Fieldbus-based office chair hybrid production automation control method and system

Abstract

The application discloses an office chair hybrid production automation control method and system based on a field bus, which comprises the following steps: a main control PLC is communicated with each station PLC and a buffer area control unit to build a data structure for each station, the data structure comprising a reference beat and a station inertia coefficient; a beat and a buffer state are collected in real time to calculate a beat deviation and a buffer occupancy rate; an adjacent matrix is established according to a production line topology, and a disturbance propagation matrix is generated in combination with the buffer occupancy rate and the station inertia coefficient; the coefficients in the matrix are compared with a threshold value, the release time of a link exceeding the threshold value is calculated and adjusted by a damping coefficient, and the adjusted release time is sent, so that the release rhythm of the station materials is accurately controlled. The method can quantize the disturbance propagation intensity, accurately control the damping, effectively inhibit the disturbance diffusion and buffer blockage, and improve the beat stability and operation efficiency of the office chair hybrid production.

Description

Technical Field

[0001] This application relates to the field of production automation control technology, specifically to a fieldbus-based method and system for automated control of hybrid production of office chairs. Background Technology

[0002] Office chair production often employs a discrete manufacturing model involving the assembly of multiple models on mixed lines. The production line mixes manual and automated workstations, with buffer devices between processes but limited physical capacity. Furthermore, it features a topological characteristic of multiple branching material flows converging, placing high demands on the automation control of the production process. Currently, such production lines mostly use fieldbuses to connect the main control PLC with the PLCs at each workstation and the buffer control units. However, existing control methods can only achieve basic equipment start-up and shutdown and material conveying control, lacking the ability to finely regulate the production cycle.

[0003] In actual production, the assembly cycle time at each workstation is easily affected by factors such as operator proficiency and equipment status, causing fluctuations. These disturbances propagate along upstream and downstream workstations, easily leading to buffer blockages. Existing methods cannot quantify the path and intensity of disturbance propagation, making precise intervention difficult. Simultaneously, the resource waiting relationships between workstations and buffers in the production line are complex. Existing control logic cannot identify potential deadlock risks and can only stop the line after a deadlock occurs, resulting in poor production continuity. Furthermore, traditional control is mostly passive response regulation, lacking advance prediction and pre-control capabilities. This makes the production line prone to cycle time oscillations and frequent blockages, leading to reduced production efficiency and increased downtime losses.

[0004] Therefore, there is an urgent need for an automated control method that is adapted to the mixed production scenario of office chairs, to solve the problems of insufficient control of cycle disturbance, lack of deadlock prediction, and lag in control response in the existing technology, and to achieve coordinated control of production line cycle stability and deadlock prevention, thereby improving the stability and continuity of production operation. Summary of the Invention

[0005] The purpose of this application is to provide a fieldbus-based automated control method and system for mixed production of office chairs to solve the problems mentioned in the background art.

[0006] According to a first aspect of this application, a fieldbus-based automated control method for hybrid production of office chairs is provided, comprising the following steps: S1: The main control PLC establishes a communication link with each workstation PLC and buffer control unit through the fieldbus, and establishes a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient of the workstation, and stores all workstation data structures in the global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of the workstation and the reference cycle time. S2: The main control PLC receives the current assembly cycle time collected and uploaded in real time by each workstation PLC through the fieldbus, calculates the cycle time deviation based on the current assembly cycle time and the reference cycle time, and writes the cycle time deviation into the corresponding workstation data structure. S3: The main control PLC receives the actual number of caches and the maximum capacity of the cache from the cache control unit in real time through the fieldbus, calculates the cache occupancy rate based on the actual number of caches and the maximum capacity of the cache, and writes the cache occupancy rate into the corresponding workstation data structure. S4: The main control PLC establishes a workstation adjacency matrix based on the production line topology. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the disturbance propagation coefficient is calculated in combination with the workstation adjacency matrix, and a disturbance propagation matrix is ​​generated and stored. S5: Compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, calculate the adjusted release time based on the disturbance propagation coefficient and the damping coefficient, and send the adjusted release time to the corresponding workstation PLC through the fieldbus. The workstation PLC adjusts the material release rhythm of the workstation according to the adjusted release time.

[0007] Preferably, the disturbance propagation coefficient is calculated by multiplying the adjacency matrix elements representing the upstream and downstream connection relationship of the workstation, the upstream workstation buffer remaining capacity coefficient, and the downstream workstation disturbance absorption coefficient. The upstream workstation buffer remaining capacity coefficient is obtained by subtracting the upstream workstation buffer occupancy rate from the value 1, and the downstream workstation disturbance absorption coefficient is obtained by subtracting the downstream workstation workstation inertia coefficient from the value 1.

[0008] Preferably, the damping coefficient is dynamically determined based on the buffer occupancy rate, and the adjusted release time is calculated by adding the current assembly beat to the product of the damping coefficient and the disturbance propagation coefficient of the trigger beat damping control.

[0009] Preferably, the method further includes calculating the logically available capacity based on the actual cache usage, the maximum cache capacity, and the number of virtual cache tokens; the logically available capacity is calculated by subtracting the actual cache usage from the maximum cache capacity and then adding it to the number of virtual cache tokens; the number of virtual cache tokens is calculated by rounding up the ratio of the beat deviation to the reference beat.

[0010] Preferably, it also includes, S6: the main control PLC receives the waiting resource identifier reported by each workstation PLC when the duration of the workstation waiting resource status exceeds a preset waiting threshold, and constructs and updates a resource waiting directed graph containing workstation nodes and cache nodes based on the waiting resource identifier; S7: The main control PLC performs loop detection in the resource waiting directed graph. If a closed loop structure is detected and the average cache occupancy rate of all cache nodes in the loop structure is greater than the preset average cache occupancy rate threshold in the loop, it is determined to be a quasi-deadlock state.

