Task flow dynamic optimization method and system for data feedback

Abstract

The invention provides a task flow dynamic optimization method and system based on data feedback, and the method comprises the steps: obtaining process feedback data, determining the process execution time consumption and a processing result through a production monitoring system log, recognizing the current processing capability of a process, and obtaining the to-be-processed workpiece amount of a downstream assembly node and the frequency fluctuation amplitude of an upstream process output workpiece; acquiring the queuing workpiece quantity and the real-time load quantity of each process corresponding to the rearranged process processing execution sequence, analyzing to obtain a process processing capacity matching degree index, and determining a global workpiece circulation stability evaluation result; and obtaining total number change data of workpieces completed in each process from the dynamic optimization implementation scheme, judging the system processing throughput improvement degree, and determining the process load balancing state of workpiece circulation.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to a task flow dynamic optimization method and system based on data feedback. BACKGROUND

[0002] Task flow dynamic optimization is a key research field in modern complex systems to improve efficiency and resource utilization, especially in high concurrency and high dynamic task processing. The stability and efficiency of task flow optimization directly determine the overall performance and response capability of the system. Task flow optimization not only concerns the processing speed of a single node, but also involves the coordination and balance of the entire system under dynamic changes. Existing methods often focus on improving the performance of a single node in task flow optimization, such as increasing resource investment or optimizing algorithms to improve the processing efficiency of a certain node. However, these methods usually ignore the interdependence between nodes, resulting in local optimization that may cause instability in the overall system. For example, after a node adjusts its processing speed through feedback data, it may cause downstream nodes to overload due to an influx of tasks, or disrupt the output rhythm of upstream nodes. This imbalance between nodes makes the system sacrifice overall coordination in pursuit of local optimization. The dynamic nature of task flow is reflected in the constant changes in real-time load, queued task number, and task completion of each node, and these changes will quickly spread due to local adjustments of a node, triggering a chain reaction. Global balance adjustment requires real-time acquisition of the running state of each node, and dynamic adjustment of the execution order and resource allocation ratio of the nodes when local optimization causes system imbalance. However, existing technologies are difficult to quickly establish an effective coordination mechanism between nodes, especially when feedback data between nodes is complex and changes rapidly, making it difficult to accurately capture and process. For example, in an intelligent manufacturing system, a machine in a certain process optimizes its processing speed through feedback data, but the downstream assembly node forms a bottleneck due to an influx of tasks that cannot be processed in time, and the upstream raw material supply node frequently stops due to mismatched rhythm. Therefore, how to dynamically coordinate the execution order and resource allocation ratio of each node while the nodes adjust themselves based on real-time feedback to avoid local optimization causing global task flow congestion or imbalance becomes a key problem in task flow dynamic optimization. SUMMARY

[0003] The present application provides a task flow dynamic optimization method based on data feedback, the method comprising: obtaining process feedback data, quantity of workpieces to be processed, and fluctuation amplitude of output workpiece frequency, wherein the process feedback data includes equipment processing time, workpiece completion status, and equipment usage frequency, the quantity of workpieces to be processed represents the accumulation of workpieces at the downstream assembly node, and the fluctuation amplitude of output workpiece frequency represents the change in output rhythm of the upstream process; determine a process load change trend according to a device utilization rate and a capacity occupation proportion, adjust a workpiece flow rate based on a comparison result of the workpiece quantity to be processed and a preset assembly threshold, and identify a potential process blocking position list and a queued workpiece quantity in combination with a fluctuation amplitude of the output workpiece frequency; extract a key process identifier and a blocking type from the potential process blocking position list, and adjust a device allocation scheme in combination with real-time load data of a device, real-time load data of a capacity, and the queued workpiece quantity, wherein the device allocation scheme comprises a device number allocation and a capacity capacity allocation; rearrange a process execution order according to a process device proportion in the device allocation scheme and a total quantity of completed workpieces, in combination with a workpiece priority and a process dependency relationship; obtain a queued workpiece quantity and a real-time load according to the process execution order, determine a process processing capacity matching degree index and a global workpiece flow stability evaluation result; adjust the device allocation scheme and the process execution order based on the key process identifier, the blocking type, and the global workpiece flow stability evaluation result, and determine a dynamic optimization implementation scheme; obtain total quantity change data of completed workpieces according to the dynamic optimization implementation scheme, and determine a process load balancing state.

[0004] Further, the process feedback data, the workpiece quantity to be processed, and the output workpiece frequency fluctuation amplitude are obtained as follows: extract running logs of each process device from a production monitoring system, calculate device processing time consumption and workpiece stay duration based on time stamps in the running logs, and determine a process processing capacity coefficient in the process feedback data; screen low processing capacity process nodes according to the process processing capacity coefficient, count the workpiece quantity to be processed, monitor upstream process output workpiece quantity changes, and determine the output workpiece frequency fluctuation amplitude based on output quantity differences of adjacent time periods.

[0005] Further, the process load change trend is determined according to a device utilization rate and a capacity occupation proportion, the workpiece flow rate is adjusted based on a comparison result of the workpiece quantity to be processed and a preset assembly threshold, and the potential process blocking position list and the queued workpiece quantity are identified in combination with the output workpiece frequency fluctuation amplitude, as follows: collect the device utilization rate and the capacity occupation proportion of each process device to form a time sequence, calculate a change rate based on the time sequence, and determine the process load change trend; when the workpiece quantity to be processed exceeds the preset assembly threshold, send a flow rate limiting instruction to an upstream process, and adjust a workpiece release frequency; Calculate the coefficient of variation based on the output time interval of the upstream process after the flow restriction instruction, determine the output rhythm disorder process, generate the potential process blocking position list combining the process load change trend and the coefficient of variation, the potential process blocking position list contains process number and queuing workpiece amount.

[0006] Further, the coefficient of variation is calculated based on the output time interval of the upstream process after the flow restriction instruction, and the output rhythm disorder process is determined, which includes: Collect the output time of the upstream process after the flow restriction instruction is executed, and calculate the coefficient of variation of the adjacent output time interval sequence; When the coefficient of variation exceeds the preset threshold, mark the rhythm disorder process, count the queuing workpiece amount before the rhythm disorder process, and generate a list containing process number and blocking degree.

[0007] Further, the key process identifier and blocking type are extracted from the potential process blocking position list, and the equipment allocation scheme is adjusted combining the real-time load data of the equipment, the real-time load data of the capacity, and the queuing workpiece amount, which includes: Extract the key process identifier and blocking type in the potential process blocking position list, obtain the real-time load data of the equipment and the real-time load data of the capacity, and calculate the ratio of the queuing workpiece amount to the buffer capacity as the queuing saturation degree; According to the weighted sum of the real-time load data of the equipment, the real-time load data of the capacity and the queuing saturation degree, determine the resource shortage degree score, adjust the equipment number and the upper limit of the capacity based on the resource shortage degree score and the blocking type, and form the equipment allocation scheme.

[0008] Further, the process execution order is rearranged combining the workpiece priority and the process dependency relationship according to the process equipment ratio in the equipment allocation scheme and the total amount of completed workpieces, which includes: Extract the process equipment ratio from the equipment allocation scheme, count the total amount of completed workpieces corresponding to the process equipment ratio, and determine the downward trend process based on the time series slope; According to the workpiece priority value and the process dependency relationship extracted from the downward trend process, construct a directed graph, generate a feasible process sequence based on topological sorting, and select the sequence with the minimum comprehensive evaluation value as the process execution order.

[0009] Further, the queuing workpiece amount and real-time load are obtained according to the process execution order, the process processing capacity matching degree index and the global workpiece flow stability evaluation result are determined, which includes: Obtain the queuing workpiece amount and real-time load under the process execution order, determine the process processing capacity value based on the comparison between the actual processing rate and the standard processing rate; According to the processing capacity value, the capacity difference of the adjacent process is calculated, the mismatch process ratio is counted, and the global workpiece flow stability evaluation result is determined by combining the buffer occupancy rate weighted calculation.

