Scheduling method and device for horizontal wafer assembling machine

Abstract

The invention provides a scheduling method and device of a wafer horizontal type assembling machine, which are applied to the technical field of wafer assembling machine scheduling, and are used for predicting the predicted arrival time of each wafer reaching the wafer horizontal type assembling machine by collecting wafer transmission state data in upstream equipment, and calculating the wafer delay time of the actual arrival time and the predicted arrival time. The method comprises the following steps: acquiring equipment state data of each wafer horizontal assembly machine, predicting idle time after each wafer horizontal assembly machine completes a current task, determining the priority of each wafer, and comprehensively considering the wafer delay time, the wafer priority and the idle time of the wafer horizontal assembly machine; calculating a comprehensive score of distributing the wafer to the corresponding wafer horizontal assembling machine; and finally, an optimal allocation scheme is selected and executed in real time, so that efficient scheduling of the wafer horizontal assembling machine is realized. Therefore, the method has the advantage of improving the production efficiency of a wafer manufacturing production line.

Description

Technical Field

[0001] This application relates to the technical field, and in particular to a scheduling method and device for a horizontal wafer assembly machine. Background Art

[0002] In a modern wafer manufacturing production line, the horizontal wafer assembly machine, as an indispensable key device, undertakes high-precision wafer assembly tasks, which is crucial for ensuring the efficient and stable operation of the production line. To meet the growing production demands, wafer manufacturing production lines usually configure multiple horizontal assembly machines to work together and highly rely on an automated material handling system to achieve fast and accurate transmission of wafers between various devices, with the aim of maximizing production efficiency. In an ideal production state, wafers should be able to arrive precisely at each assembly machine strictly according to the predetermined time plan, and the assembly machine should also be able to execute operations efficiently completely according to the established scheduling plan, thereby ensuring the smoothness and efficiency of the entire production process.

[0003] However, in the actual wafer production environment, the material handling system is often inevitably affected by various complex disturbance factors, such as slight vibrations during equipment operation, sudden failures in local areas, and other unpredictable accidental situations, etc. These factors may all lead to instantaneous delays in wafer transmission. Especially when the wafers planned to enter a certain assembly machine fail to arrive at the designated station on time, the assembly machine will have to enter a waiting state, directly affecting the production rhythm of the entire production line and ultimately resulting in a significant reduction in overall production efficiency. Traditional methods for dealing with material delays still largely rely on manual intervention for adjustment. This method not only has a relatively slow response speed but is also prone to errors due to human factors and is difficult to fully meet the stringent requirements for real-time and precision of high-tempo wafer production lines. Especially in a highly automated wafer manufacturing scenario, for the problem of occasional delays in the material handling system, there is an urgent need for an intelligent scheduling method that can achieve automation and rapid response, thereby effectively ensuring the efficient collaborative operation of the horizontal wafer assembly machine group and further improving the production efficiency and intelligent level of the entire wafer manufacturing production line.

[0004] In view of the above problems, the existing technology urgently needs to be improved. Summary of the Invention

[0005] In view of the deficiencies of the above-mentioned prior art, this application provides a scheduling method and device for a horizontal wafer assembly machine, which has the beneficial effect of improving the production efficiency of the wafer manufacturing production line.

[0006] In a first aspect, a scheduling method for a horizontal wafer assembly machine, which is used for an automated wafer manufacturing production line, the method includes the steps: S1: Collect the wafer transfer status data in the upstream equipment, predict the predicted arrival time of each wafer at the wafer horizontal assembly machine based on the transfer status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; S2: Collect the equipment status data of each wafer horizontal assembly machine, and predict the idle time of each wafer horizontal assembly machine after completing the current task; S3: Determine the priority of each wafer, and calculate the comprehensive score of allocating the wafer to the corresponding wafer horizontal assembly machine based on the wafer delay time, the priority, and the idle time of the wafer horizontal assembly machine; S4: Select the wafer with the lowest comprehensive score and the allocation scheme of the corresponding wafer horizontal assembly machine as the scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

[0007] Further, step S1 includes: S11: Collect the wafer transfer status data in the upstream equipment, and extract the handling path of the wafer, the speed of the handling equipment, and the traffic congestion situation of the handling system; S12: Predict the reference handling time of the wafer under the current handling path according to the handling path; S13: Obtain the speed data of the handling equipment in real time, calculate the deviation value between the current handling speed and the historical average handling speed, and adjust the reference handling time according to the deviation value to obtain the corrected predicted arrival time; S14: Obtain the predicted delay time according to the traffic congestion situation of the handling system, and obtain the predicted arrival time of the final wafer at the wafer horizontal assembly machine according to the corrected predicted arrival time and the predicted delay time; S15: Record the actual arrival time of the wafer at the wafer horizontal assembly machine, and calculate the difference between the actual arrival time and the predicted arrival time to obtain the wafer delay time.