[0011] Preferably, it further includes, S8: after determining that the state is quasi-deadlock, the main control PLC identifies all cache nodes in the closed loop structure, selects the node with the lowest cache occupancy rate as the target node, borrows logical capacity from the adjacent cache of the target node, updates the logical borrowable capacity of the target node according to the preset borrowing ratio and the physical remaining capacity of the adjacent cache, and sends it to the corresponding cache area control unit through the fieldbus to adjust the cache logical capacity.

[0012] Preferably, the logically available capacity of the target node after the update is calculated by adding the logically available capacity of the target node before the update to the borrowed capacity. The borrowed capacity is the product of the borrowing ratio and the physical remaining capacity of the adjacent cache. The physical remaining capacity of the adjacent cache is obtained by subtracting the actual number of adjacent caches from the maximum capacity of the adjacent cache.

[0013] Preferably, it further includes, S9: the main control PLC calculates the edge density of the resource waiting directed graph, the edge density being the ratio of the actual number of directed edges in the resource waiting directed graph to the theoretical maximum number of directed edges, and updates the damping coefficient according to the edge density to obtain the updated damping coefficient, the updated damping coefficient being obtained by multiplying the damping coefficient before the update with the value 1 plus the sum of the edge densities.

[0014] Preferably, it further includes, S10: the main control PLC inputs the cycle deviation of the workstation for the most recent specified number of cycles, the current buffer occupancy rate, and the current workstation inertia coefficient into a pre-trained long short-term memory network cycle prediction model to obtain the predicted value of the cycle deviation of the workstation for the next specified number of cycles; and triggers S5 based on the predicted value of the cycle deviation.

[0015] A second aspect of this application also provides a fieldbus-based automated control system for hybrid production of office chairs, comprising: Main control PLC, multiple station PLCs, and buffer control unit based on fieldbus connection; The main control PLC includes: The communication establishment module is used to establish communication links with the PLCs and buffer control units of each workstation via fieldbus. The data storage module is used to establish a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient for that workstation, and stores all workstation data structures in a global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of that workstation and the reference cycle time. The first data receiving module is used to receive the current assembly cycle time collected and uploaded in real time by the PLC of each workstation via the fieldbus, calculate the cycle time deviation based on the current assembly cycle time and the reference cycle time, and write the cycle time deviation into the corresponding workstation data structure. The second data receiving module is used to receive the actual number of caches occupied and the maximum capacity of the cache in real time collected and uploaded by the cache control unit through the fieldbus, calculate the cache occupancy rate based on the actual number of caches occupied and the maximum capacity of the cache, and write the cache occupancy rate into the corresponding workstation data structure. The matrix generation module is used to establish a workstation adjacency matrix based on the topology of the production line. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the module calculates the disturbance propagation coefficient in combination with the workstation adjacency matrix, generates the disturbance propagation matrix, and stores it. The disturbance control module is used to compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, the adjusted release time is calculated based on the disturbance propagation coefficient and the damping coefficient, and the adjusted release time is sent to the corresponding workstation PLC through the fieldbus. The workstation PLC is used to adjust the material release rhythm of the workstation according to the adjusted release time.

[0016] This application achieves quantitative characterization of workstation disturbance characteristics by constructing workstation inertia coefficients. Combined with the disturbance propagation matrix established by the production line topology, it can accurately locate the disturbance propagation path and intensity, realizing targeted damping control of cycle time disturbances and effectively avoiding efficiency losses caused by global speed reduction. Simultaneously, through the design of virtual cache tokens and logically borrowable capacity, the cache scheduling capability is expanded without changing the hardware structure. Combined with loop detection and quasi-deadlock determination in the directed graph of resource waiting, early warning and topology reconstruction can be completed before deadlocks occur, achieving deadlock resolution without stopping the production line. Furthermore, the dynamic updating of edge density to the damping coefficients achieves synergy between topology and dynamic control. Combined with cycle time prediction using the LSTM model, control strategies can be triggered in advance, transforming passive response into proactive prediction, significantly improving the production line's anti-interference capability and operational continuity, significantly reducing downtime losses, and effectively improving the overall efficiency and stability of mixed-line production of office chairs. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0018] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 A schematic diagram of an automated control method for hybrid production of office chairs based on fieldbus provided in this application embodiment; Figure 2 A schematic diagram illustrating the process of constructing the perturbation propagation coefficient matrix provided in this application embodiment; Figure 3 This is a schematic diagram of the quasi-deadlock state determination process provided in the embodiments of this application; Figure 4 This is a schematic diagram of the cache topology dynamic reconstruction process provided in an embodiment of this application; Figure 5 A schematic diagram of a fieldbus-based automated control system for hybrid production of office chairs is provided for embodiments of this application. Figure 6 This is a schematic diagram of the main control PLC structure provided in the embodiments of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] It should be noted that all user information (including but not limited to user device information, user personal information, object information corresponding to device usage data, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, device usage data, etc.) involved in all embodiments of this application are information and data authorized by the user or fully authorized by all parties.

[0021] This method is suitable for discrete manufacturing scenarios involving the mixed assembly of multiple office chair models. These scenarios are characterized by a mix of manual and automated workstations, limited physical capacity of buffers in each process, and the convergence of multiple branch material flows. The execution unit of the method consists of a main control programmable logic controller (PLC), slave PLCs at each workstation, buffer control units, and a fieldbus communication network. The fieldbus can adopt PROFINET, EtherCAT, or CAN open industrial fieldbus. The main control PLC, as the global decision-making core, establishes periodic communication with each execution unit through the fieldbus. Through bus semantic extension and control logic optimization, it achieves coordinated control for production line cycle stability and deadlock prevention.