[0010] Further, the device allocation scheme and process execution order are adjusted based on the key process identifier, blocking type and global workpiece flow stability evaluation result to determine a dynamic optimization implementation scheme, including: The current device allocation ratio and the process execution order are obtained, the process information corresponding to the key process identifier is extracted, and the capacity bottleneck process and resource surplus process are identified based on the blocking type and global workpiece flow stability evaluation result; Based on the capacity bottleneck process and resource surplus process, the device allocation ratio and process execution order are adjusted using a linear programming method to form the dynamic optimization implementation scheme, which includes the number of device adjustments and the position of execution order adjustment, wherein: The device allocation ratio and process execution order are adjusted based on the capacity bottleneck process and resource surplus process using a linear programming method to form the dynamic optimization implementation scheme, including: The blocking type and workpiece stagnation time corresponding to the key process identifier are extracted, the number of transmission delays is counted, the delay serious process is marked, and the number of device gaps is determined; Based on the number of device gaps and process dependency relationship, the process execution order is adjusted, the device transfer path and relocation schedule are formulated, and the dynamic optimization implementation scheme including device allocation arrangement and process priority rearrangement is formed.

[0011] Further, the total number of completed workpieces change data is obtained according to the dynamic optimization implementation scheme to determine the process load balancing state, including: The total number of completed workpieces change data is extracted from the dynamic optimization implementation scheme, the change rate and system processing throughput are calculated; Based on the change rate, the load coefficient of each process is calculated, the standard deviation of the load coefficient is counted, the overloaded and idle processes are marked based on the load coefficient distribution, and the process load balancing state is determined.

[0012] In another aspect, the present application also discloses a data feedback task flow dynamic optimization system, comprising: a feedback data acquisition and processing capability identification module, configured to acquire process feedback data, a quantity of workpieces to be processed, and a fluctuation amplitude of an output workpiece frequency, wherein the process feedback data comprises a device processing time, a workpiece completion state, and a device usage frequency, the quantity of workpieces to be processed represents a workpiece accumulation quantity of a downstream assembly node, and the fluctuation amplitude of the output workpiece frequency represents a change in an upstream process output rhythm; a load evaluation and blockage identification module, configured to determine a process load change trend according to a device usage rate and a production capacity occupancy ratio, adjust a workpiece flow based on a comparison result of the quantity of workpieces to be processed and a preset assembly threshold, and identify a list of potential process blockage positions and a queued workpiece quantity in combination with the fluctuation amplitude of the output workpiece frequency; a key process analysis and device allocation adjustment module, configured to extract a key process identifier and a blockage type from the list of potential process blockage positions, and adjust a device allocation scheme in combination with real-time load data of a device, real-time load data of a production capacity, and the queued workpiece quantity, wherein the device allocation scheme comprises a device quantity allocation and a production capacity allocation; a device allocation scheme evaluation and process sequence rearrangement module, configured to rearrange a process execution sequence based on a process device ratio and a total quantity of completed workpieces in the device allocation scheme, in combination with a workpiece priority and a process dependency relationship; a processing capability matching degree analysis module, configured to acquire a queued workpiece quantity and a real-time load according to the process execution sequence, determine a process processing capability matching degree index and a global workpiece flow stability evaluation result; a dynamic optimization scheme generation module, configured to adjust the device allocation scheme and the process execution sequence based on the key process identifier, the blockage type, and the global workpiece flow stability evaluation result, and determine a dynamic optimization implementation scheme; and a throughput evaluation and load balancing judgment module, configured to acquire total quantity of completed workpiece change data according to the dynamic optimization implementation scheme, and determine a process load balancing state. The application discloses a data feedback task flow dynamic optimization method and system, aiming at the problems of process blockage, uneven device allocation, and unstable workpiece flow in a production process, by collecting production monitoring system logs in real time, analyzing process execution time consumption, processing results, and upstream and downstream workpiece flow data, identifying key process blockage positions and types, evaluating process processing capability matching degrees and global flow stability. BRIEF DESCRIPTION OF DRAWINGS

[0013] Fig. 1 A flow chart of a data feedback task flow dynamic optimization method of the present application.

[0014] Fig. 2 A structural schematic diagram of a data feedback task flow dynamic optimization system of the present application. DETAILED DESCRIPTION

[0015] The technical solutions in the embodiments of the present application will be described clearly and in detail below with reference to the drawings in the embodiments of the present application. The described embodiments are only some of the embodiments of the present application.

[0016] As Figs. 1-2 , the data feedback task flow dynamic optimization method and system can specifically include: Step S101, obtaining process feedback data, determining process execution time consumption and processing results through a production monitoring system log, identifying current processing capacity of the process, and simultaneously obtaining the amount of workpieces to be processed at a downstream assembly node and the fluctuation amplitude of the output workpiece frequency of an upstream process.

[0017] The real-time running log of each process equipment in the production monitoring system is obtained, the start-up timestamp and the shutdown timestamp of each equipment are extracted from the log, the actual processing time consumption of the equipment is calculated according to the timestamp difference, the time when the workpiece enters the process and the time when the processing is completed are read, the residence time of a single workpiece in the current process is determined, the residence time is compared with a preset standard process time, and a process processing capacity coefficient is obtained. According to the process processing capacity coefficient, the process node with a processing capacity lower than a preset threshold is screened out, the accumulation amount of workpieces to be processed at the downstream assembly node is counted through a workshop data acquisition terminal, the number of workpieces output by the upstream process per unit time is monitored, the absolute value of the difference of the output number in adjacent time periods is calculated, and when the absolute value of the difference exceeds a preset fluctuation threshold, it is determined that the high fluctuation state exists, and an upstream process output workpiece frequency fluctuation amplitude evaluation value is obtained.

[0018] In an embodiment, the production monitoring system obtains running data in real time through data acquisition equipment deployed on each process equipment, records the start-up and shutdown timestamps of the equipment to calculate the actual processing time consumption. The entering and completing times are recorded through automatic identification technology when the workpiece is transferred between processes, and the residence time of a single workpiece is calculated.

[0019] Specifically, the preset standard process duration is determined according to the standard processing time specified in the product process card, for example, the standard duration of the drilling process of a certain part is 180 seconds. When the actual residence duration is 200 seconds, the process handling capacity coefficient is calculated as 180÷200=0.9, indicating that the actual handling capacity of the process is 90% of the standard capacity. By setting the preset threshold value to 0.85, the process nodes with a handling capacity coefficient lower than 0.85 are screened out, which are the bottleneck processes that need to be monitored. The workshop data acquisition terminal counts the number of workpieces to be processed on the conveyor belt of the downstream assembly node through an optical sensor, and triggers a warning when the accumulation amount exceeds 70% of the buffer capacity.

[0020] It should be noted that the fluctuation amplitude of the output workpiece frequency of the upstream process is evaluated by counting the output quantity every 5 minutes, and the absolute value of the difference between the output quantities of two adjacent time periods is the fluctuation value. The amount of workpieces to be processed at the downstream assembly node and the fluctuation amplitude of the output workpiece frequency of the upstream process are obtained, and the fluctuation amplitude includes two states of low-frequency stable fluctuation and high-frequency abnormal fluctuation. When the fluctuation value exceeds 20% of the average output quantity, it is determined to be in a high fluctuation state, and the obtained fluctuation amplitude evaluation value is used for subsequent adjustment of the coordination between processes.

