below.

[0116] In some embodiments, generating the influence weight matrix based on the judgment matrix includes:

[0117] The column vectors of the judgment matrix are normalized.

[0118] The average value of the row vectors of the judgment matrix after normalization is calculated to obtain the influence weight matrix.

[0119] In some specific embodiments, the influence weights are calculated by judging the matrix, which can be achieved using the arithmetic mean method. Specifically:

[0120] Step 1: Calculate the normalized matrix of column D of the judgment matrix. The above Table 4 can be equivalent to the judgment matrix D, as shown in equation (1); the following equation (2) is obtained by calculating from equation (1).

[0121] Judgment Matrix

[0122] Column normalized matrix

[0123] Step 2: Calculate the row average matrix of the column normalized matrix A. Equation (3) is obtained from equation (2) above.

[0124] Row average matrix

[0125] The row average matrix W and the column vectors corresponding to the weights of Impact1 to ImpactM are W1, W2, W3, W4, W5, W6, W7, W8, W9, W1 ... 11 W 21 ...Wm1 .

[0126] Execute step S103: Generate an operation mapping table based on the influence weight matrix.

[0127] In some embodiments, generating an operation mapping table based on the influence weight matrix includes:

[0128] Obtain the first correction operation corresponding to the maximum importance level and the first operation process corresponding to the correction operation corresponding to the maximum importance level.

[0129] The operation mapping table is generated based on the first correction operation and the first operation process.

[0130] As shown in Table 5, during system initialization, the corrective operation (action11, action12, action32, action33, actionM1, actionM3, ... at least one of each) corresponding to each impact (at least one of Impact1 to ImpactM) is defined at each process station (Ope.1, Ope.2, Ope.3, Ope.4, ..., Ope.N-1, Ope.N, Ope.N+1, Ope.N+2, Ope.N+3, ...). For example, ImpactM executes corrective operation actionM1 at process station Ope.N+1 and corrective operation actionM3 at process station Ope.N+3.

[0131] Here, for a single process station, the corrective operation performed by that single process station is explicit and unique. That is, for each impact (at least one of Impact1 to ImpactM), the corresponding corrective operation (action11, action12, action32, action33, actionM1, actionM3, ...) performed at each process station (Ope.1, Ope.2, Ope.3, Ope.4, ..., Ope.N-1, Ope.N, Ope.N+1, Ope.N+2, Ope.N+3...) is explicit and unique. In other words, for a single process station, there is only one standard rework route. Different process stations each have their own rework routes, which may be the same or different.

[0132] In other embodiments, for a single process station, there may be multiple standard rework processes for different situations. These rework processes may be the same or different. When a specific situation occurs, the corresponding rework process is executed.

[0133] Table 5

[0134] Ope.N+1 Ope.N+2 Ope.N+3 …… Impact1 - action12 - …… Impact2 action11 - - …… Impact3 action32 action33 …… …… …… …… …… …… ImpactM actionM1 - actionM3 ……

[0135] Each process station includes: the one or more process stations include Ope.1, Ope.2, Ope.3, Ope.4, ..., Ope.N-1, Ope.N, and the operation process (subsequent related process stations) corresponding to the correction operation whose importance exceeds the preset level includes Ope.N+1, Ope.N+2, Ope.N+3, ...

[0136] As shown in Table 6, process station Ope.N represents the process station where Q_Time ends. After generating the influence weight matrix, the first correction operation corresponding to the maximum importance (WmaxImpacti) and the first operation process corresponding to the correction operation (Action) corresponding to the maximum importance are extracted. For example, the first correction operation action1 corresponding to the maximum importance is extracted, and the first operation process Ope.N+1 corresponding to the first correction operation action1 is extracted.

[0137] As shown in Table 6, an operation mapping table is constructed based on the first correction operation and the first operation process. The operation mapping table is used to guide the product to perform corresponding correction operations at subsequent relevant process stations.

[0138] Table 6

[0139] <![CDATA[W max Impact i ]]> Action Ope.N+1 actioni1 Ope.N+2 - Ope.N+3 actioni3 …… ……

[0140] It should be noted that an operation mapping table will be generated whenever Q_time ends.

