Servo press control system and method based on time sensitive networking

Abstract

The application provides a time-sensitive network-based servo press control system and method, relates to the technical field of intelligent manufacturing and industrial automation control, and comprises the following steps: a time-sensitive network (TSN) is used to provide global time synchronization and deterministic data transmission; a main controller of a multi-axis servo control unit is used to issue motion control instructions and feeding and discharging control instructions through the TSN; a robot controller in a robot feeding and discharging unit is used to drive an industrial robot to perform sheet material taking and placing operations under the action of the feeding and discharging control instructions; a visual monitoring unit comprises an industrial camera connected to the TSN, is used to perform image acquisition, and transmits image data to an edge computing node through the TSN; the edge computing node is used to process the image data to generate compensation instructions, and feeds back the compensation instructions through the TSN; and an upper monitoring system is used to subscribe to real-time data published by the edge computing node, so that the control range of the servo press control system, the convenience of data transmission and interaction are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent manufacturing and industrial automation control, and particularly relates to a servo press control system and method based on a time-sensitive network. BACKGROUND

[0002] As the core equipment of modern intelligent manufacturing, servo presses are widely used in stamping production lines in the fields of automobiles, aerospace, home appliances, etc. due to their high flexibility, high efficiency, energy saving and environmental protection, etc.

[0003] In related technologies, traditional servo press control systems focus on press body control and adopt architectures of special buses, such as programmable logic controllers (PLC) or special motion controllers, which are connected to servo drives, I / O modules, etc. through industrial real-time Ethernet buses. Such architectures are relatively mature in early applications, but as the requirements of intelligent manufacturing for production flexibility, OT / IT fusion and system maintainability continue to increase, their inherent interaction defects gradually become prominent. Specifically, modern stamping production covers press body control, robot loading and unloading, visual monitoring and detection, upper monitoring system and other units. Under the traditional architecture, different network protocols are often used to transmit data, eventually forming multiple "information islands". The interconnection between different networks needs to rely on complex gateway devices, which not only increases the complexity, deployment cost and fault risk of the system, but also directly hinders the efficient and deterministic interaction of control instructions, visual data and information management data, resulting in a decrease in overall control reliability. Therefore, how to improve the control range, data transmission and interaction convenience of the servo press control system, and thus enhance the accuracy and reliability of control, has become a key problem to be solved at present. SUMMARY

[0004] The present application provides a servo press control system and method based on a time-sensitive network.

[0005] According to a first aspect of the present application, a servo press control system based on a time-sensitive network is provided, which comprises:

[0006] a time-sensitive network, a multi-axis servo control unit, a robot loading and unloading unit, a visual monitoring unit, an edge computing node and an upper monitoring system;

[0007] The time-sensitive network is connected to the multi-axis servo control unit, the robot loading and unloading unit, the visual monitoring unit, the edge computing node and the upper monitoring system, respectively, for providing global time synchronization and deterministic data transmission.

[0008] The multi-axis servo control unit comprises a main controller and a plurality of servo drives, the main controller is configured to issue motion control instructions to each servo drive and feeding / unloading control instructions to the robot feeding / unloading unit based on the global time synchronization through the time-sensitive network;

[0009] The robot feeding / unloading unit comprises an industrial robot and a robot controller, the robot controller is configured to drive the industrial robot to perform sheet material taking / placing operation under the action of the feeding / unloading control instructions;

[0010] The visual monitoring unit comprises an industrial camera connected to the time-sensitive network, the industrial camera is configured to perform image acquisition under the action of the global time synchronization trigger signal and transmit the acquired image data to the edge computing node through the time-sensitive network;

[0011] The edge computing node is configured to receive and process the image data, generate compensation instructions, and feed back the compensation instructions to the main controller or the robot controller through the time-sensitive network;

[0012] The upper monitoring system is configured to subscribe to real-time data published by the edge computing node.

[0013] Optionally, when transmitting data through the time-sensitive network, the periodic servo control stream transmitted between the main controller and each servo drive is of the first priority, the image data stream is of the second priority, the production information data stream transmitted between the edge computing node and the upper monitoring system is of the third priority, and the configuration and diagnosis data is of the fourth priority, wherein the fourth priority data is transmitted through a best-effort channel.

[0014] Optionally, the main controller is configured to start a stamping task based on a global synchronization clock and synchronously issue motion control instructions to each servo drive through the time-sensitive network at the beginning of each control cycle, wherein the motion control instructions comprise at least one of the following: position instructions, speed instructions, and torque instructions.

[0015] Each servo drive is configured to drive the corresponding servo motor to work based on the global synchronization clock after receiving the motion control instructions, update local state data within each control cycle, and send the local state data to the main controller through the time-sensitive network, wherein the local state data comprises at least one of the following: actual position, actual speed, actual torque, and drive state.