[0021] Step S102, according to the process load change trend recorded by the equipment utilization rate and the capacity occupation ratio, if the amount of workpieces to be processed at the downstream assembly node exceeds the preset assembly threshold value, the workpiece flow is limited, and the output workpiece frequency fluctuation amplitude is evaluated to assess the degree of output rhythm disorder of the upstream process, and a list of potential process blocking positions and the amount of queued workpieces of each process are identified.

[0022] The use rate data and capacity occupancy ratio data of each process equipment in the past time period are acquired, time series are formed by collecting data every fixed time interval, the ratio of the difference between adjacent data points in the time series to the time interval is calculated to obtain the change rate, the load rising or falling trend of the process is judged according to the positive or negative value of the change rate, and the real-time workpiece quantity to be processed of the downstream assembly node is read, if the workpiece quantity to be processed exceeds the preset assembly threshold, a flow restriction instruction is sent to the upstream process to reduce the workpiece release frequency to a preset proportion of the original frequency. After the flow restriction instruction is executed, the workpiece output time interval of each process in the upstream is monitored, and the stability index of the output rhythm is calculated by statistical analysis method, when the stability index exceeds the preset threshold, it is judged that the output rhythm of the process is disorderly, and the disorderly process number and disorder degree value are recorded. Combined with the process load change trend and the disorder degree value, the ratio of the average arrival rate to the average service rate of each process is calculated as the utilization rate index, when the utilization rate index is greater than the preset critical value, it is marked as a potential blocking process, the queuing workpiece quantity before the potential blocking process is counted, and a process blocking position list is generated according to the queuing workpiece quantity, the utilization rate index and the disorder degree value, the list includes the process number, the blocking severity score and the current queuing workpiece quantity.

[0023] In an embodiment, the production monitoring system collects equipment use rate and capacity occupancy ratio data every 5 minutes through data collection terminals deployed at each process to form a continuous time series.

[0024] Specifically, the equipment use rate refers to the ratio of the actual running time to the planned running time, and the capacity occupancy ratio refers to the ratio of the actual output to the rated capacity. By calculating the data difference between the adjacent two collection time points divided by the time interval, the change rate value is obtained. When the change rate is positive, it indicates that the process load is rising; when the change rate is negative, it indicates that the process load is falling.

[0025] It should be noted that the preset assembly threshold of the downstream assembly node is usually set to 80% of the assembly buffer capacity. When it is monitored that the workpiece quantity to be processed exceeds this threshold, the system automatically sends a flow restriction instruction to each process in the upstream to reduce the workpiece release frequency to 70% of the original frequency. This dynamic control mechanism can avoid the blocking of the production line caused by excessive accumulation of workpieces at the downstream node.

[0026] Preferably, after executing the flow restriction, the system continuously collects the workpiece output time of each process in the upstream at 10 time points to form a time interval sequence. The coefficient of variation is obtained by calculating the ratio of the standard deviation to the average value of the time interval sequence, which can eliminate the dimension influence and objectively reflect the relative fluctuation degree of the output rhythm. When the coefficient of variation of a certain process exceeds 0.3, it is judged that the output rhythm of the process is disorderly, and the system records the process number and the specific coefficient of variation value.

[0027] In a possible implementation, the utilization index is calculated by comprehensively considering the dynamic load state of the process. The average arrival rate is obtained by counting the number of workpieces arriving at the process per unit time, and the average service rate is obtained by counting the number of workpieces completed by the process per unit time. When the utilization index exceeds 0.95, it indicates that the process is close to a saturated state and is prone to form a blockage. The blockage severity score is calculated by using a weighted method, in which the queue workpiece amount accounts for 50% of the weight, the utilization index accounts for 30% of the weight, and the turbulence degree value accounts for 20% of the weight, and a score value of 0-100 is obtained after normalization processing.

[0028] For example, the utilization index of a certain stamping process reaches 0.98, the front queue workpiece amount is 45, and the coefficient of variation is 0.35. After weighted calculation, the blockage severity score is 85, which belongs to a serious blockage state. The system records the process information in the blockage position list to provide data support for subsequent dynamic optimization adjustment.

[0029] The change amplitude of the output workpiece quantity per unit time of the upstream process is monitored, and when the change amplitude is greater than a preset change threshold, the process is marked as a rhythm turbulence process. The number of workpieces currently being processed and the accumulation amount of workpieces to be processed of each process are checked, the total number of workpieces waiting in the queue in front of each process is counted, the process position whose queue workpiece amount exceeds the normal processing capacity is identified, the process whose device utilization rate exceeds the saturated state and whose queue workpiece amount continuously increases is marked as a potential blockage point, and a position list containing the process name and the blockage is established.

[0030] The output workpiece quantity of each process at a preset time interval is monitored, the difference value between the output quantities of adjacent time intervals and the output quantity of the previous time interval is calculated as the change amplitude, and if the change amplitude exceeds a preset change threshold, the process is marked as a rhythm turbulence process. According to the marking of the rhythm turbulence process, the number of workpiece sensor signals being processed in the device of each process is read, the accumulation amount of workpieces to be processed in the process buffer is counted, the total number of workpieces waiting in the queue in front of the process is obtained by adding the number of workpieces being processed and the accumulation amount, and when the total number of workpieces exceeds the rated processing capacity of the process, it is determined to be in an overload state. The real-time data of the device utilization rate and the historical queue workpiece quantity change sequence of the process in the overload state are obtained, and if the device utilization rate exceeds a preset saturation threshold and the total number of queue workpieces continuously increases in a continuous monitoring period, the process is marked as a potential blockage point. The blockage type is determined according to the rhythm turbulence marking, the overload state, and the device saturation state, and a position list containing the process name and the blockage type is established.

[0031] In an embodiment, the production monitoring system counts the number of workpieces passing through every 5 minutes by installing a photoelectric counter at the outlet of each process.

[0032] Specifically, the system records the output quantity in the Nth time interval as Qn, the output quantity in the N-1th time interval as Qn-1, and the change amplitude as |Qn-Qn-1| / Qn-1. When the ratio exceeds 0.25, it indicates that the process output rhythm fluctuates significantly, and the system automatically marks the process as a rhythm disorder process.

[0033] It should be noted that the workpieces being processed in the process equipment are detected by pressure sensors or position sensors installed inside the equipment.

[0034] For example, the pressure sensors in the punch press can sense the presence of workpieces, and each sensor signal corresponds to a workpiece being processed. The accumulation amount of the process buffer zone is counted by an infrared reflection sensor array, and the sensor array is arranged at a certain interval in the buffer zone. The number of blocked sensors reflects the number of accumulated workpieces.

[0035] Preferably, the rated processing capacity is determined according to the design capacity of the process and the physical capacity of the buffer zone. The rated processing capacity of a certain drilling process is 30 pieces, and when the total number of queued workpieces reaches 35 pieces, the system determines that the process is in an overload state. This overload determination mechanism can identify potential risks in advance before the process is completely blocked.

[0036] In one possible implementation, the preset saturation threshold of equipment utilization is set to 0.9, indicating that equipment utilization above 90% is considered saturated. The system analyzes the change in the number of queued workpieces in the last 10 monitoring periods, and if more than 8 periods show a growing trend, it is determined to be continuously increasing. The blocking types are mainly divided into three categories: rhythm imbalance type blocking, which is caused by the blocking of upstream output rhythm disorder; capacity shortage type blocking, which is caused by the long-term saturation of the equipment; and cumulative type blocking, which is caused by the continuous accumulation of queued workpieces that cannot be digested.

[0037] For example, the position list is stored in a structured data form, including fields such as process number, process name, blocking type, current queue number, equipment utilization, rhythm disorder mark, etc. By establishing such a list, the production scheduling system can quickly locate the bottleneck process on the production line and achieve precise capacity allocation and process optimization.