[0141] In some specific embodiments, the operation mapping table corresponds to multiple different dimensions of process information, i.e., multiple different dimensions of process information generated under multiple Q_time termination conditions, including waiting time, process duration, measurement result alarm status, and whether a timeout occurred (at least one of Condition1 to ConditionN); it also includes whether the judgment nodes (Node1 to NodeK) before each process station have experienced Q_Time termination (at least one of Node1 Over Time to NodeK Over Time). The fact that the judgment nodes before each process station have experienced Q_Time termination is included in multiple different dimensions of process information, thus incorporating the impact of whether or not a Q_Time termination has occurred into the operation mapping table.

[0142] In other specific embodiments, the operation mapping table corresponds to multiple different dimensions of process information, i.e., multiple different dimensions of process information generated under multiple Q_time termination conditions, including waiting time, process duration, measurement result alarm status, timeout status, etc. (at least one of Condition1 to ConditionN); whether the judgment node before each process station has experienced Q_Time termination is not included in the multiple different dimensions of process information. Before generating the operation mapping table, it can be determined whether the judgment node before each process station has experienced Q_Time termination, thereby generating an operation mapping table corresponding to whether the judgment node before each process station has experienced Q_Time termination; in this way, the impact caused by whether or not Q_Time termination has occurred can also be included in the operation mapping table.

[0143] like Figure 2 As shown, in some embodiments, after generating the operation mapping table, the method further includes:

[0144] The location of the product is determined to determine whether it has reached the target site; the target site is the last of the plurality of process sites.

[0145] If the product does not reach the target site, continue to acquire process information of multiple different dimensions corresponding to each operation process;

[0146] When the product arrives at the target site, the corresponding correction operation of the target site is executed according to the information in the operation mapping table, or the product production process of the target site is controlled normally.

[0147] like Figure 2 As shown, in some embodiments, performing corresponding corrective operations or normally controlling the product manufacturing process of the target site based on the information in the operation mapping table includes:

[0148] When the duration of the timekeeping is greater than or equal to the preset duration, the corresponding correction operation of the target station is performed according to the information in the operation mapping table;

[0149] When the duration of the timing is less than the preset duration, the product manufacturing process at the target site is controlled normally.

[0150] This disclosure provides a method to set multiple Q_Time duration end determination nodes within a Q_Time interval. After the duration of each Q_Time ends before the target site, a matrix model determination mechanism can be invoked to draw a conclusion and dynamically match the subsequent process (shipping) logic. At the same time, the end flag and related data of each Q_Time in the process are retained for the determination of the end of Q_Time at the target site.

[0151] On the one hand, it addresses the limitation of Lot being able to determine only a single condition (whether the duration of Q_Time has ended) within the Q_Time execution interval. There are various reasons why Lot might end Q_Time, but not all reasons will affect process quality. For example, Lot might not be able to deliver in time due to machine issues, leading to the end of Q_Time. Therefore, it is necessary to obtain multi-dimensional information during the Lot delivery process and establish a matrix model judgment mechanism when the duration of Q_Time ends to obtain accurate conclusions.

[0152] On the other hand, this invention addresses the issue that only one correction operation can be executed after Q_Time ends, and that manual intervention is required. This embodiment matches a set of dispatch logic based on the judgment result of a matrix model. This set of dispatch logic can be a collection of subsequent site-specific correction operations. The judgment result and dispatch logic correspond through an operation mapping table, and the system executes them automatically using the operation mapping table.

[0153] This allows for precise and automatic determination of the reasons for the end of Q_Time, statistically analyzing the factors contributing to the end of Q_Time over a timeline, reducing the reliance on subjective human judgment. The system dynamically generates corresponding Lot dispatch logic, calculating the impact weights using a judgment matrix after Q_Time ends, qualitatively and quantitatively analyzing influencing factors to match appropriate handling methods. Operation mapping tables are recorded on the Lots, enabling precise control over each Lot batch. The entire inventory control process is completed by the system, reducing the risk of human error and judgment.