[0016] Optionally, the main controller is configured to synchronously issue motion control instructions to each servo drive through the time-sensitive network within a first time window of each control cycle.

[0017] Each of the servo drives is used to send the local status data to the main controller via the time-sensitive network within a second time window of each control cycle;

[0018] The main controller is also configured to receive local status data of each servo drive via the time-sensitive network within a third time window of each control cycle.

[0019] Optionally, the edge computing node is used to process the image data to determine the deviation between the actual position and the theoretical position of the current sheet material. Combined with the real-time motion state of the press, it generates trajectory compensation instructions or slider trajectory fine-tuning instructions for the robot loading and unloading unit. The trajectory compensation instructions are sent to the robot controller through the time-sensitive network, or the slider trajectory fine-tuning instructions are sent to the main controller through the time-sensitive network.

[0020] Optionally, the edge computing node is used to encapsulate the received trajectory compensation value, visual detection result, device status data, and raw motion control data, and publish them through the time-sensitive network using a publish-subscribe pattern.

[0021] Optionally, the visual monitoring unit may further include an image acquisition card;

[0022] The industrial camera is used to acquire image data under the action of the global time synchronization trigger signal, and after preprocessing the image data through the image acquisition card, it is sent to the edge computing node through the time-sensitive network in a streaming mode. The image data includes at least one of the following: sheet metal positioning image data of the mold area, dynamic image data of the stamping process, and workpiece quality inspection image data after stamping.

[0023] Optionally, the main controller is used to determine the multi-axis synchronization error value of the current cycle based on the actual position feedback of each servo axis in the current cycle, construct an objective function based on the multi-axis synchronization error value, the system state-space model, the network delay measurement value and the disturbance value, and determine the speed correction amount of each servo axis by minimizing the objective function.

[0024] According to a second aspect of this application, a servo press control method based on time-sensitive networks is provided, applied to any of the aforementioned servo press control systems based on time-sensitive networks, comprising:

[0025] Global time synchronization is performed based on a global synchronization clock, and a deterministic communication channel is established;

[0026] The main controller, based on a global synchronous clock, controls the multi-axis servo control unit and the robot loading and unloading unit to perform stamping operations through a time-sensitive network, and triggers the vision monitoring unit to acquire images.

[0027] The edge computing node receives and processes image data, and feeds back the generated compensation instructions to the main controller or robot controller through a time-sensitive network.

[0028] The main controller, in conjunction with the compensation command, sends motion control commands to each servo driver to drive the servo motor to move the slider to perform the stamping operation, and corrects the multi-axis synchronization error in each control cycle.

[0029] After stamping is completed, the visual monitoring unit collects image data, and the edge computing node processes the data and encapsulates and publishes the detection results and equipment status data.

[0030] The supervisory control system subscribes to and updates monitoring data to complete a single stamping cycle.

[0031] According to a third aspect of this application, an electronic device is provided, comprising: a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements any of the above-described time-sensitive network-based servo press control methods.

[0032] According to a fourth aspect of this application, a computer-readable storage medium is provided, on which computer program instructions are stored, which, when executed by a processor, implement any of the above-described time-sensitive network-based servo press control methods.

[0033] In summary, the servo press control system and method based on time-sensitive networks provided in this application have at least the following beneficial effects: The servo press control system based on time-sensitive networks can include a time-sensitive network, a multi-axis servo control unit, a robot loading and unloading unit, a vision monitoring unit, an edge computing node, and a host monitoring system. Each unit is connected to the time-sensitive network, thereby enabling global time synchronization and deterministic data transmission among the units. This improves the control range of the servo press control system, the convenience of data transmission and interaction, and also enhances the accuracy and reliability of the control. Attached Figure Description

[0034] To more clearly illustrate the specific embodiments of this application or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 A schematic diagram of a servo press control system based on a time-sensitive network provided for an embodiment of this application;

[0036] Figure 2 A flowchart illustrating a servo press control method based on a time-sensitive network, provided for embodiments of this application;

[0037] Figure 3 This is a structural diagram of an electronic device provided as an embodiment of the present application. Detailed Implementation

[0038] To make the above and other features and advantages of this application clearer, the application is further described below with reference to the accompanying drawings. It should be understood that the specific embodiments given herein are for the purpose of explanation to those skilled in the art, and are exemplary only, not restrictive.

[0039] In the following description, numerous specific details are set forth to provide a thorough understanding of this application. However, it will be apparent to those skilled in the art that the specific details are not required to practice this application. In other instances, well-known steps or operations have not been described in detail to avoid obscuring this application.

[0040] The time-sensitive network-based servo press control method provided in this application embodiment can be executed by the time-sensitive network-based servo press control device provided in this application embodiment, which can be configured in an electronic device.

[0041] refer to Figure 1 This application provides a servo press control system based on time-sensitive network, which includes: time-sensitive network, multi-axis servo control unit, robot loading and unloading unit, vision monitoring unit, edge computing node and upper monitoring system.