[0038] Step S103, extract the key process identifier and blocking type from the potential process blocking position list, obtain the real-time load data of each process equipment and the real-time load data of the capacity through the production scheduling center, evaluate the current equipment occupancy ratio of each process by combining the queued workpiece quantity, and adjust the equipment allocation.

[0039] The key process identifier and the corresponding blockage type code are extracted from the potential process blockage position list, the real-time load data of the equipment and the real-time load data of the capacity of the key process are obtained through the data interface of the production scheduling center, the current number of queued workpieces of each process is read, and the ratio of the number of queued workpieces to the capacity of the process buffer is calculated as the queuing saturation degree. According to the real-time load data of the equipment, the real-time load data of the capacity and the queuing saturation degree, the resource shortage degree score of each process is obtained by multiplying the three data by the corresponding preset weight coefficient and summing them up, wherein the sum of the weight coefficients of the equipment load, the capacity load and the queuing saturation degree is 1. The resource shortage process and the resource surplus process are identified based on the resource shortage degree score, the optimal scheme of transferring equipment from the resource surplus process to the resource shortage process is determined according to the blockage type code, the optimal scheme of transferring equipment includes the number of allocated equipment and the capacity of the capacity, the equipment number transfer plan is obtained, and the upper limit of the capacity of each process is adjusted according to the blockage type code and the resource shortage degree score. The equipment number transfer plan and the capacity upper limit adjustment value are integrated to form an equipment allocation scheme including an equipment reallocation list and a capacity adjustment parameter table, and the scheme specifies the new equipment configuration number and the maximum capacity output value allowed for each process.

[0040] In an embodiment, the production scheduling center maintains a real-time updated potential process blockage position list, which records the process information of all processes at risk of blockage. The system extracts key process identifiers from the list, each identifier corresponding to a unique process number, and extracts blockage type codes, including equipment failure type, capacity deficiency type and material shortage type. Through the data bus of the scheduling center, the system collects real-time equipment load data of each process, which reflects the current operation intensity of the equipment; at the same time, it obtains capacity load data, which represents the ratio of actual output to rated capacity. The calculation of queuing saturation degree is based on the physical capacity limit of the process buffer. When the number of queued workpieces approaches the buffer capacity, the saturation degree value approaches 1, indicating that the process has approached the blockage critical state. By multiplying each indicator by the corresponding weight and summing them up, a resource shortage degree score value between 0 and 100 is obtained, and the higher the score, the more resource-intensive the process.

[0041] It should be noted that the system determines the optimal allocation scheme by evaluating the equipment transfer benefits between processes. The transfer benefits take into account the resource shortage degree of the target process, the resource surplus degree of the source process and the priority of the blockage type.

[0042] For example, two devices are transferred from the surplus deburring process to the tight drilling process, effectively relieving the production bottleneck.

[0043] In one possible implementation, the adjustment of the upper limit of the capacity is performed by using a dynamic threshold mechanism. The system determines the adjustment range according to the congestion type code and the resource shortage degree score. For the process of the equipment failure type congestion, the upper limit of the capacity is reduced to 70% of the original value to avoid the overload operation of the equipment. For the process of the capacity deficiency type congestion, the upper limit of the capacity can be increased to 120% of the original value by adding equipment. For the process of the material shortage type congestion, the capacity is maintained unchanged, and the problem of material supply is solved.

[0044] Exemplarily, the final formation of the equipment allocation scheme includes two core components. The equipment reallocation list records the equipment change of each process, including the number of equipment transferred in, the number of equipment transferred out, and the final configuration number. The capacity adjustment parameter table records the upper limit of the capacity after adjustment, the recommended running rate, and the buffer capacity configuration of each process.

[0045] It can be understood that after the implementation of the scheme, the dynamic balance of the production line resources can be achieved. The system can complete the resource reconfiguration of the whole production line within 10 minutes, improve the overall equipment utilization rate from 85% to 92%, and reduce the number of blocked processes by more than 60%. Further, the system also establishes a feedback mechanism for the implementation of the scheme, and monitors the running state of each process after adjustment in real time. When a new resource imbalance phenomenon is found, the next round of optimization adjustment process is automatically triggered.

[0046] In step S104, the proportion of equipment in the equipment allocation scheme and the corresponding total amount of completed workpieces are identified. If the total amount of completed workpieces corresponding to the proportion of equipment in the process shows a downward trend, the execution order of the process is rearranged according to the workpiece priority and the process dependency relationship.

[0047] The device proportion value of each process in the device allocation scheme is identified, the total amount of completed workpieces in the past multiple monitoring periods corresponding to the device proportion value is counted, the time series slope of the total amount of completed workpieces is calculated by linear regression, if the slope is negative, it is determined that there is a downward trend, and the process number with a downward trend is recorded. According to the process number with a downward trend, the workpiece priority value related to these processes is extracted from the production management database, the priority value is obtained by comprehensive evaluation according to the delivery time urgency, customer importance and product profit rate, and the dependency relationship between processes is obtained, and a directed graph structure representing the constraint between processes is constructed. Based on the workpiece priority value and the directed graph of process dependency relationship, a feasible process sequence set satisfying the dependency constraint is generated, for each feasible sequence, the product sum of the expected completion time of all workpieces and the corresponding priority value is calculated as the comprehensive evaluation value of the sequence. From the feasible process sequence set, the sequence with the minimum comprehensive evaluation value is selected as the optimized process execution order, the position of the process with a downward trend in the new order is adjusted, and a rearranged process execution order table is formed, which specifies the execution priority of each process.

[0048] In an embodiment, the device allocation scheme records the ratio of the number of devices configured for each process to the total number of devices in the production line, i.e. the device proportion value. The system counts the total amount of completed workpieces for each process every hour, records data for 24 monitoring periods continuously, and forms a time series data set. Linear regression fitting is performed by least squares method to calculate the slope value of the total amount of completed workpieces with respect to time. When the slope is negative and the absolute value exceeds a preset threshold, it indicates that the production efficiency of the process presents a downward trend, and the system automatically marks it as a target process that needs to be optimized.

[0049] Specifically, the linear regression model takes time as the independent variable and the total amount of completed workpieces as the dependent variable, and determines the optimal fitting straight line by minimizing the squared error between the predicted value and the actual value. The slope reflects the change rate of the total amount of completed workpieces per unit time, and a negative slope means that the production capacity is gradually decreasing. The completion amount of a certain stamping process decreases from 120 pieces per hour to 95 pieces in 24 hours, and the linear regression slope is -1.04, which obviously presents a downward trend and needs to be adjusted.

[0050] It should be noted that the workpiece priority evaluation adopts multi-factor comprehensive evaluation.

[0051] For example, a batch of auto parts is finally determined to have a priority value of 73.5 according to the comprehensive evaluation of the delivery urgency, customer importance and product profit rate, and the processing is arranged in priority in the process rearrangement. Preferably, the construction of the directed graph of process dependency relationship is based on the constraint conditions of the production process flow. Each node in the graph represents a process, and the directed edge represents the precedence constraint relationship between processes. If process A must be completed before process B, a directed edge is drawn from node A to node B. The system automatically extracts the precedence and postcondition relationships between processes by parsing the process route file, and constructs a complete dependency relationship graph. The processing flow of a certain product includes five processes of blanking, turning, milling, drilling and assembly, among which blanking must be performed first, turning and milling can be performed in parallel but must be performed after blanking, drilling must be performed after turning is completed, and assembly must wait for all processing processes to be completed. These constraint relationships are accurately expressed in the directed graph.