[0154] Figure 3 This disclosure provides a flowchart of a script mounting process for a product manufacturing process system.

[0155] like Figure 3 As shown, the above embodiments of this disclosure are implemented using script programs (hereinafter referred to as scripts), that is, the script is attached to the process station corresponding to the process in the product manufacturing process system, and the script is triggered to execute after the product moves to the process station.

[0156] In some embodiments, scripts are only mounted on associated process stations, minimizing the resource consumption of the product manufacturing process system. Triggered by the Lot, they provide real-time and precise control over the correction operations of each batch of Lots. In practical applications, scripts should be mounted on important process stations within the Q_Time execution interval, depending on process requirements, to determine the end of Q_Time, collect process information from multiple dimensions, or control the Lot to perform corresponding correction operations. Therefore, multiple scripts may work together.

[0157] In some embodiments, the script execution intervals are a front interval (the interval from the start to the completion of the operation at the current process station) and a back interval (the interval from the completion of the operation at the current process station to the start of the next process station). The front interval is used to control the Lot to perform correction operations, and the back interval is used to collect process information of multiple different dimensions as the Lot travels through the current process station.

[0158] like Figure 3 As shown in this embodiment of the disclosure, the one or more process stations include a startup station, Ope.1, Ope.2, Ope.3, Ope.4, ..., Ope.N-1, Ope.N, and a target station.

[0159] The embodiments disclosed above can be implemented using scripts. The script mounting status, script execution mechanism, and method flow within the Q_Time execution interval are detailed below:

[0160] After the Q_Time site starts, attach a Q_Time trigger script (QTriS, Q_Time Trigger Script) to trigger the Q_Time timer on the Lot.

[0161] The following script is pre-installed at each process station within the Q_Time execution interval:

[0162] The Q_Time Over Judge Script (QOJS) is used to determine whether the Q_Time of a Lot has expired.

[0163] The Criterion Matrices Create Script (CMCS) is used to generate the criterion matrix at the end of Q_Time.

[0164] The Judgment Matrices Create Script (JMCS) is used to generate judgment matrices and calculate influence weights.

[0165] The Action Mapping Table Create Script (AMTCS) is used to generate an action mapping table, determine the production process logic of the Lot, and guide the Lot to perform corresponding corrective operations at subsequent related process stations.

[0166] A ConditionRecord Script (CRS) is attached after each process station within the Q_Time execution interval to record process information of multiple different dimensions at that station.

[0167] The Q_Time target script (QTarS, Q_Time Target Script) AP_M_Qtime_Target_P1 is pre-attached to the Q_Time target site. The Q_Time target script QTarS is used to clear the end marker of Q_Time and process information of multiple different dimensions on the Lot.

[0168] Combination Figure 2 and Figure 3 The above embodiments of this disclosure can implement the flow of another product manufacturing process control method by attaching a script to a product manufacturing process system, as detailed below:

[0169] Execution step S201: Start the site and start the Q_Time timer:

[0170] At the Q_Time start site, start the Q_Time timer for the Lot that passes through the site, and record the Q_Time start time, i.e. the system time S_Qtime_StartTime and the Q_Time time limit C_Qtime_DurationTime (determined by system initialization). This is achieved by attaching the Q_Time trigger script QTriS after the start site.

[0171] Execution step S202: Recording process information from multiple different dimensions:

[0172] As shown in Table 7, within the Q_Time execution interval, process information of the Lot at each process station is recorded in different dimensions. The process information in different dimensions includes multi-dimensional data that affects the end of Q_Time. For example, the waiting time at the station, the process time, the measurement result alarm status, whether the timeout occurred, etc. This information is recorded on each batch of Lots.

[0173] The dimensions of information to be recorded are determined by the system initialization; the information recording script CRS is implemented when the Lot passes through the station. When the Lot passes through the station, the information recording script CRS will record multiple different dimensions of process information of the station on the Lot according to the preset information dimensions.