[0022] The following detailed description, in conjunction with specific embodiments, illustrates the implementation process of the fieldbus-based automated control method for hybrid production of office chairs as described in this application. It should be noted that these embodiments are merely illustrative of this application and not intended to limit its scope of protection. Any conventional adjustments or substitutions made by those skilled in the art to the steps without departing from the concept of this application should be included within the scope of protection of this application.

[0023] like Figure 1 As shown in the figure, this application discloses a schematic diagram of an automated control method for hybrid production of office chairs based on fieldbus, including the following method steps: S1: The main control PLC establishes a communication link with each workstation PLC and buffer control unit through the fieldbus, and establishes a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient of the workstation, and stores all workstation data structures in the global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of the workstation and the reference cycle time. S2: The main control PLC receives the current assembly cycle time collected and uploaded in real time by each workstation PLC through the fieldbus, calculates the cycle time deviation based on the current assembly cycle time and the reference cycle time, and writes the cycle time deviation into the corresponding workstation data structure. S3: The main control PLC receives the actual number of caches and the maximum capacity of the cache from the cache control unit in real time through the fieldbus, calculates the cache occupancy rate based on the actual number of caches and the maximum capacity of the cache, and writes the cache occupancy rate into the corresponding workstation data structure. S4: The main control PLC establishes a workstation adjacency matrix based on the production line topology. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the disturbance propagation coefficient is calculated in combination with the workstation adjacency matrix, and a disturbance propagation matrix is ​​generated and stored. S5: Compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, calculate the adjusted release time based on the disturbance propagation coefficient and the damping coefficient, and send the adjusted release time to the corresponding workstation PLC through the fieldbus. The workstation PLC adjusts the material release rhythm of the workstation according to the adjusted release time.

[0024] In some embodiments, for step S1, a communication link is established and a workstation data structure is constructed. For example, after the main control PLC completes its system self-test upon power-up, it sends a link establishment request frame to each workstation PLC and the buffer control unit via the fieldbus. Upon receiving the request frame, each slave unit sends back a response frame containing its own device identifier and communication parameters. The main control PLC completes the handshake establishment of the communication link based on the response frame. The preset communication period can be configured to milliseconds to ensure real-time data interaction.

[0025] Once the communication link is established, the main control PLC assigns a unique digital identifier to each workstation on the production line. And construct the workstation data structure based on this identifier. This data structure serves as the data carrier for all subsequent control logic, and its core contains the reference clock cycle. With workstation inertia coefficient Two parameters; the main control PLC will control all workstations. Summarize, build, and store global data tables. .

[0026] Specifically, refer to the beat The determination method is as follows: the main control PLC retrieves the assembly cycle time data of this workstation during the stable production phase from the historical production database. The criteria for determining the stable production phase are that the production line runs continuously for a preset duration without any downtime or buffer blockage alarms. The main control PLC statistically analyzes the assembly cycle time within this period, calculates its arithmetic mean, and assigns this value... . The update mechanism is offline, which can be recalculated and written based on the latest historical production data when the production line is shut down for maintenance. Its value range matches the actual assembly process time of the workstation.

[0027] Workstation inertia coefficient This is a key parameter characterizing the response characteristics of a workstation to cycle time disturbances. Its value is calculated based on the variance of the assembly cycle time of that workstation and the reference cycle time. In the specific calculation process, the main control PLC first calculates the workstation's response based on historical stable production data. Variance of assembly cycle time This variance reflects the historical dispersion of the assembly cycle time at the workstation. The workstation inertia coefficient can be calculated using the following formula. : In the formula, For workstation The variance of assembly cycle time. For workstation The reference beat. As can be seen from the formula, The range of values ​​is When the assembly cycle time at the workstation is highly discrete, i.e. When the value is large, A value approaching 1 indicates that the workstation has a strong ability to amplify external clock disturbances and is easily a source of disturbance; when When the value is small, A value close to 0 indicates that the assembly process at this station is stable and has a strong ability to absorb disturbances.

[0028] This calculation method solves the problem of the inability to quantify the disturbance characteristics of work stations in traditional control. By converting the historical performance of work stations into numerical coefficients, the principle is to eliminate the influence of differences in the base number of different work stations by using the normalization of variance and reference cycle time, thus achieving horizontal comparability of work station characteristics.

[0029] The main control PLC completes and After the calculation, write it into the corresponding... Data structure, and store in a global data table. The data table uses a cyclic overwrite storage method to ensure that the main control PLC can retrieve the parameters of any station in subsequent steps with constant time complexity.

[0030] In some embodiments, step S2 involves cycle time data acquisition and cycle time deviation calculation. According to embodiments of this disclosure, each workstation PLC is connected to a sensor signal connection with the workstation actuator to acquire assembly cycle time data in real time. The specific acquisition logic is as follows: the workstation PLC detects the start and finish trigger signals of the assembly process, calculates the time interval between the two signals using an internal timer, and determines this time interval as the cycle time of the workstation. Current assembly cycle time After an assembly cycle is completed, the workstation PLC, according to a preset communication cycle, communicates via the fieldbus. With their own workstation identification Package and upload to the main control PLC.

[0031] The main control PLC receives the data uploaded by each workstation. Afterwards, data verification is performed to remove outliers caused by communication interference. The rules for determining outliers can be configured to deviate from the target range. Data reaching the preset ratio. After data verification, the main control PLC determines the data based on the workstation identifier. In the global data table Retrieve the corresponding Data structure, read the reference beat from it .