[0052] In one possible implementation, the process execution sequence satisfying all dependency constraints is generated by traversing the directed graph through depth-first search. First, the in-degree of each node, i.e. the number of edges pointing to the node, is calculated. The nodes with an in-degree of 0 represent no precedence constraint and can be executed in priority. These nodes are added to the queue, and each time a node is taken out from the queue to be added to the sorting sequence, the in-degree of all successor nodes of the node is reduced by 1. When the in-degree of a certain successor node becomes 0, it is added to the queue. Repeat this process until all nodes are processed to obtain one or more feasible process execution sequences. For the case where there are multiple nodes with an in-degree of 0, permutations and combinations are generated to form a set of feasible sequences.

[0053] Exemplarily, the calculation of the comprehensive evaluation value fully considers the influence of the workpiece priority on production scheduling. For each feasible sequence, the system simulates the expected completion time of each workpiece when the sequence is executed, and calculates the cumulative time of each workpiece passing through all processes according to the standard processing time of the process and the device configuration. Multiply the completion time of each workpiece by its priority value, and the higher the priority of the workpiece, the greater the influence of its completion time on the evaluation value. The sum of the products of all workpieces gives the comprehensive evaluation value of the sequence, and the smaller the value, the higher the priority of the workpiece that can be completed earlier, and the better the scheduling scheme.

[0054] It can be understood that the system selects the sequence with the minimum comprehensive evaluation value from all feasible sequences as the final execution scheme. For the process showing a downward trend, the system will specially adjust its position in the new sequence to balance the load by changing the execution order to avoid the process becoming a production bottleneck. Further, the process execution order table not only specifies the execution priority of each process, but also contains dynamic scheduling parameters such as the maximum allowed waiting time, processing rules for emergency orders, etc., realizing fine management of production scheduling. Through this dynamic optimization mechanism based on data driving, the overall efficiency of the production line is improved by more than 15%, and the on-time delivery rate is increased from 88% to 96%.

[0055] In step S105, the queuing workpiece quantity and real-time load quantity corresponding to the rearranged process execution order are obtained, and the process processing capacity matching degree index is analyzed to determine the global workpiece flow stability evaluation result.

[0056] The rearranged process execution order is obtained, the queuing workpiece quantity and device real-time load quantity of each process under the execution order are read, the actual processing rate is calculated according to the number of workpieces completed per unit time, and compared with the standard processing rate designed in the process, to obtain the processing capacity value of each process. The production capacity difference between adjacent processes is calculated according to the processing capacity value, the processing capacity value of the upstream process is compared with the processing capacity value of the downstream process, and if the absolute value of the difference exceeds the preset matching threshold, it is marked as a mismatched process pair. The proportion of the number of mismatched process pairs to the total number of process pairs is calculated to obtain the process processing capacity matching degree index. Based on the process processing capacity matching degree index and the occupation rate of each process buffer, the matching degree index is multiplied by a preset weight coefficient and added to the weighted average value of the buffer occupation rate to obtain the global flow stability score. When the score is higher than the preset stability threshold, it is determined to be in a stable state, otherwise it is determined to be in an unstable state, and the global workpiece flow stability evaluation result is determined.

[0057] In an embodiment, the system obtains the rearranged process execution order, which has been optimized and adjusted. The current queuing workpiece quantity of each process is read through the real-time monitoring terminal of the workshop, and the real-time load quantity data of each device is obtained from the device management module. The actual processing rate is calculated by counting the number of workpieces completed in the last hour, and the standard processing rate is derived from the theoretical production capacity value specified in the process design document. The actual processing rate is divided by the standard processing rate to obtain the processing capacity value, which reflects the proportional relationship between the actual production capacity and the design capacity of the process.

[0058] Specifically, the calculation of the processing capacity value provides a quantitative basis for the evaluation of the running state of each process. When the processing capacity value of a certain process is 0.8, it means that the process has only reached 80% of the designed capacity, and there may be problems such as equipment aging, unskilled operation, or delayed material supply. The system identifies the capacity bottleneck by comparing the processing capacity values of adjacent processes. When the processing capacity value of the upstream process is 0.95 and that of the downstream process is only 0.75, the difference between them reaches 0.2, which exceeds the preset matching threshold of 0.15. The system will mark this pair of processes as a mismatched process pair.

[0059] It should be noted that the process processing capacity matching degree index evaluates the coordination degree of the entire production line by counting the proportion of mismatched process pairs. A certain production line has 10 pairs of adjacent processes, of which 3 pairs are identified as mismatched process pairs, and the matching degree index is calculated as 70%, indicating that the production line has a relatively obvious capacity imbalance phenomenon.

[0060] Preferably, the calculation of the global flow stability score adopts a two-factor weighting method. The weight coefficient of the matching degree index is set to 0.6, reflecting the importance of the coordination between processes; the weight coefficient of the weighted average of the buffer occupancy rate is set to 0.4, reflecting the influence of the buffer state on system stability. When the matching degree index is 70% and the average occupancy rate of the buffer is 65%, the stability score is calculated as 0.6x70+0.4x65=68. The system sets the stability threshold to 75, and the current score of 68 is lower than the threshold, determining that the system is in an unstable state.

[0061] In one possible implementation, the evaluation result triggers the corresponding adjustment mechanism. In an unstable state, the system automatically starts the capacity rebalancing program to improve the matching degree by adjusting the device allocation or modifying the process order until the stability score meets the preset requirements, achieving dynamic optimization of the production process.

[0062] Step S106, the production scheduling center re-identifies the device allocation ratio and the process processing execution order, combines the key process identification, the blocking type information, and the global workpiece flow stability evaluation result to determine the dynamic optimization implementation scheme, and maintains the process processing throughput.

[0063] The current equipment allocation ratio and process processing execution order are reacquired through the data interface of the production scheduling center, the equipment configuration quantity of each process in the equipment allocation ratio is read, and the execution priority order of each process in the execution order table is extracted, and the two data are combined to form a process resource configuration basic data set. From the process resource configuration basic data set, the process information corresponding to the key process identifier is screened out, the adjustment priority is determined in combination with the blocking type code of each process, the capacity bottleneck process that needs to increase the equipment and the resource surplus process that can reduce the equipment are identified according to the unstable factors in the global workpiece flow stability evaluation result. Based on the capacity bottleneck process and the resource surplus process, a resource configuration optimization model is established by using a mathematical optimization method, and under the premise of meeting the total equipment quantity constraint and the basic configuration requirement of the process, a new equipment allocation ratio and an adjusted process execution order are obtained. The new equipment allocation ratio and the adjusted process execution order are integrated into a dynamic optimization implementation scheme, which specifies the equipment adjustment quantity and the execution order adjustment position of each process, and after implementation, the actual throughput is monitored, and when the throughput remains in the preset range, it is confirmed that the target of maintaining the process processing throughput is reached.

[0064] In an embodiment, the production scheduling center is the core control unit of the entire production system, which acquires various production data in real time through a standardized data interface. The system first reads the current equipment allocation ratio data, which records the number of devices configured for each process, the device model, the rated capacity and other information. At the same time, the process execution order table is extracted from the execution management module, which specifies the sequence and priority of each process in the production process in detail. Through data integration program, the equipment configuration information and the execution order information are associated and matched according to the process number to form a unified process resource configuration basic data set, which serves as the basis for subsequent optimization decisions.

[0065] It should be noted that the selection of key processes follows the principle of multiple standard judgment. The system extracts records with key process identifiers from the basic data set, which have been marked in the previous blocking analysis process. Each key process is associated with a specific blocking type code, including device failure type, capacity insufficient type, material shortage type, rhythm mismatch type and other types. According to the severity and influence range of the blocking type, the system automatically calculates the adjustment priority score, the device failure type is given the highest priority 1.5, the capacity insufficient type is 1.2, the material shortage type is 1.0, and the rhythm mismatch type is 0.8.