[0174] For example, the process information for process station Ope.6 includes a process type of Ope.Type as Measurement Mea., a waiting time of 20 minutes, a processing time of 50 minutes, a measurement result alarm status of Mea.result indicating OOS occurred, and an Over Flag indicating no timeout. The process information for process station Ope.7 includes a process type of Ope.Type as Measurement Mea., a waiting time of 20 minutes, a processing time of 50 minutes, a measurement result alarm status of Mea.result indicating no OOS occurred, and an Over Flag indicating no timeout (Y) (indicating a timeout).

[0175] For example, the Over Flag condition of timeout occurred 3 times: at process station Ope.7, Ope.N-1, and Ope.N. The Measurement Result Alarm condition Mea.result occurred 2 times: at process station Ope.2 and Ope.6.

[0176] Table 7

[0177] Ope.Type Process Mea. Mea. Mea. Process Mea. Mea. …… Process Process Waiting Time 10min 20min 20min 20min 10min 20min 20min …… 10min 10min Processing Time 100min 50min 50min 50min 100min 50min 50min …… 100min 50min Mea.result 00S 00S …… Over Flag Y …… Y Y Ope.1 Ope.2 Ope.3 Ope.4 Ope.5 Ope.6 Ope.7 …… Ope.N-1 Ope.N

[0178] Execution step S203, Q_Time end judgment:

[0179] When a process station enters the Lot within the Q_Time execution interval, it is determined whether Q_Time has ended. The Q_Time end judgment script QOJS is pre-attached to the station to obtain the current system time S_CurrentTime, calculate whether the difference between the current time S_CurrentTime and the system time S_Qtime_StartTime is greater than the Q_Time limit time C_Qtime_DurationTime, and if so, Q_Time ends. The Lot's Q_Time end marker is recorded.

[0180] Execute step S204, generate the criterion matrix:

[0181] As shown in Table 8, the Criterion Matrix Creation Script (CMCS) is cascaded after the script for determining the end of Q_Time. When the marker for the end of Q_Time is "Y", the criterion matrix generation is triggered. The CMCS script will, based on the process information of different dimensions of the current station recorded in the multi-dimensional judgment condition record, find the thresholds for the process information of different dimensions in the criterion matrix, determine whether the thresholds are exceeded, and extract the corresponding effects of the process information of different dimensions that exceed the thresholds to form the criterion matrix. For example, the data structure of the criterion matrix model is the same as the criterion matrix, with values ​​being thresholds, requiring initial system determination. The initial system determination affects the correspondence between the multi-dimensional data of the end of Q_Time and the correction operation.

[0182] As shown in Table 8, the multi-dimensional data affecting the end of Q_Time, cascaded after the script for determining the end of Q_Time, includes at least one of the following: Waiting Time Over; Processing Time Over; Measurement Result Alarm Mea.result OOS; Node1 OverTime; Node2 OverTime. Corrective actions include at least one of the following: Rework; Branch; Lot Hold; Skip Mea.

[0183] For example, a Processing Time Over corresponds to re-executing the Rework operation. A Measurement Result Alarm Mea.result OOS corresponds to batch execution of the Branch operation. Node Timeouts (Node1Over Time, Node2Over Time) both correspond to holding the current operation (Lot Hold). A Waiting Time Over corresponds to skipping the measurement process (Skip Mea.).

[0184] For example, the "Skip Mea" condition occurs twice: when the node timeout is determined (Node1 Over Time, Node2 Over Time). The "Rework" operation, the "Branch" operation, and the "Skip Mea" condition each occur once.

[0185] Table 8

[0186]

[0187] Execute step S205: Generate the influence weight matrix:

[0188] The judgment matrix creation script JMCS is concatenated after the script that generates the criterion matrix. The implementation method of the judgment mechanism is detailed below:

[0189] 1. Calculate the Impact scale

[0190] As shown in Table 9, the Impact term is obtained through the criterion matrix, and the Impact scale, i.e., D, is calculated. Impact i =L Impact i *Count Condition , where D Impact i L is the scale of Impact i Impact i For the priority of Impact (set during system initialization), Count Condition The Impact scaling table is shown in Table 9, which specifies the number of conditions that cause the Impact.