[0008] Further, step S14 includes: S141: Monitor the traffic congestion situation of the handling system, count the number of handling equipment passing through the key nodes per unit time, and if the number exceeds the preset threshold, it is determined to be congested; S142: When the traffic congestion situation of the handling system is congested, determine the congestion degree, and obtain the predicted delay time according to the congestion degree. The sum of the corrected predicted arrival time and the predicted delay time is the predicted arrival time.

[0009] Further, step S142 includes: S1421: When the traffic congestion situation of the handling system is congested, record the congestion duration; the congestion duration reflects the congestion degree; S1422: Query the congestion delay schedule, and determine the predicted delay time of the wafer at the congestion critical node according to the congestion duration; the congestion delay schedule records the corresponding relationship between the congestion duration and the predicted delay time; S1423: Accumulate the predicted delay times of the wafer at each congestion critical node to obtain the total congestion delay time, and add the total congestion delay time to the predicted arrival time to obtain the final predicted arrival time.

[0010] Further, step S2 includes: S21: Collect the equipment status data of each wafer horizontal assembly machine, and read the task type and task volume of the task currently being executed according to the equipment status data; S22: Query the preset corresponding relationship table between the task type and the standard processing time according to the task type, and determine the standard processing time corresponding to the current task; S23: Calculate the remaining processing time of the current task according to the task volume and the standard processing time; S24: Predict the completion time of the current task by the wafer horizontal assembly machine based on the remaining processing time, and subtract the current time from the completion time to obtain the idle time of each wafer horizontal assembly machine after completing the current task.

[0011] Further, step S3 includes: S31: Obtain the customer order priority, product type urgency, and process step criticality corresponding to each wafer, construct a wafer priority evaluation model, determine the priority of the wafer according to the wafer priority evaluation model, and obtain the priority coefficient of the wafer; S32: Construct a dynamic scoring function according to the wafer delay time, the priority coefficient of the wafer, and the idle time of the wafer horizontal assembly machine; S33: Calculate the comprehensive score assigned to each corresponding wafer horizontal assembly machine according to the dynamic scoring function.

[0012] Further, step S323 includes: S321: Collect the historical processing data of the wafer horizontal assembly machine, where the historical processing data includes the wafer type, processing time, and yield rate; S322: Calculate the average processing time and average yield rate of each wafer horizontal assembly machine for different wafer types; S323: Determine the adaptability score between the currently to-be-allocated wafer type and each wafer horizontal assembly machine according to the average processing time and the average yield rate; S324: Normalize the adaptability score to obtain the assembly machine adaptability coefficient; S325: Construct a dynamic scoring function based on the wafer delay time, the priority coefficient of the wafer, the idle time of the wafer horizontal assembly machine, and the adaptability coefficient of the assembly machine.

[0013] Further, step S333 includes: S3231: Assign weights to the average processing time and the average yield respectively, and calculate the initial adaptability score of each assembly machine for the currently to-be-allocated wafer type. The sum of the weights of the average processing time and the yield is 1; S3232: Statistically calculate the standard deviations of the average processing time and the yield of all wafer types for each assembly machine. If the average processing time and the yield corresponding to the currently to-be-allocated wafer type are less than the corresponding standard deviations, reduce the initial adaptability score of this assembly machine. Conversely, if they are greater than the corresponding standard deviations, increase the initial adaptability score of this assembly machine to obtain the final adaptability score.

[0014] Further, step S4 includes: S41: Real-time monitor the communication link status between the scheduling system and each of the wafer horizontal assembly machines; the communication link status includes packet loss rate, delay time, and connection status; S42: When the communication link status meets any of the following conditions, start the instruction retransmission mechanism: the packet loss rate exceeds the first preset threshold, the delay time exceeds the second preset threshold, or the connection status is disconnected. The instruction retransmission mechanism includes: the scheduling system records the scheduling instructions that have been sent but not confirmed, and retransmits them at a preset time interval until the confirmation information from the assembly machine is received. <>

[0015] In a second aspect, a scheduling device for a wafer horizontal assembly machine is applied to the steps of any one of the above-mentioned scheduling methods for a wafer horizontal assembly machine. The device includes: An arrival time prediction module, configured to collect wafer transmission status data in upstream devices, predict the predicted arrival time of each wafer at the wafer horizontal assembly machine based on the transmission status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; An assembly machine status monitoring module, configured to collect the device status data of each wafer horizontal assembly machine and predict the idle time of each wafer horizontal assembly machine after completing the current task; A task allocation module, configured to determine the priority of each wafer, and calculate the comprehensive score of allocating the wafer to the corresponding wafer horizontal assembly machine based on the wafer delay time, the priority, and the idle time of the wafer horizontal assembly machine; A task execution module, configured to select the allocation scheme of the wafer with the lowest comprehensive score and the corresponding wafer horizontal assembly machine as the scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

[0016] Beneficial effects: A scheduling method and device for a wafer horizontal assembly machine proposed in this application realizes the automatic scheduling of the wafer horizontal assembly machine by predicting the wafer arrival time, the idle time of the assembly machine, considering the wafer priority, and calculating a comprehensive score, and has the beneficial effect of improving the production efficiency of the wafer manufacturing production line. Description of the drawings

[0017] Figure 1 It is a flowchart of a scheduling method for a wafer horizontal assembly machine proposed in this application.