[0032] The main control PLC calculates the station position using the following formula. rhythm deviation : In the formula, For workstation The current assembly cycle time, For workstation Reference beat. The value can be positive or negative, positive This indicates that the current assembly cycle is slower than the reference cycle, indicating a hysteresis disturbance; negative... This indicates that the current assembly cycle time is faster than the reference cycle time, indicating a leading disturbance. This parameter is a core indicator for quantifying the deviation of the real-time operating status of the chemical level.

[0033] The main control PLC will calculate the result Write to global data table The corresponding The data structure overwrites the original values. This embodiment realizes real-time quantification of production line cycle disturbances, solving the problem that traditional control can only judge the busyness of the workstation through switching quantities and cannot detect small cycle fluctuations, enabling the control system to capture early signals of deadlock and oscillation.

[0034] In some embodiments, for step S3, the buffer status is acquired and capacity parameters are calculated. Specifically, the buffer control unit is connected to the sensors of the buffer devices corresponding to each workstation, and the sensors monitor the quantity of semi-finished products in the buffer in real time. The buffer control unit uploads buffer status data to the main control PLC according to a communication cycle consistent with the fieldbus. This data includes the actual quantity of buffers occupied. With maximum cache capacity .in, This is a physical attribute of the cache device, determined by the hardware structure. It is a fixed value and is written to the cache control unit during system initialization. These are real-time changing values, obtained from sensor counting.

[0035] The main control PLC receives and Then, based on the correspondence between the cache and the workstation, the global data table is located. The corresponding Data structure. Calculate cache utilization using the following formula. : In the formula, This represents the actual amount of cache used. This is the maximum cache capacity. The range of values ​​is Its value directly reflects the congestion level of the buffer and is a key basis for judging the material flow status of the production line. The main control PLC will calculate the value. Write the corresponding Data structures.

[0036] In embodiments that include virtual cache tokens, the main control PLC also needs to read... Number of virtual cache tokens generated in the data structure And calculate the logically available capacity. : In the formula, The physical remaining capacity of the cache. This represents the number of virtual cache tokens. Logical borrowable capacity. The principle behind this is to logically expand the cache's scheduling capacity using virtual tokens without altering its physical structure. This parameter solves the congestion sensitivity problem caused by insufficient physical capacity in small cache production lines. It can absorb upstream materials through logical capacity allocation before the cache reaches its physical capacity, preventing hard blockages. The main control PLC will calculate the... Write the corresponding Data structures.

[0037] In some embodiments, for step S4, please refer to Figure 2 , Figure 2 This is a schematic diagram illustrating the process of constructing the disturbance propagation coefficient matrix provided in an embodiment of this application. For example, in S201, the main control PLC establishes a workstation adjacency matrix based on the physical layout and material flow direction of the office chair production line. This matrix is ​​a two-dimensional square matrix, with its rows and columns corresponding to the workstation identifiers on the production line. and Matrix elements The rule for determining the value is: if the workstation For workstation The direct upstream workstation, and the two have a direct material transfer relationship, then If workstation With workstation If there is no direct upstream and downstream material transfer relationship, then Workstation adjacency matrix Once the production line topology is determined, it is preset in the main control PLC. If the production line process changes, it can be modified through offline configuration.

[0038] In S202, calculate the disturbance propagation coefficient between any two workstations with upstream and downstream relationships. This is done after obtaining the workstation adjacency matrix. Then, the main control PLC retrieves data from the global data table. Retrieve cache usage rate for all workstations With workstation inertia coefficient Calculate the disturbance propagation coefficient between any two workstations with an upstream-downstream relationship. : In the formula, These are the elements of the workstation adjacency matrix. For upstream workstations The remaining cache capacity coefficient, For downstream workstations The disturbance absorption coefficient. Its principle lies in coupling and modeling the production line topology, the buffer capacity of the buffer, and the disturbance characteristics of the workstation. Buffer remaining capacity coefficient. The disturbance absorption coefficient represents the physical buffering capacity of the upstream buffer against disturbances; the larger the value, the stronger the buffering capacity and the weaker the disturbance propagation. It characterizes the downstream workstation's inherent ability to absorb disturbances. The larger the value, the stronger the absorption capacity and the weaker the disturbance propagation.

[0039] In S203, all calculation results are arranged in order of workstation identification to generate a disturbance propagation matrix. The main control PLC controls all... and Complete the combination After the calculation, all calculation results are arranged in the order of workstation identifiers to generate a disturbance propagation matrix. This matrix and the workstation adjacency matrix With the same dimensions, each element precisely quantifies the upstream workstations. The rhythm disturbance to the downstream work station The intensity of the propagation. The main control PLC will use the disturbance propagation matrix. By storing the data, this matrix solves the problem of the inability to quantify the propagation path and intensity of disturbances in traditional control, enabling the control system to accurately locate the key propagation links of disturbances, achieve precise damping control, and avoid the efficiency loss caused by global deceleration.

[0040] In some embodiments, for step S5, cycle damping control and virtual buffer token generation are performed. According to embodiments of this disclosure, the main control PLC first retrieves a preset disturbance propagation coefficient threshold. The threshold is determined by multiplying the average disturbance propagation coefficient during the historical stable operation of the production line by a preset safety factor. Its value range is... It is used to determine whether the propagation of disturbances has reached a level that requires intervention.

[0041] The main control PLC will propagate the disturbance matrix. Each element in and Perform a comparison one by one. If any one exists... Greater than Then determine the workstation To the workstation The risk of disturbance propagation is too high, so the main control PLC triggers the cycle damping control process.