[0066] Specifically, the identification of capacity bottleneck and resource surplus processes is based on multiple indicators in the global workpiece flow stability evaluation results. When the utilization rate of a process device continuously exceeds 95%, the number of queued workpieces exceeds 80% of the buffer capacity, and it is marked as an unstable factor in the stability evaluation, the system identifies it as a capacity bottleneck process, indicating that the process urgently needs to increase device resources. Conversely, a process with a device utilization rate of less than 60% and an empty buffer, and with normal upstream and downstream processes, is identified as a resource surplus process, and its device configuration can be reduced. The drilling process of a certain assembly line has a device utilization rate of 98%, and 120 pieces of workpieces are accumulated in front of it, which is identified as a typical capacity bottleneck. The adjacent deburring process has a device utilization rate of only 45%, which is identified as a resource surplus process.

[0067] Preferably, the application of the linear programming method involves a complex mathematical modeling process. The objective function is set to maximize the total throughput of the entire production line, expressed as the weighted sum of the actual output of each process, with the weights determined according to the contribution of the process to the product value chain. The constraint conditions include three levels: first, the total device quantity limit, i.e., the sum of the number of devices in all processes cannot exceed the total number of available devices in the workshop; second, the minimum configuration requirement of each process, at least one device is reserved for each process to maintain basic operation; third, the capacity balance constraint, the difference in capacity between adjacent processes cannot exceed the pre-set balance threshold, to avoid the creation of new bottlenecks. The linear programming problem is solved by the simplex algorithm or the interior point method to obtain the optimal number of devices for each process. At the same time, according to the new capacity distribution, the system recalculates the execution priority of each process, and the processes with lower capacity obtain higher execution priority to fully utilize their limited processing capacity.

[0068] In one possible implementation, the development of the dynamic optimization implementation scheme needs to consider the feasibility of actual operation. The scheme clearly specifies the specific content of device adjustment: 2 devices are transferred from the resource surplus deburring process to the capacity bottleneck drilling process; the originally fifth-ranked milling process is adjusted to the third execution to alleviate the waiting time of the downstream assembly process. Each adjustment is accompanied by detailed implementation time window, and device relocation is usually arranged during production gaps to avoid affecting normal production.

[0069] Illustratively, the system tracks the implementation effect of the scheme through a real-time monitoring mechanism. Within 24 hours after the completion of device adjustment, the system collects actual throughput data of each process every hour, and calculates the average throughput of the entire line. The pre-set throughput maintenance range is between 95% and 105% of the original throughput. When the throughput of the last three monitoring periods remains within this range and no new blocking points appear, the system determines that the goal of maintaining process handling throughput has been achieved.

[0070] It can be understood that the dynamic optimization mechanism can adaptively adjust the resource configuration according to the real-time changes of the production state, avoiding the resource waste and capacity bottleneck problem under the traditional fixed configuration mode. Through continuous monitoring and optimization, the overall equipment comprehensive efficiency and product delivery punctuality rate of the production line can be effectively improved. Further, the system also establishes an optimization effect evaluation archive, records the specific parameters and implementation effect of each adjustment, and provides historical experience reference for subsequent optimization decision.

[0071] The key process name and corresponding blocking type in the production line are extracted, the stagnation time and transfer delay times of the current workpiece between processes are obtained, the adjustment content of device reallocation and the rearrangement scheme of process execution sequence are formulated, the process position requiring to increase the number of devices and the specific number requiring to adjust the capacity allocation are determined, and a complete implementation plan including device allocation time arrangement and process priority rearrangement is formed.

[0072] The key process name and corresponding blocking type in the production line are extracted, the stagnation time record of each workpiece between processes is read from the production monitoring database, the transfer delay occurrence times of each process are counted, the process whose stagnation time is more than twice the average stagnation time is marked as a serious delay process, and the blocking and delay information are combined to form a process problem diagnosis table. According to the blocking severity in the process problem diagnosis table and the current capacity demand, the theoretical device number required by each process is calculated by dividing the capacity demand by the processing capacity of a single device, the device gap number is determined by comparing the existing device configuration, and the reordered execution sequence is obtained based on the dependency relationship between processes and the delay times. Based on the device gap number and the adjusted process execution sequence, the specific number and path of device transfer from surplus processes to gap processes are formulated, the device relocation standard time is queried according to the device model, the allocation time window is determined in combination with the shutdown maintenance period in the production plan, and the relocation start and end time table of each device is compiled. The device transfer path, relocation time table and new process execution sequence are integrated, the target capacity value of each process is recalculated according to the adjusted device configuration, the allowed fluctuation range is set as the upper and lower preset percentage of the target value, and a complete implementation plan including device allocation arrangement and process priority rearrangement is formed.

[0073] In one embodiment, the system extracts the basic information of all key processes from the production line monitoring database, including process name, belonging workshop, current device configuration, etc. Each key process is associated with a specific blocking type, such as reduced processing speed caused by device performance degradation, increased rework caused by process parameter deviation, and prolonged waiting time caused by material supply delay, etc. The system calculates the stagnation time of each workpiece between adjacent processes by analyzing the workpiece flow log, i.e. the time interval from the completion of the previous process to the start of the next process. When the stagnation time of a workpiece reaches more than twice the average stagnation time, the system records a transfer delay event.

[0074] It should be noted that the identification of the delay serious process adopts a dynamic threshold determination mechanism. The system first calculates the average idle time of all processes. The average idle time of a certain stamping production line is 15 minutes, and 30 minutes is set as the delay determination threshold. When the idle time of a certain process exceeds the threshold for 5 consecutive workpieces, or the number of delays in a single day exceeds 10 times, the process is marked as a delay serious process. The process problem diagnosis table adopts a structured data format, including process number, process name, blocking type, average idle time, delay number, severity level and other fields, providing comprehensive data support for subsequent optimization decisions.

[0075] Specifically, the calculation of the theoretical number of devices is based on the relationship between the capacity demand and the processing capacity of a single device. The system first extracts the daily capacity demand of each process from the production plan, for example, the drilling process needs to complete 1200 products. According to the device model, the standard processing capacity of a single device is queried. A certain type of drilling machine can process 60 pieces per hour, and according to 20 hours of work per day, the daily capacity of a single device is 1200 pieces.

[0076] Preferably, the reordering of the process execution sequence takes into account the inter-process dependency relationship and delay severity. The system constructs a process priority scoring matrix, with the dependency strength valued between 0 and 1, with strong dependency being 1 and weak dependency being 0.3; the delay number is normalized as another dimension of the score. The two dimensions are weighted and summed to obtain a comprehensive priority score, with weights of 0.4 and 0.6 respectively. Processes with higher scores get a higher execution position, but must meet the hard constraints of the process flow, such as heat treatment must be performed after machining. In this way, both the process requirements and the bottleneck processes are prioritized.

[0077] In one possible implementation, the preparation of the device relocation schedule requires accurate calculation of the time consumption of each link. The device relocation standard duration is determined according to factors such as device model, weight, relocation distance, etc. The relocation of a large punch press requires 8 hours, including 2 hours of disassembly, 3 hours of transportation, and 3 hours of installation and debugging. The system identifies the downtime maintenance window from the production plan, which is usually arranged during the weekend or night shift handover period. The devices that need to be relocated are sorted according to the urgency, and the devices that have the greatest impact on capacity are relocated first. The relocation plan for each device includes the start disassembly time, the expected completion time, the responsible team, the required tool list and other detailed information. When the actual capacity remains within this range, the adjustment scheme is considered to be successfully executed.

[0078] It can be understood that the formation of the complete implementation plan is a system integration process. The plan document includes execution summary, equipment deployment table, process sequence adjustment description, time schedule Gantt chart, risk response plan and other parts. Each adjustment clearly defines the responsible department and the completion time limit to ensure the executability of the scheme. Further, the system establishes an implementation effect tracking mechanism, which daily statistics the actual capacity of each process, equipment utilization rate, delay times and other key indicators, compares and analyzes with the planned target, finds out the deviation in time and takes corrective measures to effectively improve the overall efficiency of the production line.