[0018] Figure 2 It is a structural diagram of a scheduling device for a wafer horizontal assembly machine proposed in this application.

[0019] Label description: 201, arrival time prediction module; 202, assembly machine status monitoring module; 203, task assignment module; 204, task execution module. Detailed implementation manners

[0020] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Usually, the components of the embodiments of the present application described and marked in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the present application to be protected, but only represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0021] It should be noted that: Similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the present application, the terms "first", "second", etc. are only used for distinguishing descriptions, and cannot be understood as indicating or implying relative importance.

[0022] Please refer to Figure 1 , in the first aspect, a scheduling method for a wafer horizontal assembly machine, used for an automated wafer manufacturing production line, the method includes the steps of: S1: Collect the wafer transfer status data in the upstream equipment, predict the predicted arrival time of each wafer arriving at the wafer horizontal assembly machine based on the transfer status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; S2: Collect the equipment status data of each wafer horizontal assembly machine, and predict the idle time of each wafer horizontal assembly machine after completing the current task; S3: Determine the priority of each wafer, and calculate the comprehensive score for allocating the wafer to the corresponding wafer horizontal assembly machine based on the wafer delay time, priority, and the idle time of the wafer horizontal assembly machine; S4: Select the allocation plan of the wafer with the lowest comprehensive score and the corresponding wafer horizontal assembly machine as the scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

[0023] Among them, in step S1, the upstream equipment is the production line for transporting wafers, and the wafer transfer status data is collected, which can include information such as the transfer path of the wafer, the speed of the transfer equipment, and the traffic congestion situation of the transfer system. The predicted arrival time is predicted based on these data. For example, the transfer path determines the benchmark transfer time, the deviation of the transfer equipment speed can be used to adjust the benchmark transfer time, and the traffic congestion situation can further introduce the prediction of the delay time. The wafer delay time is calculated by comparing the actual arrival time and the predicted arrival time, which provides quantitative delay information for subsequent scheduling decisions.

[0024] In step S2, the equipment status data is collected from each wafer horizontal assembly machine, and the equipment status data reflects the current task status and processing capacity of the assembly machine. The prediction of the idle time is performed based on the equipment status data. For example, by analyzing the task type, task volume of the current task, and the processing speed of the equipment, the time required for the assembly machine to complete the current task can be predicted, and then the idle time can be obtained.

[0025] In step S3, the wafer priority is determined, which can be evaluated based on factors such as the priority of the customer order, the urgency of the product type, and the criticality of the process steps. The calculation of the comprehensive score comprehensively considers the wafer delay time, wafer priority, and the idle time of the wafer horizontal assembly machine. The design of the scoring function can be adjusted according to the actual production requirements. For example, wafers with longer delay times and higher priorities, when allocated to assembly machines with shorter idle times, can obtain lower comprehensive scores.

[0026] In step S4, the allocation plan with the lowest comprehensive score is selected as the scheduling instruction. The scheduling instruction is then sent to the corresponding wafer horizontal assembly machine for execution in real time, thereby realizing the automation and real-time nature of the scheduling. Through the above steps, the efficient scheduling of the wafer horizontal assembly machine can be achieved, and the impact of material delay on production efficiency can be reduced.

[0027] Further, step S1 includes: S11: Collect the wafer transfer status data in the upstream equipment, and extract the transfer path of the wafer, the speed of the transfer equipment, and the traffic congestion situation of the transfer system; S12: Predict the benchmark transfer time of the wafer under the current transfer path according to the transfer path; S13: Obtain the speed data of the handling equipment in real time, calculate the deviation value between the current handling speed and the historical average handling speed, and adjust the reference handling time according to the deviation value to obtain the corrected predicted arrival time; S14: Obtain the predicted delay time according to the traffic congestion situation of the handling system, and obtain the predicted arrival time when the wafer finally arrives at the wafer horizontal assembly machine according to the corrected predicted arrival time and the predicted delay time; S15: Record the actual arrival time when the wafer actually arrives at the wafer horizontal assembly machine, calculate the difference between the actual arrival time and the predicted arrival time, and obtain the wafer delay time.

[0028] Among them, in step S11, the wafer transfer status data may include the ID information of the current upstream equipment where the wafer is located, the batch information of the wafer, and the process flow information of the wafer. The handling path can be extracted as a series of coordinate points or a sequence of equipment IDs, which is used to describe the movement trajectory of the wafer in the handling system. The handling equipment speed can be obtained by real-time collection using a speed sensor installed on the handling equipment. The traffic congestion situation of the handling system can use the sensor data deployed at the key nodes of the handling system, such as optoelectronic sensors or cameras, to monitor the number of wafers or handling equipment passing through per unit time, so as to evaluate the congestion degree.

[0029] In step S12, the reference handling time can be set in advance. For example, it can be calculated according to the average handling time of different handling paths statistically based on historical data, or according to the theoretical length of the handling path and the rated speed of the handling equipment.