[0191] The priorities for Rework, Branch, Lot Hold, and Skip Mea are 7, 5, 3, and 1, respectively. Lot Hold occurs twice, while Rework, Branch, and Skip Mea each occur once.

[0192] Therefore, the scale value for maintaining the current operation Lot Hold is the product of the priority value of 3 corresponding to maintaining the current operation Lot Hold and the number of times the current operation Lot Hold occurs, which is 2, i.e., 3*2.

[0193] Table 9

[0194] Scale Calculation method Rework 7 7*1 Branch 5 5*1 Lot Hold 6 3*2 Skip Mea. 1 1*1 …… …… ……

[0195] 2. Construct the judgment matrix

[0196] As shown in Table 10, the judgment matrix is ​​a square matrix, and both the row vector and column vector are Impact, which are used for pairwise comparison of Impact. Dij represents the importance of the comparison between Impact i and Impact j.

[0197] Table 10

[0198] Rework Branch Lot Hold Skip Mea. …… Rework 1 2 4 6 …… Branch 1 / 2 1 2 4 …… Lot Hold 1 / 4 1 / 2 1 2 …… Skip Mea. 1 / 6 1 / 4 1 / 2 1 …… …… …… …… …… …… ……

[0199] Calculation method: if Di>Dj, Dij=Di-Dj; else Dij=1 / Di-Dj. An example of the Impact scaling matrix D is shown in Table 10.

[0200] 3. Calculate the Impact weight

[0201] Generate an influence weight matrix based on the judgment matrix: Calculate the influence weights using the judgment matrix. The following example uses the arithmetic mean method.

[0202] Step 1: Calculate the normalized matrix of column D of the judgment matrix. Table 10 above can be equivalent to the judgment matrix D, as shown in equation (4); equation (5) is obtained by calculating equation (4).

[0203] Judgment Matrix

[0204] Column normalized matrix

[0205] Step 2: Calculate the row average matrix of the column normalized matrix A. Equation (5) above is used to calculate the following equation (6).

[0206] Row average matrix

[0207] The row average matrix W and the column vectors corresponding to Impact1 to ImpactM have weights of 0.51, 0.28, 0.14, and 0.08, respectively.

[0208] Execute step S206: Generate operation mapping table:

[0209] As shown in Table 11, the script for generating the judgment matrix is ​​followed by the script for generating the operation mapping table (AMTCS). The implementation method of the AMTCS for generating the operation mapping table is detailed below:

[0210] Table 11

[0211] Ope.N+1 Ope.N+2 Ope.N+3 …… 返工 - 返工路线 - …… 分支 子路线 - - …… 批次暂停 - 暂停操作1 暂停操作2 …… 跳过测量 - 门卡 - …… …… …… …… …… ……

[0212] As shown in Table 11, during system initialization, the corrective operations (Rework route, Subroute, Hold Ope1, Hold Ope2, Gate pass, etc.) corresponding to each effect (Rework operation, Branch operation, Lot Hold current operation, Skip Mea. at least one) are defined at each process station (Ope.N+1, Ope.N+2, Ope.N+3... at least one) and executed at each process station (Ope.N+3). For example, Rework operation executes the corrective operation Rework route at process station Ope.N+2, but there are no corrective operations at process stations Ope.N+1 and Ope.N+3.

[0213] Each process station includes: the one or more process stations include Ope.1, Ope.2, Ope.3, Ope.4, ..., Ope.N-1, Ope.N, and the operation process (subsequent related process stations) corresponding to the correction operation whose importance exceeds the preset level includes Ope.N+1, Ope.N+2, Ope.N+3, ...

[0214] As shown in Table 12, based on the influence weight matrix generated by the matrix judgment model, the correction operation corresponding to the maximum weight is selected (the system initializes and presets the correction operation corresponding to each influence to be executed at each process station), an operation mapping table is generated, and the system controls the production process based on the operation mapping table carried on the Lot.