[0079] Step S107, obtaining the total number of completed workpieces of each process from the dynamic optimization implementation scheme, judging the system processing throughput improvement degree, and determining the process load balancing state of workpiece flow.

[0080] The total number of completed workpieces of each process before and after the implementation is obtained from the execution record of the dynamic optimization implementation scheme. The change rate is obtained by subtracting the value before the implementation from the value after the implementation and then dividing by the value before the implementation. The total amount of workpieces passing through the entire production line per unit time is taken as the system processing throughput. The improvement degree is obtained by subtracting the throughput before the implementation from the throughput after the implementation and then dividing by the throughput before the implementation. According to the improvement degree and the change rate of the completed workpieces of each process, the load coefficient of each process is calculated. The load coefficient is the ratio of the actual completion amount to the rated capacity of the process. The standard deviation of the load coefficients of all processes is calculated. When the standard deviation is lower than a preset balancing threshold, it indicates that the load distribution of each process is uniform. Based on the distribution of the load coefficients, the processes with load coefficients greater than a preset upper limit are marked as overloaded, and the processes with load coefficients less than a preset lower limit are marked as idle. The proportion of the number of overloaded and idle processes to the total number of processes is calculated. When the proportion is lower than a preset standard, the process load balancing state of workpiece flow is determined to be good, otherwise it is determined to be unbalanced.

[0081] In one embodiment, the system extracts the historical production records of each process from the database of the dynamic optimization implementation scheme. The data collection period before the implementation is the average value of the month before the optimization, and the data after the implementation is the actual value of the first week after the optimization. A certain drilling process completes 800 pieces per day before the implementation, and 960 pieces per day after the implementation, with a change rate of (960-800) / 800x100%=20%. The system processing throughput is obtained by counting the output of the entire production line, which is 150 pieces per hour before the implementation and 180 pieces per hour after the implementation, with an improvement degree of 20%.

[0082] It should be noted that the load factor reflects the actual load level of the process. Each process has a designed rated capacity, for example, the rated capacity of a certain milling process is 50 pieces per hour. When the process actually completes 45 pieces, the load factor is 45 / 50 = 0.9. After the system calculates the load factors of all processes, it evaluates the degree of data dispersion by the standard deviation formula. The smaller the standard deviation, the closer the load of each process, the more balanced the production line. The preset balance threshold is usually set to 0.15, and when the standard deviation is lower than this value, it is considered that the load distribution is uniform.

[0083] Specifically, the determination of overload and idle state is based on the interval division of load factor. The process with load factor greater than 0.95 is in overload state, indicating that the process is close to full load operation and is easy to become a production bottleneck; the process with load factor less than 0.6 is in idle state, which exists capacity waste. The system counts the number of overloaded processes as 2, the number of idle processes as 1, and the total number of processes as 10, and the proportion of abnormal processes is 30%.

[0084] Preferably, the determination of process load balancing state adopts a comprehensive evaluation method. The preset standard is usually that the proportion of abnormal processes does not exceed 20%. The current proportion of 30% exceeds the standard, and the system determines that it is an unbalanced state. Such imbalance will lead to uncoordinated production rhythm, increase the waiting time of part of the processes, and reduce the overall efficiency.

[0085] In one possible implementation, the system also analyzes the specific reasons for imbalance. Overloaded processes may need to increase equipment or optimize process parameters, and idle processes can consider undertaking other production tasks or reducing equipment configuration. Through this dynamic monitoring and evaluation mechanism, the production management department can discover problems in time and take adjustment measures to maintain the efficient and stable operation of the production line.

[0086] The application also provides a data feedback task flow dynamic optimization system, mainly comprising: a feedback data acquisition and processing capacity identification module, which is used for acquiring process feedback data, determining process execution time consumption and processing results through production monitoring system logs, identifying current processing capacity of the process, and simultaneously acquiring the amount of workpieces to be processed at a downstream assembly node and the fluctuation amplitude of output workpiece frequency of an upstream process; a load evaluation and blockage identification module, which is used for recording process load variation trends according to device utilization and capacity occupation proportion, limiting workpiece flow if the amount of workpieces to be processed at the downstream assembly node exceeds a preset assembly threshold, evaluating the degree of upstream process output rhythm disorder according to the fluctuation amplitude of output workpiece frequency, identifying a list of potential process blockage positions and the amount of workpieces queued at each process; a key process analysis and device allocation adjustment module, which is used for extracting key process identifiers and blockage types from the list of potential process blockage positions, acquiring real-time load data of devices at each process and real-time load data of capacity at each process through a production scheduling center, evaluating the current device occupation proportion at each process in combination with the amount of workpieces queued, and adjusting device allocation; a device allocation scheme evaluation and process sequence rearrangement module, which is used for identifying the process device proportion in a device allocation scheme and the corresponding total amount of completed workpieces, and rearranging process execution sequences according to workpiece priorities and process dependency relationships if the total amount of completed workpieces corresponding to the process device proportion shows a downward trend; a processing capacity matching degree analysis module, which is used for acquiring the amount of workpieces queued at each process and real-time load corresponding to the rearranged process execution sequences, obtaining a process processing capacity matching degree index through analysis, and determining a global workpiece flow transfer stability evaluation result; a dynamic optimization scheme generation module, which is used for re-identifying device allocation proportions and process execution sequences through the production scheduling center, determining a dynamic optimization implementation scheme in combination with key process identifiers, blockage type information and the global workpiece flow transfer stability evaluation result, and maintaining process processing throughput; a throughput evaluation and load balancing judgment module, which is used for acquiring total number of completed workpieces at each process from the dynamic optimization implementation scheme, judging the improvement degree of system processing throughput, and determining the process load balancing state of workpiece flow transfer. The above description is only the preferred embodiments of the application and the explanation of the applied technical principles. Those skilled in the art should understand that the application range involved in the application is not limited to the technical solutions formed by the specific combinations of the above technical features, and meanwhile, should also cover other technical solutions formed by any combinations of the above technical features or equivalent features without departing from the concept of the application. For example, the above features are replaced with the technical features disclosed in the application (but not limited to) having similar functions to form technical solutions.

Claims

1. A method for dynamic optimization of task flow with data feedback, characterized in that, The method comprises: acquiring process feedback data, a quantity of workpieces to be processed, and a fluctuation amplitude of an output workpiece frequency, wherein the process feedback data comprises equipment processing time, workpiece completion status, and equipment usage frequency, the quantity of workpieces to be processed represents a workpiece accumulation quantity of a downstream assembly node, and the fluctuation amplitude of the output workpiece frequency represents a change in an output rhythm of an upstream process; determining a process load change trend according to equipment usage rate and capacity occupancy proportion, adjusting a workpiece flow based on a comparison result of the quantity of workpieces to be processed and a preset assembly threshold, and identifying a list of potential process blocking positions and a queued workpiece quantity in combination with the fluctuation amplitude of the output workpiece frequency; extracting a key process identifier and a blocking type from the list of potential process blocking positions, and adjusting an equipment allocation scheme in combination with real-time load quantity data of equipment, real-time load quantity data of capacity, and the queued workpiece quantity, wherein the equipment allocation scheme comprises equipment quantity allocation and capacity capacity allocation; rearranging a process execution order based on a process equipment proportion in the equipment allocation scheme and a total quantity of completed workpieces in combination with workpiece priority and process dependency; acquiring a queued workpiece quantity and a real-time load quantity according to the process execution order, determining a process processing capacity matching degree index and a global workpiece flow stability evaluation result; adjusting the equipment allocation scheme and the process execution order based on the key process identifier, the blocking type, and the global workpiece flow stability evaluation result, and determining a dynamic optimization implementation scheme; acquiring total quantity change data of completed workpieces according to the dynamic optimization implementation scheme, and determining a process load balancing state.