[0030] In step S13, the historical average handling speed can be the average value statistically for a long time, or the average value in a recent period of time. The deviation value can be the difference between the current handling speed and the historical average handling speed, or the ratio of the two. Adjusting the reference handling time can adopt a linear adjustment or a non-linear adjustment method. For example, when the deviation value is positive, reduce the reference handling time, and when the deviation value is negative, increase the reference handling time.

[0031] In step S14, the predicted delay time can be determined according to a preset correspondence table between the traffic congestion level and the delay time. The traffic congestion level can be divided according to the traffic congestion situation data collected in step S11. For example, it can be divided into multiple levels such as unobstructed, slightly congested, moderately congested, and severely congested.

[0032] In step S15, the actual arrival time can be detected by using a sensor installed at the entrance of the wafer horizontal assembly machine to detect the arrival moment of the wafer. The wafer delay time can be the difference obtained by subtracting the predicted arrival time from the actual arrival time.

[0033] Further, step S14 includes: S141: Monitor the traffic congestion situation of the handling system, count the number of handling devices passing through the key nodes within a unit time. If the number exceeds the preset threshold, it is determined as congestion. S142: When the traffic congestion situation of the handling system is congested, determine the congestion level, and obtain the predicted delay time according to the congestion level. Modify the sum of the predicted arrival time and the predicted delay time to be the predicted arrival time.

[0034] Among them, in step S141, the monitoring of the traffic congestion situation of the handling system is realized. Specifically, key nodes are preset on the handling path. For example, they can be positions prone to congestion such as path intersections and equipment entrances. Monitoring devices are deployed at these key nodes to count the number of handling devices passing through within a unit time in real time. The preset threshold is used as the standard for judging whether there is congestion, and this threshold can be set according to historical data, simulation analysis or actual experience. For example, if the average number of handling devices passing through the key node within a unit time under normal circumstances is 15, the preset threshold can be set to a value slightly lower than the average, such as 10. When it is monitored that the number passing through within a unit time is less than 10, the system is determined to be congested.

[0035] In step S142, the determination of the congestion level and the obtaining of the predicted delay time are realized. Further, the congestion level can be quantified into multiple levels. For example, mild congestion, moderate congestion and severe congestion. The congestion level can be divided according to the deviation degree of the number of handling devices passing through the key node within a unit time from the preset threshold. For example, when the passing number is lower than the preset threshold but higher than 80% of the threshold, it is determined as mild congestion; when it is lower than 80% but higher than 50%, it is determined as moderate congestion; when it is lower than 50%, it is determined as severe congestion. For each congestion level, the predicted delay time is preset. For example, mild congestion corresponds to a 1-minute delay, moderate congestion corresponds to a 3-minute delay, and severe congestion corresponds to a 5-minute delay. A congestion delay time table is established to store the correspondence between the congestion level and the predicted delay time. The system queries this table to obtain the corresponding predicted delay time according to the determined congestion level. Finally, the predicted delay time is added to the corrected predicted arrival time to obtain a more accurate predicted arrival time for the wafer to reach the wafer horizontal assembly machine.

[0036] Further, step S142 includes: S1421: When the traffic congestion situation of the handling system is congested, record the congestion duration; the congestion duration reflects the congestion level; S1422: Query the congestion delay time table, and determine the predicted delay time of the wafer at the congested key node according to the congestion duration; the congestion delay time table records the correspondence between the congestion duration and the predicted delay time; S1423: Accumulate the predicted delay time of the wafer at each congestion key node to obtain the total congestion delay time, add the total congestion delay time to the predicted arrival time, and obtain the final predicted arrival time.

[0037] In step S1421, the congestion duration can be recorded in the following manner: a timer is set inside the system, and when the transport system is determined to be in traffic congestion, the timer starts counting until the congestion is resolved, and the timer stops. The recorded time is the congestion duration. In step S1422, the congestion delay table can be pre-set, and the congestion duration and the predicted delay time are mapped to each other. For example, when the congestion duration is 10 seconds, the predicted delay time is 2 seconds; when the congestion duration is 20 seconds, the predicted delay time is 5 seconds.

[0038] For step S1423 , the total congestion delay time is obtained by summing up the predicted delay times of the wafer at all congestion key nodes, and the final predicted arrival time is obtained by adding the total congestion delay time to the corrected predicted arrival time.

[0039] Furthermore, step S2 includes: S21: collecting equipment status data of each wafer horizontal assembly machine, and reading the task type and task amount of the task currently being executed according to the equipment status data; S22: According to the task type, query the preset task type and standard processing time correspondence table to determine the standard processing time corresponding to the current task; S23: Calculate the remaining processing time of the current task based on the task volume and the standard processing time; S24: Based on the remaining processing time, predict the completion time of the wafer horizontal assembly machine to complete the current task, subtract the current time from the completion time, and obtain the idle time of each wafer horizontal assembly machine after completing the current task.