[0215] As shown in Table 12, process station Ope.N represents the process station where Q_Time ends. After generating the influence weight matrix, the first correction operation (Action) corresponding to the maximum importance (WmaxImpacti) and the first operation process corresponding to the correction operation corresponding to the maximum importance are extracted. For example, the first correction operation (Action) corresponding to the maximum importance is extracted, and the first operation process corresponding to the first correction operation (Action) is as follows: no correction operation at process station Ope.N+1, the correction operation Rework route is executed at process station Ope.N+2, and no correction operation at process station Ope.N+3.

[0216] Table 12

[0217] <![CDATA[W max Impact i ]]> 行动 操作N+1 - 操作N+2 返工路线 操作N+3 - …… ……

[0218] Execute step S207, determining whether the target station has been reached:

[0219] If the product does not reach the target site, return to step S202 to continue obtaining process information from multiple different dimensions corresponding to each operation process;

[0220] When the product arrives at the target site, step S208 is executed, and the corresponding correction operation of the target site is performed or the product production process of the target site is normally controlled according to the information in the operation mapping table.

[0221] According to a second aspect of the present disclosure, a product manufacturing process system is provided, the product manufacturing process system being configured to perform the following operations:

[0222] Obtain multiple different dimensions of process information for the product at one or more process stations within a preset time period;

[0223] Based on the process information of multiple different dimensions, an influence weight matrix is ​​generated; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions.

[0224] An operation mapping table is generated based on the influence weight matrix; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes for correction operations whose importance exceeds a preset level; the operation mapping table is used to guide the product to perform the corresponding correction operations.

[0225] In this embodiment of the disclosure, the product manufacturing process system may include: an information acquisition module, an influence weight matrix generation module, and an operation mapping table generation module; wherein...

[0226] The information acquisition module is configured to record process information of the product at each process station in different dimensions within the Q_Time execution interval. The process information in different dimensions includes multi-dimensional data that affects the end of Q_Time.

[0227] The influence weight matrix generation module is configured to trigger a judgment mechanism at any Q_Time end point within the Q_Time execution interval. When Q_Time ends, an influence weight matrix is ​​generated based on process information from multiple different dimensions.

[0228] The operation mapping table generation module is configured to generate an operation mapping table based on the influence weight matrix. The operation mapping table is used to guide the product to perform corresponding correction operations when the product enters the station in subsequent related process stations.

[0229] In some embodiments, the operations performed by the product manufacturing process tracking system are implemented through scripts. For details, please refer to the above. Figure 3 The relevant descriptions will not be repeated here.

[0230] This allows for accurate and real-time assessment of the production process status of each batch of products, and dynamic allocation of optimal control logic to the production process of each batch, playing a crucial role in improving the production efficiency of each batch.

[0231] It should be understood that the phrase "an embodiment" or "one embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this disclosure. Therefore, "in one embodiment" or "one embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this disclosure, the sequence numbers of the above-described processes do not imply a sequential order of execution; the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this disclosure. The sequence numbers of the above-described embodiments are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0232] The above description is merely a preferred embodiment of this disclosure and does not limit the patent scope of this disclosure. Any equivalent structural transformations made using the contents of this specification and drawings under the inventive concept of this disclosure, or direct / indirect applications in other related technical fields, are included within the patent protection scope of this disclosure.

Claims

1. A method for controlling a product manufacturing process, characterized in that, include: Obtain multiple different dimensions of process information for the product at one or more process stations within a preset time period; Based on the process information of multiple different dimensions, an influence weight matrix is ​​generated; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions. Based on the influence weight matrix, an operation mapping table is generated; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes of correction operations whose importance exceeds a preset level. The operation mapping table is used to guide the product to perform corresponding correction operations; The process of generating an influence weight matrix based on the process information of multiple different dimensions includes: generating a criterion matrix based on the process information of multiple different dimensions; the criterion matrix including the relationship between the process information satisfying different preset conditions and the various correction operations to be performed; calculating the scale of each correction operation to be performed based on the criterion matrix, the scale representing the priority and trigger number of the correction operation; forming a judgment matrix based on the scale of each correction operation to be performed, the judgment matrix including the comparison between any two scales among the multiple scales corresponding to each correction operation to be performed; and generating the influence weight matrix based on the judgment matrix.