[0042] In cycle damping control, the damping coefficient These are the core control parameters, and their values ​​are determined based on the upstream workstation. Corresponding cache usage Dynamically determined. Specifically, it is determined using a linear mapping relationship. ,in The coefficient is a proportionality coefficient, calibrated using historical production data. For example, data on the natural fluctuations in workstation cycle time are collected during the process of increasing buffer occupancy from 0.2 to 0.8 on the production line without external intervention. The least squares method is used to fit a linear relationship between buffer occupancy and the additional release time required to suppress disturbances. The slope of this fitted line is the initial value of $k$. In subsequent operation, the value of $k$ can be fine-tuned based on the actual control effect, such as the amplitude of production line oscillations. The fine-tuning step size can be set to 0.01. In this embodiment, the typical range of $k$ is [0,1], ensuring... The final value falls on Within the range. The larger the value, the more congested the representation cache becomes, and the higher the risk of perturbation propagation. The larger the value, the stronger the damping control.

[0043] Determine the damping coefficient Afterwards, the main control PLC calculates the station. Adjusted release time : In the formula, For workstation The current assembly cycle time, The damping coefficient is... This is the disturbance propagation coefficient that triggers this control. The principle is to dynamically increase the material release interval of the upstream station based on the disturbance propagation strength and the congestion level of the buffer, achieving soft blocking delay. Compared to traditional hard blocking control, this method can avoid severe start-stop oscillations in the production line.

[0044] The main control PLC will calculate the result Send to the corresponding workstation via fieldbus The PLC at the workstation receives the instruction and adjusts its internal material release timer accordingly. The duration of the control mechanism is used to start the material conveying mechanism, thereby slowing down the material output speed to the downstream workstation and suppressing the propagation of disturbances.

[0045] While generating the adjusted release time, the main control PLC also executes the virtual buffer token count. Calculation: In the formula, It is a rounding function. For workstation The rhythm deviation, For workstation The reference beat. When If the value is negative, meaning the workstation cycle time is faster than the reference cycle time, the calculation result will be non-positive. In this case, the main control PLC will... Assign a value of 0; when A positive value indicates that the workstation cycle time is lagging. This is a positive integer, and its value is positively correlated with the degree of lag. The main control PLC will generate... Write to global data table The corresponding The data structure, whose parameter is involved in the calculation of the available capacity, enables the coordinated linkage between clock damping control and buffer capacity management.

[0046] In some embodiments, step S6 is further included, specifically, each workstation PLC monitors its own operating status in real time. When a workstation is unable to continue the assembly process due to waiting for materials, equipment, or manpower resources, the workstation PLC starts its internal waiting timer. When the duration of the waiting state exceeds a preset waiting threshold... At that time, the workstation PLC will generate a resource waiting identifier. The identifier contains the identifier of the waiting workstation and the code of the required resources.

[0047] Preset waiting threshold The method of determination is based on the reference beat. The preset ratio, whose value is determined by statistical analysis of historical production line data, ensures that instantaneous resource contention is filtered out, and only persistent waiting states are reported. The workstation PLC identifies waiting resources via the fieldbus. Upload to the main control PLC.

[0048] The main control PLC receives the waiting resource identifier Then, combined with the global data table The mapping between workstations and caches in the storage is used to construct a directed graph of resource waiting. The directed graph contains all workstation nodes and buffer nodes of the production line, and each node has a unique identifier. The edges of this directed graph are directed edges, used to represent the waiting relationships between nodes. If a workstation... Waiting for cache Release materials at the node With nodes Establish a line between them point to Directed edges; if cached Waiting workstation Input materials to create a process from... point to The directed edge.

[0049] The main control PLC waits for resources in a directed graph according to the communication cycle of the fieldbus. Perform dynamic updates. If a wait-to-cancel signal is received from the workstation PLC, delete the corresponding directed edge; if a new wait resource identifier is received... If so, then a corresponding directed edge will be added.

[0050] This embodiment transforms the resource dependencies in the production line into a graph structure, solving the problem that traditional PLCs cannot identify complex resource waiting relationships. This allows the loop detection algorithm to be directly applied to production control for early warning of deadlock.

[0051] In some embodiments, step S7 is further included, which further includes a quasi-deadlock state determination. See also... Figure 3 , Figure 3 This is a schematic diagram of the quasi-deadlock state determination process provided in an embodiment of this application. For example, in S301, the main control PLC completes the resource waiting directed graph... After the system is built and updated, a cycle detection algorithm is executed. The cycle detection algorithm can employ a depth-first search (DFS) algorithm, which starts from the resource-waiting directed graph. Starting from any node, traverse along the directed edges. If you return to a previously visited node during the traversal, then the graph contains a closed cycle.

[0052] In S302, if a closed loop structure is detected, the main control PLC further extracts all buffer nodes contained in that loop structure. For each buffer node, the main control PLC retrieves data from the global data table. Retrieve the corresponding cache usage rate Subsequently, the average cache occupancy rate within the ring structure was calculated. : In the formula, This represents the total number of cache nodes in the closed-loop structure. For the first in the ring The cache utilization rate of each cache node.

[0053] In S303, the average buffer occupancy rate in the loop is compared with a threshold to determine a near-deadlock state. The main control PLC retrieves the preset average buffer occupancy rate threshold in the loop. This threshold is obtained by calibrating the cached state data before historical deadlocks occurred on the production line, and its value range is [value range missing]. The main control PLC will calculate the... and Perform a comparison, if If so, the production line is determined to be in a near-deadlock state.

[0054] The principle behind this judgment logic lies in combining the topological characteristics of resource waiting with the congestion status of the buffer. A simple ring structure only indicates the possibility of resource contention, while a high buffer occupancy rate means that material flow is close to stagnation. Only by combining the two can deadlock formation be accurately predicted. This embodiment solves the drawback of traditional control methods that can only stop the line after deadlock occurs. It issues an early warning at the critical state before deadlock formation, buying time for topology reconfiguration.