[0040] In step S21, the equipment status data can be collected in real time via the assembly machine's own sensor system or a host computer monitoring system. Information about the task type and task volume can be parsed from the equipment status data. For example, the equipment control system typically records the type of task currently being performed (e.g., alignment, bonding, etc.) and the number of wafers to be processed.

[0041] In step S22, the preset table of correspondences between task types and standard processing times can be a data table or configuration file stored in the scheduling system database. This table predefines various task types and their corresponding standard processing times. Once a task type is obtained, the system can quickly obtain the standard processing time for that task type by querying this table.

[0042] In step S23, for the calculation of the remaining processing time, specifically, if the task volume is represented by the number of wafers to be processed, the remaining processing time can be obtained by multiplying the standard processing time by the task volume. For example, if the standard processing time is 5 minutes per wafer and the current task volume is 10 wafers, the remaining processing time is 50 minutes.

[0043] In step S24, the predicted completion time is obtained by adding the remaining processing time to the current time. For the calculation of the idle time, it is the predicted completion time minus the current time, which is actually equivalent to the remaining processing time. Thus, the predicted value of the idle time of each wafer horizontal assembly machine after completing the current task can be obtained.

[0044] Furthermore, step S3 includes: S31: Obtain the customer order priority, product type urgency, and process step criticality corresponding to each wafer, construct a wafer priority evaluation model, determine the priority of the wafer according to the wafer priority evaluation model, and obtain the priority coefficient of the wafer; S32: Construct a dynamic scoring function based on the wafer delay time, the priority coefficient of the wafer, and the idle time of the wafer horizontal assembly machine; S33: Calculate the comprehensive score of each wafer assigned to the corresponding wafer horizontal assembly machine according to the dynamic scoring function.

[0045] Among them, in step S31, the construction of the wafer priority evaluation model can be implemented as follows: First, determine the weights of the customer order priority, product type urgency, and process step criticality respectively. For example, the weight of the customer order priority is 0.4, the weight of the product type urgency is 0.3, and the weight of the process step criticality is 0.3. Then, score the customer order priority, product type urgency, and process step criticality of each wafer. The scoring levels can be divided into three levels: high, medium, and low, corresponding to the scoring values 3, 2, and 1 respectively. The priority coefficient of the wafer is obtained by weighted summation, and the calculation formula can be: Priority coefficient = Customer order priority scoring value * 0.4 + Product type urgency scoring value * 0.3 + Process step criticality scoring value * 0.3.

[0046] In step S32, the construction of the dynamic scoring function can be implemented by using the method of linear weighting. For example, the comprehensive score = Wafer delay time * The first weight + Wafer priority coefficient * The second weight + Idle time of the wafer horizontal assembly machine * The third weight. The first weight, the second weight, and the third weight can be adjusted according to the actual production requirements to balance the influence of the delay time, wafer priority, and assembly machine idle time in the scheduling decision.

[0047] Furthermore, step S32 includes: S321: Collect the historical processing data of the wafer horizontal assembly machine, where the historical processing data includes wafer type, processing time, and yield rate; S322: Calculate the average processing time and average yield rate of each wafer horizontal assembly machine for different wafer types; S323: Determine the adaptability scores of the currently to-be-allocated wafer type and each wafer horizontal assembly machine based on the average processing time and average yield rate; S324: Perform normalization processing on the adaptability scores to obtain the assembly machine adaptability coefficient; S325: Construct a dynamic scoring function based on the wafer delay time, the priority coefficient of the wafer, the idle time of the wafer horizontal assembly machine, and the assembly machine adaptability coefficient.

[0048] Among them, in step S321, the historical processing data is periodically collected by the data acquisition module of the assembly machine from each wafer horizontal assembly machine. The data acquisition period can be set to once per hour, or adjusted according to the actual data volume of the production line. The collected historical data should cover at least the processing records of the past week to ensure that the data volume can fully reflect the processing performance of the assembly machine for different types of wafers.

[0049] In step S322, the calculation methods of the average processing time and average yield rate are as follows: for each assembly machine, statistical analysis is performed on the historical processing data of each wafer type. For example, for assembly machine A and wafer type X, all the records of processing wafer type X in the past week are statistically analyzed. The average value of the processing times in these records is calculated as the average processing time of assembly machine A for processing wafer type X, and the average value of the yield rates is calculated as the average yield rate.

[0050] In step S323, the determination method of the adaptability score can be designed to be negatively correlated with the average processing time and positively correlated with the average yield rate. Specifically, the adaptability score can be calculated by a formula. For example, adaptability score = w1 * (1 / average processing time) + w2 * average yield rate, where w1 and w2 are preset weight coefficients used to adjust the relative importance of the average processing time and average yield rate in the adaptability score, and the sum of w1 and w2 is 1.

[0051] In step S324, the normalization processing adopts the maximum-minimum normalization method to linearly map the adaptability scores calculated in step S333 to the interval from 0 to 1. The normalization formula is: assembly machine adaptability coefficient = (adaptability score - minimum value) / (maximum value - minimum value), where the maximum value and minimum value are respectively the maximum value and minimum value among the adaptability scores of all assembly machines for all wafer types.