2. The control method according to claim 1, characterized in that, Based on the process information from multiple different dimensions, a criterion matrix is ​​generated, including: Obtain the threshold values ​​of different preset conditions in the multiple different dimensions of process information; At the end of the preset time period, process information that exceeds the corresponding threshold is filtered out from the multiple different dimensions of process information; Based on the process information that exceeds the corresponding threshold, determine the corrective operations to be performed on the product.

3. The control method according to claim 1, characterized in that, Based on the judgment matrix, the influence weight matrix is ​​generated, including: The column vectors of the judgment matrix are normalized. The average value of the row vectors of the judgment matrix after normalization is calculated to obtain the influence weight matrix.

4. The control method according to claim 1, characterized in that, The acquisition of multiple different dimensions of process information corresponding to each of the multiple process stations within a preset time period includes: When the product travels to the start station, the timing is started. The start station is the first process station among the plurality of process stations. When the product travels to the end station, the timing is stopped. The end station is the process station after the start station among the plurality of process stations. The duration is determined when the product travels to any process station after the starting station, or when the product completes each process station after the starting station. If the timing duration has not reached the preset duration, continue to acquire process information of multiple different dimensions corresponding to each process station; The step of generating an influence weight matrix based on the process information from multiple different dimensions includes: When the timing duration reaches or exceeds the preset duration, the influence weight matrix is ​​generated based on the acquired process information from multiple different dimensions.

5. The control method according to claim 4, characterized in that, After generating the operation mapping table, the method further includes: The location of the product is determined to determine whether it has reached the target site; the target site is the last of the plurality of process sites. If the product does not reach the target site, continue to acquire process information of multiple different dimensions corresponding to each operation process; When the product arrives at the target site, the corresponding correction operation of the target site is executed according to the information in the operation mapping table, or the product production process of the target site is controlled normally.

6. The control method according to claim 5, characterized in that, Executing corresponding corrective operations at the target site or normally controlling the product manufacturing process at the target site based on the information in the operation mapping table includes: When the duration of the timekeeping is greater than or equal to the preset duration, the corresponding correction operation of the target station is performed according to the information in the operation mapping table; When the duration of the timing is less than the preset duration, the product manufacturing process at the target site is controlled normally.

7. The control method according to claim 1, characterized in that, Based on the influence weight matrix, an operation mapping table is generated, including: Obtain the first correction operation corresponding to the maximum importance level and the first operation process corresponding to the correction operation corresponding to the maximum importance level. The operation mapping table is generated based on the first correction operation and the first operation process.

8. A product manufacturing process system, characterized in that, The product manufacturing process system is configured to perform the following operations: Obtain multiple different dimensions of process information for the product at one or more process stations within a preset time period; Based on the process information of multiple different dimensions, an influence weight matrix is ​​generated; the influence weight matrix includes the importance of each correction operation to be performed on the product when the process information of multiple different dimensions meets different preset conditions. Based on the influence weight matrix, an operation mapping table is generated; the operation mapping table includes correction operations whose importance exceeds a preset level and the corresponding operation processes of correction operations whose importance exceeds a preset level. The operation mapping table is used to guide the product to perform corresponding correction operations; The process of generating an influence weight matrix based on the process information of multiple different dimensions includes: generating a criterion matrix based on the process information of multiple different dimensions; the criterion matrix including the relationship between the process information satisfying different preset conditions and the various correction operations to be performed; calculating the scale of each correction operation to be performed based on the criterion matrix, the scale representing the priority and trigger number of the correction operation; forming a judgment matrix based on the scale of each correction operation to be performed, the judgment matrix including the comparison between any two scales among the multiple scales corresponding to each correction operation to be performed; and generating the influence weight matrix based on the judgment matrix.

Images