[0055] In some embodiments, dynamic S8 cache topology reconfiguration is also included. See [link / reference] Figure 4 , Figure 4 This is a schematic diagram of the cache topology dynamic reconstruction process provided in an embodiment of this application. According to an embodiment of this disclosure, when the main control PLC determines that the production line is in a quasi-deadlock state, the cache topology dynamic reconstruction process is immediately initiated.

[0056] In S401, the node with the lowest buffer occupancy rate is selected as the target node. The main control PLC identifies all buffer nodes in the closed-loop structure corresponding to the near deadlock and retrieves them from the global data table. Retrieve the cache usage of these nodes .

[0057] The main control PLC controls all buffer nodes in the ring. Sort the nodes in ascending order and select the first node in the sorted list, i.e., the node with the lowest cache usage, as the target node. The principle behind selecting this target node is that it has the most remaining physical capacity, which can accommodate additional logical capacity, thereby effectively breaking the bottleneck of material flow.

[0058] In S402, the physical remaining capacity of adjacent buffers is calculated. The main control PLC identifies the target node. Adjacent cache, defined as the cache in the production line topology that is adjacent to the target node. Buffers with direct material transfer relationships exist. For each adjacent buffer, the main control PLC retrieves its maximum buffer capacity. Compared with the actual number of occupants And calculate the physical remaining capacity of adjacent caches. : In the formula, The maximum capacity of adjacent caches. This represents the actual number of adjacent caches occupied.

[0059] The main control PLC retrieves the preset borrowing ratio. This parameter is a weighting coefficient, and its value range is... Its value is obtained through self-learning by statistically analyzing the historical reconstruction effect of the production line, and is used to balance the capacity requirements of the target node and the self-security of adjacent caches. For example, the self-learning acquisition method refers to using actual operational data from historical cache topology reconstructions on the production line as samples, judging the effectiveness of each reconstruction through preset evaluation metrics, and then dynamically adjusting the value based on the feedback. Specifically, evaluation metrics for reconstruction effectiveness are first set, such as the time to resolve the near-deadlock state after reconstruction, the time for the production line cycle time to recover to a stable state, and the probability of cache blocking recurring. The degree of metric achievement is used as the judgment criterion. If the metric performs well after a reconstruction, the current value is retained or slightly adjusted in that direction; if the metric performs poorly, the value is adjusted in the opposite direction. Through continuous learning and iterative optimization of historical reconstruction cases, the optimal value is finally obtained.

[0060] Calculate borrowed capacity : In the formula, For the borrowing ratio, This represents the physical remaining capacity of the adjacent cache.

[0061] In S403, the updated logically available capacity is calculated. The main control PLC retrieves the data from the global data table. Retrieve target node The logic before the update can borrow capacity And calculate the updated logically available capacity. : In the formula, The available capacity is used for the logic prior to the target node update. To borrow capacity.

[0062] The main control PLC will calculate the result Data is sent to the target node via fieldbus. The corresponding buffer control unit. The buffer control unit is configured according to... Update its capacity determination logic to allow more materials to enter the cache.

[0063] This embodiment changes the reachability path of resources in the ring structure by borrowing logical capacity across nodes, breaking the deadlock loop of mutual waiting at the topology level. Without stopping the production line, it successfully resolves the deadlock risk and significantly improves the continuous operation capability of the production line.

[0064] In some embodiments, the method further includes S9, which updates the damping coefficient based on edge density. Specifically, the main control PLC completes the resource waiting directed graph... After the update, calculate the edge density of the directed graph. Edge density is a parameter characterizing the density of resource waiting relationships in a production line. It is calculated as the number of directed edges actually existing in the resource waiting directed graph. Compared with the theoretical maximum number of directed edges The ratio of [the number of directed edges]. The theoretical maximum number of directed edges. Based on the total number of nodes The decision is made for a mixed node set that includes workstations and caches. Calculate edge density : In the formula, This represents the actual number of directed edges. This represents the theoretical maximum number of directed edges. The range of values ​​is The larger the value, the more intense the resource competition in the production line and the more strained the topology.

[0065] The main control PLC reads the current damping coefficient in step S5. And calculate the updated damping coefficient. : In the formula, The damping coefficient before the update. This refers to edge density. The principle is that the topological tension of the production line is used inversely to affect the strength of the cycle damping control. When the edge density... Increasing the damping coefficient signifies intensified resource competition and a higher risk of deadlock. In this case, increasing the damping coefficient is necessary. This can further reduce the material release rate of upstream workstations, and alleviate resource competition by cooperating with topology reconfiguration from a dynamic behavior perspective.

[0066] The main control PLC is based on the updated damping coefficient. Used for release time after the next adjustment The calculation. This step realizes bidirectional collaboration between the topology layer and the dynamic behavior layer, solving the problem of the two layers being independent and having contradictory control effects in traditional control, and improving the anti-interference capability of the production line under complex working conditions.

[0067] In some embodiments, the method further includes S10, which involves beat prediction and pre-trigger control based on a Long Short-Term Memory (LSTM) network. For example, while performing the aforementioned steps, the main control PLC runs a Long Short-Term Memory (LSTM) beat prediction model in parallel. This model has a hierarchical, serial structure, consisting of an input layer, a first LSTM layer, a second LSTM layer, a fully connected layer, and an output layer, from input to output. The number of neurons in the input layer is consistent with the dimension of the input features, which is the nearest workstation. Cycle deviation Current cache usage Current workstation inertia coefficient , The preset sliding window length. The first LSTM layer contains... One neuron is used to extract shallow features from time-series data; the second LSTM layer contains Each neuron is used to extract deep long-term dependency features from time-series data. The fully connected layer contains... The output layer has neurons, using the Rectified Linear Array (ReLU) as the activation function to achieve feature fusion mapping. , To predict the step size, a linear activation function is used, and the output is the future position value. Predicted value of cycle deviation for each period .