[0052] In step S325, the dynamic scoring function is constructed as a function that comprehensively considers the wafer delay time, the wafer priority coefficient, the assembler idle time, and the assembler adaptability coefficient. For example, the dynamic scoring function can be designed as: Comprehensive score = a * wafer delay time - b * wafer priority coefficient + c * assembler idle time - d * assembler adaptability coefficient, where a, b, c, and d are preset weight coefficients used to adjust the influence degree of each factor in the comprehensive score. The values of the weight coefficients can be adjusted according to the scheduling requirements and optimization objectives of the actual production line.

[0053] Further, step S323 includes: S3231: Assign weights to the average processing time and the average yield respectively, and calculate the initial adaptability score of each assembler for the currently to-be-allocated wafer type. The sum of the weights of the average processing time and the yield is 1; S3232: Statistically calculate the standard deviations of the average processing time and the yield of all wafer types for each assembler. If the average processing time and the yield corresponding to the currently to-be-allocated wafer type are less than the corresponding standard deviations, then reduce the initial adaptability score of this assembler; conversely, if they are greater than the corresponding standard deviations, then increase the initial adaptability score of this assembler to obtain the final adaptability score.

[0054] Among them, in step S3231, the calculation method of the initial adaptability score is elaborated in detail. Specifically, for the average processing time and the average yield, their respective weight coefficients are preset, and the sum of the two weight coefficients is constant at 1. For example, if more emphasis is placed on processing efficiency, the weight of the average processing time can be set to 0.7, and the weight of the average yield can be set to 0.3. Conversely, if more importance is attached to product quality, the weight allocation can be adjusted. Through this weighted average calculation, the initial adaptability score of each assembler for the currently to-be-allocated wafer type can be obtained.

[0055] In step S3232, the initial adaptability score is further corrected. Specifically, first, the standard deviations of the average processing time and the average yield rate of each assembly machine during the past processing of all wafer types are statistically calculated to evaluate the stability of the processing performance of the assembly machine. Subsequently, the average processing time and the average yield rate corresponding to the current wafer type to be allocated are compared with their respective standard deviations. If the average processing time is less than its standard deviation and the average yield rate is also less than its standard deviation, it means that the performance of the assembly machine in processing this type of wafer may be lower than the average level or fluctuate greatly. Therefore, the initial adaptability score will be reduced. On the contrary, if both the average processing time and the average yield rate are greater than their respective standard deviations, it indicates that the performance of the assembly machine in processing this type of wafer may be better than the average level or more stable. At this time, the initial adaptability score will be increased. By introducing the standard deviation, the accuracy and reliability of the adaptability score are improved, thus providing a more effective reference basis for subsequent task allocation decisions.

[0056] Further, step S4 includes: S41: Real-time monitor the communication link status between the scheduling system and each wafer horizontal assembly machine; The communication link status includes packet loss rate, latency, and connection status; S42: When the communication link status meets any of the following conditions, start the instruction retransmission mechanism: the packet loss rate exceeds the first preset threshold, the latency exceeds the second preset threshold, or the connection status is disconnected. The instruction retransmission mechanism includes: the scheduling system records the scheduling instructions that have been sent but not confirmed and resends them at a preset time interval until the confirmation information from the assembly machine is received.

[0057] Among them, in response to the possible communication link instability problem faced by the scheduling instruction sending, the real-time monitoring of the communication link status and the instruction retransmission mechanism are added. The real-time monitoring of the communication link status is implemented through step S41, aiming to grasp the quality of the current communication link and provide a basis for starting the instruction retransmission mechanism. The monitored communication link status parameters include packet loss rate, latency, and connection status. The packet loss rate can be understood as the proportion of lost data packets during data transmission, reflecting the reliability of network transmission; the latency is the time required for a data packet to travel from the sending end to the receiving end, reflecting the speed of network transmission; the connection status indicates whether the network connection between the scheduling system and the wafer horizontal assembly machine is normal.

[0058] Step S42 is started when the communication link status is abnormal. The specific abnormal situations are set as any one of the conditions that the packet loss rate exceeds the first preset threshold, the delay time exceeds the second preset threshold, or the connection status is disconnected. The first preset threshold and the second preset threshold are specific values set in advance and are used to define the abnormal degree of the packet loss rate and the delay time. The specific content of the instruction retransmission mechanism includes: the scheduling system records the dispatched instructions that have been sent but not confirmed, and re-sends them at a preset time interval until the confirmation information from the assembly machine is received. The preset time interval is the time frequency of instruction retransmission and can be adjusted according to the actual application scenario. The confirmation information of the assembly machine refers to the feedback signal sent by the assembly machine to the scheduling system after receiving the dispatched instruction, indicating that the instruction has been successfully received.