2. The method of claim 1, wherein, The acquiring of the process feedback data, the quantity of workpieces to be processed, and the fluctuation amplitude of the output workpiece frequency comprises: extracting running logs of each process equipment from a production monitoring system, calculating equipment processing time consumption and workpiece stay duration based on time stamps in the running logs, and determining a process processing capacity coefficient in the process feedback data; screening low processing capacity process nodes according to the process processing capacity coefficient, counting the quantity of workpieces to be processed, monitoring a change in the number of output workpieces of an upstream process, and determining the fluctuation amplitude of the output workpiece frequency based on a difference value of output quantities in adjacent time periods.

3. The method of claim 1, wherein, The determining of the process load change trend according to the equipment usage rate and the capacity occupancy proportion, the adjusting of the workpiece flow based on the comparison result of the quantity of workpieces to be processed and the preset assembly threshold, and the identification of the list of potential process blocking positions and the queued workpiece quantity in combination with the fluctuation amplitude of the output workpiece frequency comprise: forming a time sequence of the equipment usage rate and the capacity occupancy proportion of each process equipment, calculating a change rate based on the time sequence, and determining the process load change trend; when the quantity of workpieces to be processed exceeds the preset assembly threshold, sending a flow limitation instruction to an upstream process, and adjusting a workpiece release frequency; calculating a coefficient of variation based on an output time interval of an upstream process after the flow limitation instruction, determining an output rhythm disorder process, and generating the list of potential process blocking positions in combination with the process load change trend and the coefficient of variation, wherein the list of potential process blocking positions comprises a process number and a queued workpiece quantity.

4. The method of claim 3, wherein, The method comprises the following steps of: calculating a variation coefficient based on an output time interval of a process after the flow restriction instruction, determining a rhythm disorder process, comprising: Collecting the output time of the process after the flow restriction instruction, calculating the variation coefficient of the adjacent output time interval sequence; When the variation coefficient exceeds a preset threshold, mark the rhythm disorder process, count the queuing workpiece amount before the rhythm disorder process, and generate a list containing the process number and the blocking degree.

5. The method of claim 1, wherein, The method comprises the following steps of: extracting the key process identifier and the blocking type from the potential process blocking position list, combining the device real-time load data, the capacity real-time load data and the queuing workpiece amount to adjust the device allocation scheme, comprising: Extract the key process identifier and the blocking type in the potential process blocking position list, obtain the device real-time load data and the capacity real-time load data, and calculate the ratio of the queuing workpiece amount to the buffer capacity as the queuing saturation degree; According to the weighted sum of the device real-time load data, the capacity real-time load data and the queuing saturation degree, determine the resource shortage degree score, adjust the device number and the capacity upper limit based on the resource shortage degree score and the blocking type, and form the device allocation scheme.

6. The method of claim 1, wherein, The method comprises the following steps of: according to the process device ratio and the total amount of completed workpieces in the device allocation scheme, combining the workpiece priority and the process dependency relationship to rearrange the process execution order, comprising: Extract the process device ratio from the device allocation scheme, count the total amount of completed workpieces corresponding to the process device ratio, and determine the downward trend process based on the time series slope; According to the workpiece priority value and the process dependency relationship extracted from the downward trend process, construct a directed graph, generate a feasible process sequence based on topological sorting, and select the sequence with the minimum comprehensive evaluation value as the process execution order.

7. The method of claim 1, wherein, The method comprises the following steps of: according to the queuing workpiece amount and the real-time load under the process execution order, determining the process processing capacity matching degree index and the global workpiece flow stability evaluation result, comprising: Obtain the queuing workpiece amount and the real-time load under the process execution order, compare the actual processing rate with the standard processing rate to determine the process processing capacity value; According to the processing capacity value, calculate the capacity difference of adjacent processes, count the mismatch process pair, and determine the global workpiece flow stability evaluation result by combining the buffer occupancy rate weighted calculation.

8. The method of claim 1, wherein, The method comprises the following steps of: adjusting the device allocation scheme and the process execution order based on the key process identifier, the blocking type and the global workpiece flow stability evaluation result, determining a dynamic optimization implementation scheme, comprising: Obtain the current device allocation ratio and the process execution order, extract the process information corresponding to the key process identifier, and identify the capacity bottleneck process and the resource surplus process by combining the blocking type and the global workpiece flow stability evaluation result; Based on the capacity bottleneck process and the resource surplus process, adjust the device allocation ratio and the process execution order by using the linear programming method to form the dynamic optimization implementation scheme, and the dynamic optimization implementation scheme contains the device adjustment number and the execution order adjustment position. The linear programming method is used to adjust the equipment distribution ratio and the process execution sequence based on the capacity bottleneck process and the resource surplus process, and a dynamic optimization implementation scheme is formed, which includes: Extract the corresponding blocking type and workpiece stagnation time of the key process identifier, count the number of transmission delays, mark the delay serious process, and determine the number of equipment gaps; Based on the number of equipment gaps and process dependence relationship, adjust the process execution sequence, develop the equipment transfer path and relocation schedule, and form the dynamic optimization implementation scheme including equipment deployment arrangement and process priority rearrangement.

9. The method of claim 1, wherein, The total number of completed workpieces is obtained based on the dynamic optimization implementation scheme, and the process load balancing state is determined, which includes: Extract the total number of completed workpieces from the dynamic optimization implementation scheme, calculate the change rate and system processing throughput; Based on the change rate, calculate the load coefficient of each process, count the standard deviation of the load coefficient, mark the overload and idle process based on the load coefficient distribution, and determine the process load balancing state.

10. A data feedback task flow dynamic optimization system, characterized in that, The system comprises: a feedback data acquisition and processing capacity identification module for acquiring process feedback data, workpiece quantity to be processed and output workpiece frequency fluctuation amplitude, wherein the process feedback data includes equipment processing time, workpiece completion state and equipment use frequency, the workpiece quantity to be processed represents the workpiece accumulation quantity of the downstream assembly node, and the output workpiece frequency fluctuation amplitude represents the change of the upstream process output rhythm; a load evaluation and blockage identification module for determining the process load change trend according to the equipment utilization rate and the capacity occupation proportion, adjusting the workpiece flow based on the comparison result of the workpiece quantity to be processed and the preset assembly threshold, and identifying the list of potential process blockage positions and the queuing workpiece quantity in combination with the output workpiece frequency fluctuation amplitude; a key process analysis and equipment distribution adjustment module for extracting the key process identifier and the blocking type from the list of potential process blockage positions, and adjusting the equipment distribution scheme in combination with the real-time load quantity data of the equipment, the real-time load quantity data of the capacity and the queuing workpiece quantity, wherein the equipment distribution scheme includes equipment number distribution and capacity capacity distribution; an equipment distribution scheme evaluation and process sequence rearrangement module for rearranging the process execution sequence according to the process equipment proportion and the total quantity of completed workpieces in the equipment distribution scheme, in combination with the workpiece priority and the process dependence relationship; a processing capacity matching degree analysis module for obtaining the queuing workpiece quantity and the real-time load quantity according to the process execution sequence, determining the process processing capacity matching degree index and the global workpiece flow stability evaluation result; a dynamic optimization scheme generation module for adjusting the equipment distribution scheme and the process execution sequence based on the key process identifier, the blocking type and the global workpiece flow stability evaluation result, and determining a dynamic optimization implementation scheme; a throughput evaluation and load balancing judgment module for obtaining the total number of completed workpieces based on the dynamic optimization implementation scheme, and determining the process load balancing state.

Images