[0068] In one embodiment, the training process of the model involves using historical operating data from the production line as the training set, dividing the data into training, validation, and test sets according to a preset ratio. The loss function is mean squared error (MSE), and the optimizer uses adaptive moment estimation (Adam). The model's weights and biases are iteratively updated using the backpropagation algorithm until the loss function of the validation set converges, completing the model training. The trained model can then be embedded and deployed in the embedded computing core of the main control PLC.

[0069] During production line operation, the main control PLC retrieves data from the global data table. Input features are extracted in real time, normalized using Z-Score, and then fed into the LSTM model. The model outputs future... Predicted value of cycle deviation for each period The main control PLC compares the predicted value with a preset prediction threshold. If the predicted value indicates that a clock disturbance exceeding the threshold will occur in the future, or if, based on the predicted clock deviation sequence and through simulation calculations in steps S1-S7, a quasi-deadlock loop structure meeting the S7 judgment condition is predicted to form in the future control cycle, the main control PLC will trigger the clock damping control in step S5 or the buffer topology dynamic reconstruction in step S8 in advance. By introducing artificial intelligence prediction, the control system is transformed from a passive response to an active prediction, solving the response lag problem of traditional control, further reducing the probability of disturbance propagation and deadlock formation, and improving the intelligence and foresight of production line control.

[0070] This method achieves precise damping and suppression of cycle time oscillations by constructing a disturbance propagation model; it mitigates deadlocks in advance through loop detection and logical capacity borrowing in the resource waiting graph; and it achieves self-optimization and foresight of the control strategy through reverse adjustment of edge density and prediction using the LSTM model. It effectively solves the cascading problems of cycle time disturbances, buffer blocking, and deadlocks in production lines, significantly improving the operational stability, continuity, and production efficiency of the production line, and reducing production losses caused by line downtime.

[0071] It should be noted that although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted in the flowchart can be performed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0072] Please see Figure 5 , Figure 5This application provides a block diagram of a fieldbus-based automated control system for hybrid production of office chairs. The system specifically includes: Main control PLC, multiple station PLCs, and buffer control unit based on fieldbus connection; Among them, such as Figure 6 As shown, Figure 6 This is a schematic diagram of the main control PLC structure provided in an embodiment of this application. The main control PLC includes: The communication establishment module 601 is used to establish a communication link with the PLC of each workstation and the buffer control unit via the fieldbus. The data storage module 602 is used to establish a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient for that workstation, and stores all workstation data structures in a global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of that workstation and the reference cycle time. The first data receiving module 603 is used to receive the current assembly cycle time collected and uploaded in real time by the PLC of each workstation via the fieldbus, calculate the cycle time deviation based on the current assembly cycle time and the reference cycle time, and write the cycle time deviation into the corresponding workstation data structure. The second data receiving module 604 is used to receive the actual number of caches and the maximum capacity of the cache collected and uploaded in real time by the cache control unit via the fieldbus, calculate the cache occupancy rate based on the actual number of caches and the maximum capacity of the cache, and write the cache occupancy rate into the corresponding workstation data structure. The matrix generation module 605 is used to establish a workstation adjacency matrix according to the topology of the production line. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the disturbance propagation coefficient is calculated in combination with the workstation adjacency matrix to generate and store the disturbance propagation matrix. The disturbance control module 606 is used to compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, the adjusted release time is calculated based on the disturbance propagation coefficient and the damping coefficient, and the adjusted release time is sent to the corresponding workstation PLC through the fieldbus. The workstation PLC is used to adjust the material release rhythm of the workstation according to the adjusted release time.

[0073] It should be noted that the working process of each module in the fieldbus-based automated control system for mixed production of office chairs described in this embodiment can refer to the working process of the fieldbus-based automated control method for mixed production of office chairs described in the above embodiments, and the technical effects achieved are the same as those of the fieldbus-based automated control method for mixed production of office chairs described in the above embodiments, so they will not be repeated here.

[0074] The above description represents the preferred embodiments of the present invention. It should be noted that, for those skilled in the art, various improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A fieldbus-based automated control method for hybrid production of office chairs, characterized in that, Includes the following steps: S1: The main control PLC establishes a communication link with each workstation PLC and buffer control unit through the fieldbus, and establishes a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient of the workstation, and stores all workstation data structures in the global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of the workstation and the reference cycle time. S2: The main control PLC receives the current assembly cycle time collected and uploaded in real time by each workstation PLC through the fieldbus, calculates the cycle time deviation based on the current assembly cycle time and the reference cycle time, and writes the cycle time deviation into the corresponding workstation data structure. S3: The main control PLC receives the actual number of caches and the maximum capacity of the cache from the cache control unit in real time through the fieldbus, calculates the cache occupancy rate based on the actual number of caches and the maximum capacity of the cache, and writes the cache occupancy rate into the corresponding workstation data structure. S4: The main control PLC establishes a workstation adjacency matrix based on the production line topology. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the disturbance propagation coefficient is calculated in combination with the workstation adjacency matrix, and a disturbance propagation matrix is ​​generated and stored. S5: Compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, calculate the adjusted release time based on the disturbance propagation coefficient and the damping coefficient, and send the adjusted release time to the corresponding workstation PLC through the fieldbus. The workstation PLC adjusts the material release rhythm of the workstation according to the adjusted release time.