[0059] Please refer to Figure 2 , on the second aspect, a scheduling device for a wafer horizontal assembly machine is applied to the steps of a scheduling method for a wafer horizontal assembly machine as described in any one of the above. The device includes: An arrival time prediction module 201, configured to collect wafer transfer status data in the upstream equipment, predict the predicted arrival time of each wafer at the wafer horizontal assembly machine based on the transfer status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; An assembly machine status monitoring module 202, configured to collect the equipment status data of each wafer horizontal assembly machine and predict the idle time of each wafer horizontal assembly machine after completing the current task; A task assignment module 203, configured to determine the priority of each wafer, and calculate the comprehensive score of the wafer assigned to the corresponding wafer horizontal assembly machine based on the wafer delay time, the priority, and the idle time of the wafer horizontal assembly machine; A task execution module 204, configured to select the wafer with the lowest comprehensive score and the corresponding wafer horizontal assembly machine assignment scheme as the scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

[0060] Among them, the arrival time prediction module 201 can be configured to collect wafer transfer status data in real time from the data interface of the material handling system of the upstream equipment. The transfer status data may include information such as the current position, handling speed, and target assembly machine of the wafer. Based on these data, the prediction algorithm inside the module, such as the Kalman filter or the time series prediction model, can predict the predicted arrival time of the wafer at a specific wafer horizontal assembly machine. The calculation of the wafer delay time can be achieved by comparing the recorded actual arrival time and the predicted arrival time of the wafer.

[0061] The assembly machine status monitoring module 202 can establish a communication connection with the equipment controllers of each wafer horizontal assembly machine, and periodically or in real-time collect equipment status data. The equipment status data may include the operating status of the equipment, the task type and task volume of the currently executing task, the processed time, etc. The module can maintain a correspondence table between task types and standard processing times, and combine the task volume and processed time of the current task to predict the remaining time for the assembly machine to complete the current task, and then obtain the idle time of the assembly machine.

[0062] The task assignment module 203 can first determine the priority of each wafer to be scheduled according to a preset priority evaluation model. The priority evaluation model can comprehensively consider factors such as the priority of customer orders, the urgency of product types, and the criticality of process steps. Then, the module can construct a dynamic scoring function that takes the wafer delay time, the wafer priority coefficient, and the idle time of the wafer horizontal assembly machine as inputs and calculates the comprehensive score for allocating the wafer to different assembly machines. The design goal of the dynamic scoring function is to enable wafers with longer delay times, higher priorities, and that can be processed by idle assembly machines earlier to obtain lower comprehensive scores.

[0063] The task execution module 204, after the task assignment module calculates the comprehensive scores of all possible wafer-assembly machine allocation schemes, selects the allocation scheme with the lowest comprehensive score and generates the corresponding scheduling instruction. The scheduling instruction may include information such as the wafer ID, the target assembly machine ID, and processing parameters. The module sends the scheduling instruction to the equipment controller of the target wafer horizontal assembly machine in real-time through a network communication protocol, such as the TCP / IP or industrial Ethernet protocol, to control the assembly machine to execute the corresponding scheduling task.

[0064] In this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations.

[0065] The above are only embodiments of the present application and are not used to limit the protection scope of the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A scheduling method for a horizontal wafer assembly machine, used in an automated wafer manufacturing production line, characterized in that, The method includes the steps of: S1: Collect the wafer transfer status data in the upstream equipment, predict the predicted arrival time of each wafer at the wafer horizontal assembly machine based on the transfer status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; S2: Collect the equipment status data of each wafer horizontal assembly machine, and predict the idle time of each wafer horizontal assembly machine after completing the current task; S3: Determine the priority of each wafer, and calculate the comprehensive score of the wafer assigned to the corresponding wafer horizontal assembly machine based on the wafer delay time, the priority, and the idle time of the wafer horizontal assembly machine; S4: Select the wafer with the lowest comprehensive score and the corresponding wafer horizontal assembly machine allocation scheme as the scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

2. The scheduling method of a horizontal wafer assembly machine according to claim 1, wherein Step S1 includes: S11: Collect the wafer transfer status data in the upstream equipment, and extract the handling path of the wafer, the speed of the handling equipment, and the traffic congestion situation of the handling system; S12: Predict the reference handling time of the wafer under the current handling path according to the handling path; S13: Obtain the speed data of the handling equipment in real time, calculate the deviation value between the current handling speed and the historical average handling speed, and adjust the reference handling time according to the deviation value to obtain the corrected predicted arrival time; S14: Obtain the predicted delay time according to the traffic congestion situation of the handling system, and obtain the predicted arrival time of the final wafer at the wafer horizontal assembly machine according to the corrected predicted arrival time and the predicted delay time; S15: Record the actual arrival time of the wafer at the wafer horizontal assembly machine, and calculate the difference between the actual arrival time and the predicted arrival time to obtain the wafer delay time.

3. The scheduling method of a horizontal wafer assembly machine according to claim 2, characterized in that, Step S14 includes: S141: Monitor the traffic congestion situation of the handling system, count the number of handling equipment passing through the key node per unit time, and if the number exceeds the preset threshold, it is determined to be congested; S142: When the traffic congestion situation of the handling system is congested, determine the congestion degree, and obtain the predicted delay time according to the congestion degree. The sum of the corrected predicted arrival time and the predicted delay time is the predicted arrival time.

4. The scheduling method of a horizontal wafer assembly machine according to claim 3, characterized in that, Step S142 includes: S1421: When the traffic congestion situation of the handling system is congested, record the congestion duration; the congestion duration reflects the congestion degree; S1422: Query the congestion delay time table, and determine the predicted delay time of the wafer at the congested key node according to the congestion duration; the congestion delay time table records the corresponding relationship between the congestion duration and the predicted delay time; S1423: Accumulate the predicted delay times of the wafer at each congested key node to obtain the total congestion delay time, and add the total congestion delay time to the predicted arrival time to obtain the final predicted arrival time.

5. The scheduling method of a horizontal wafer assembly machine according to claim 1, characterized in that, Step S2 includes: S21: Collect the equipment status data of each wafer horizontal assembly machine, and read the task type and task volume of the current task being executed according to the equipment status data; S22: Query the pre-set correspondence table between task types and standard processing times according to the task type, and determine the standard processing time corresponding to the current task; S23: Calculate the remaining processing time of the current task based on the task volume and the standard processing time; S24: Based on the remaining processing time, predict the completion time of the current task by the wafer horizontal assembly machine, and subtract the current time from the completion time to obtain the idle time of each wafer horizontal assembly machine after completing the current task.

6. The scheduling method of a horizontal wafer assembly machine according to claim 1, characterized in that Step S3 includes: S31: Obtain the customer order priority, product type urgency, and process step criticality corresponding to each wafer, construct a wafer priority evaluation model, determine the priority of the wafer according to the wafer priority evaluation model, and obtain the priority coefficient of the wafer; S33: Construct a dynamic scoring function based on the wafer delay time, the priority coefficient of the wafer, and the idle time of the wafer horizontal assembly machine; S34: Calculate the comprehensive score assigned to each corresponding wafer horizontal assembly machine for each wafer according to the dynamic scoring function.

7. A scheduling method for a wafer horizontal assembly machine according to claim 6, characterized in that, Step S32 includes: S321: Collect the historical processing data of the wafer horizontal assembly machine, where the historical processing data includes wafer type, processing time, and yield rate; S322: Calculate the average processing time and average yield rate of each wafer horizontal assembly machine for different wafer types; S323: Determine the adaptability score between the currently to-be-allocated wafer type and each wafer horizontal assembly machine according to the average processing time and the average yield rate; S324: Normalize the adaptability score to obtain the assembly machine adaptability coefficient; S325: Construct a dynamic scoring function based on the wafer delay time, the priority coefficient of the wafer, the idle time of the wafer horizontal assembly machine, and the assembly machine adaptability coefficient.

8. A scheduling method for a horizontal wafer assembly machine according to claim 7, characterized in that, Step S323 includes: S3231: Assign weights to the average processing time and the average yield rate respectively, and calculate the initial adaptability score of each assembly machine for the currently to-be-allocated wafer type, where the sum of the weights of the average processing time and the yield rate is 1; S3232: Statistically calculate the standard deviations of the average processing time and yield rate of all wafer types for each assembly machine. If the average processing time and yield rate corresponding to the currently to-be-allocated wafer type are less than the corresponding standard deviations, reduce the initial adaptability score of this assembly machine. Conversely, if they are greater than the corresponding standard deviations, increase the initial adaptability score of this assembly machine to obtain the final adaptability score.

9. A scheduling method for a horizontal wafer assembly machine according to claim 1, characterized in that, Step S4 includes: S41: Real-time monitor the communication link status between the scheduling system and each wafer horizontal assembly machine; the communication link status includes packet loss rate, delay time, and connection status; S42: When the communication link status meets any of the following conditions, start the instruction retransmission mechanism: the packet loss rate exceeds the first preset threshold, the delay time exceeds the second preset threshold, or the connection status is disconnected. The instruction retransmission mechanism includes: the scheduling system records the scheduling instructions that have been sent but not confirmed, and re-sends them at a preset time interval until the confirmation information from the assembly machine is received.

10. A scheduling device for a horizontal wafer assembly machine, characterized in that, In the steps of a scheduling method applied to a wafer horizontal assembly machine according to any one of the above claims 1-9, the device includes: An arrival time prediction module, configured to collect wafer transfer status data in an upstream device, predict the predicted arrival time of each wafer at the wafer horizontal assembly machine based on the transfer status data, and calculate the wafer delay time between the actual arrival time and the predicted arrival time; An assembly machine status monitoring module, configured to collect the device status data of each wafer horizontal assembly machine and predict the idle time of each wafer horizontal assembly machine after completing the current task; A task allocation module, configured to determine the priority of each wafer, and calculate the comprehensive score of the wafer assigned to the corresponding wafer horizontal assembly machine based on the wafer delay time, the priority, and the idle time of the wafer horizontal assembly machine; A task execution module, configured to select the wafer with the lowest comprehensive score and the allocation scheme of the corresponding wafer horizontal assembly machine as a scheduling instruction, and send the scheduling instruction to each wafer horizontal assembly machine for execution in real time.

Images