2. The automated control method for hybrid production of office chairs based on fieldbus as described in claim 1, characterized in that, The disturbance propagation coefficient is calculated by multiplying the adjacency matrix elements representing the upstream and downstream connection relationship of the workstation, the upstream workstation buffer remaining capacity coefficient, and the downstream workstation disturbance absorption coefficient. The upstream workstation buffer remaining capacity coefficient is obtained by subtracting the upstream workstation buffer occupancy rate from the value 1, and the downstream workstation disturbance absorption coefficient is obtained by subtracting the downstream workstation workstation inertia coefficient from the value 1.

3. The automated control method for hybrid production of office chairs based on fieldbus as described in claim 1, characterized in that, The damping coefficient is dynamically determined based on the buffer occupancy rate, and the adjusted release time is calculated by adding the current assembly beat to the product of the damping coefficient and the disturbance propagation coefficient of the trigger beat damping control.

4. The automated control method for hybrid production of office chairs based on fieldbus as described in claim 3, characterized in that, It also includes calculating the logically available capacity based on the actual cache usage, the maximum cache capacity, and the number of virtual cache tokens; the logically available capacity is calculated by subtracting the actual cache usage from the maximum cache capacity and then adding it to the number of virtual cache tokens. The number of virtual cache tokens is calculated by rounding up the ratio of the beat deviation to the reference beat.

5. The automated control method for hybrid production of office chairs based on fieldbus according to claim 4, characterized in that, It also includes, S6: The main control PLC receives the waiting resource identifier reported by each workstation PLC when the duration of the workstation waiting resource status exceeds the preset waiting threshold, and constructs and updates the resource waiting directed graph containing workstation nodes and cache nodes based on the waiting resource identifier; S7: The main control PLC performs loop detection in the resource waiting directed graph. If a closed loop structure is detected and the average cache occupancy rate of all cache nodes in the loop structure is greater than the preset average cache occupancy rate threshold in the loop, it is determined to be a quasi-deadlock state.

6. The automated control method for hybrid production of office chairs based on fieldbus according to claim 5, characterized in that, It also includes S8: After determining that the state is quasi-deadlock, the main control PLC identifies all the cache nodes in the closed loop structure, selects the node with the lowest cache occupancy rate as the target node, borrows logical capacity from the adjacent cache of the target node, updates the logical borrowable capacity of the target node according to the preset borrowing ratio and the physical remaining capacity of the adjacent cache, and sends it to the corresponding cache area control unit through the fieldbus to adjust the cache logical capacity.

7. The automated control method for hybrid production of office chairs based on fieldbus according to claim 6, characterized in that, The updated logically available capacity of the target node is calculated by adding the original logically available capacity of the target node to the available capacity. The available capacity is the product of the borrowing ratio and the remaining physical capacity of the adjacent cache. The remaining physical capacity of the adjacent cache is obtained by subtracting the actual number of adjacent caches from the maximum capacity of the adjacent cache.

8. The automated control method for hybrid production of office chairs based on fieldbus according to claim 7, characterized in that, It also includes, S9: The main control PLC calculates the edge density of the resource waiting directed graph, the edge density being the ratio of the actual number of directed edges in the resource waiting directed graph to the theoretical maximum number of directed edges, and updates the damping coefficient based on the edge density to obtain the updated damping coefficient, which is obtained by multiplying the damping coefficient before the update by the value 1 plus the sum of the edge density.

9. The automated control method for hybrid production of office chairs based on fieldbus according to claim 1, characterized in that, It also includes S10: The main control PLC inputs the cycle deviation of the workstation for the most recent specified number of cycles, the current buffer occupancy rate, and the current workstation inertia coefficient into the pre-trained long short-term memory network cycle prediction model to obtain the predicted value of the cycle deviation of the workstation for the next specified number of cycles; and triggers S5 based on the predicted value of the cycle deviation.

10. A fieldbus-based automated control system for hybrid production of office chairs, characterized in that, include: Main control PLC, multiple station PLCs, and buffer control unit based on fieldbus connection; The main control PLC includes: The communication establishment module is used to establish communication links with the PLCs and buffer control units of each workstation via fieldbus. The data storage module is used to establish a workstation data structure for each workstation. The workstation data structure includes the reference cycle time and the workstation inertia coefficient for that workstation, and stores all workstation data structures in a global data table. The workstation inertia coefficient is calculated based on the variance of the assembly cycle time of that workstation and the reference cycle time. The first data receiving module is used to receive the current assembly cycle time collected and uploaded in real time by the PLC of each workstation via the fieldbus, calculate the cycle time deviation based on the current assembly cycle time and the reference cycle time, and write the cycle time deviation into the corresponding workstation data structure. The second data receiving module is used to receive the actual number of caches occupied and the maximum capacity of the cache in real time collected and uploaded by the cache control unit through the fieldbus, calculate the cache occupancy rate based on the actual number of caches occupied and the maximum capacity of the cache, and write the cache occupancy rate into the corresponding workstation data structure. The matrix generation module is used to establish a workstation adjacency matrix based on the topology of the production line. The elements of the workstation adjacency matrix represent the upstream and downstream connection relationships between different workstations. Based on the buffer occupancy rate and workstation inertia coefficient, the module calculates the disturbance propagation coefficient in combination with the workstation adjacency matrix, generates the disturbance propagation matrix, and stores it. The disturbance control module is used to compare each disturbance propagation coefficient in the disturbance propagation matrix with a preset disturbance propagation coefficient threshold. If there is a disturbance propagation coefficient greater than the disturbance propagation coefficient threshold, the adjusted release time is calculated based on the disturbance propagation coefficient and the damping coefficient, and the adjusted release time is sent to the corresponding workstation PLC through the fieldbus. The workstation PLC is used to adjust the material release rhythm of the workstation according to the adjusted release time.

Images