[0042] The time-sensitive network is connected to the multi-axis servo control unit, the robot loading and unloading unit, the vision monitoring unit, the edge computing node, and the upper-level monitoring system to provide global time synchronization and deterministic data transmission.

[0043] Time-Sensitive Networking (TSN) is a collection of IEEE 802.1 standards based on standard Ethernet. It enhances and optimizes the communication mechanisms of traditional Ethernet, supporting both high-real-time deterministic communication and best-effort general communication within the same network. Therefore, the servo press control system based on TSN provided in this application constructs a communication architecture centered on the TSN network. The multi-axis servo control unit, robot loading / unloading unit, vision monitoring unit, edge computing nodes, and upper-level monitoring system are all equipped with TSN communication interfaces, achieving unified data interaction and command transmission between all units of the system through the TSN network.

[0044] Optionally, the TSN network can be constructed from industrial switches supporting key TSN standards, forming the communication backbone of the entire system. It provides a global, high-precision time synchronization mechanism, establishing a unified clock reference for all devices across the network. Furthermore, through a time-aware shaper and a gating scheduling mechanism, it reserves dedicated, protected time windows for time-critical periodic control data streams, ensuring transmission with extremely low latency and zero packet loss, avoiding interference from other non-critical data streams, thereby achieving microsecond-level deterministic communication. Additionally, enhanced and automated flow reservation protocols simplify network configuration, supporting automatic data stream discovery and resource reservation, enabling the system to dynamically manage and guarantee the quality of service for critical data streams.

[0045] Therefore, the servo press control system based on Time-Sensitive Networking (TSN) provided in this application embodiment constructs a communication architecture with TSN network as its core. The multi-axis servo control unit, robot loading / unloading unit, vision monitoring unit, edge computing node, and upper-level monitoring system are all equipped with TSN communication interfaces, and unified data interaction and command transmission between all units of the entire system are achieved through the TSN network. The switches in the TSN network backbone, acting as the master clock for the Generalized Precision Time Protocol (GPP), can distribute synchronization clock signals to all terminal devices in the network, including the main controller, all servo drives, robot controllers, industrial cameras, edge computing nodes, etc. All devices adjust their local clocks accordingly to achieve time synchronization.

[0046] In addition, a multi-axis servo control unit may include a main controller and multiple servo drives.

[0047] Among them, there are usually multiple servo drives. The specific number can be determined by the model specifications and control requirements of the servo press. For example, for a press that uses a slider four-corner drive, four servo drives can usually be configured. For complex multi-link or redundant drive structures, the number may be increased accordingly. This application does not limit this.

[0048] The main controller can be used to send motion control commands to each servo driver and loading / unloading control commands to the robot loading / unloading unit based on global time synchronization and time-sensitive network.

[0049] The motion control command may include at least one of position command, speed command and torque command, and the loading and unloading control command may include loading control command, unloading control command, etc., which are not limited in this application.

[0050] Understandably, each servo driver is connected to a servo motor, and the output shaft of the servo motor is the servo axis of the servo press. After receiving the motion control command sent by the main controller through the TSN network, the servo driver can drive the corresponding servo motor to move the servo axis, thereby achieving accurate control of the press actuator.

[0051] In addition, the robot loading and unloading unit may include an industrial robot and a robot controller. The robot controller can be used to drive the industrial robot to perform sheet metal picking and placing operations under the action of loading and unloading control commands.

[0052] The loading operation can be understood as transporting the sheet metal to be stamped from the raw material table to the designated position of the press mold, and the unloading operation can be understood as removing the stamped workpiece from the mold and transporting it to the finished product table. Therefore, in this embodiment, the robot controller can establish a communication connection with the main controller via a TSN network, and based on the received loading and unloading control commands and a global time synchronization signal, drive the industrial robot to perform the loading and unloading operations of the sheet metal according to a preset trajectory and gripping parameters. After the industrial robot completes the operation, the robot controller can send an operation completion signal back to the main controller, thereby achieving time-series coordination with the multi-axis servo control unit.

[0053] In addition, the visual monitoring unit may include an industrial camera connected to a time-sensitive network. The industrial camera can be used to acquire images under the action of a global time synchronization trigger signal, and transmit the acquired image data to the edge computing node through the time-sensitive network.

[0054] The triggering methods for industrial cameras can be varied, including position triggering and time triggering. For example, the main controller can send a trigger signal when the slider moves to a preset trigger position, or the main controller can send a trigger signal at preset time intervals based on a global synchronization clock. Upon receiving the aforementioned trigger signal, the industrial camera can complete multi-stage image acquisition according to a preset acquisition frequency. This application does not limit the specific methods used.

[0055] In addition, edge computing nodes can be used to receive and process image data, generate compensation instructions, and feed the compensation instructions back to the main controller or robot controller through a time-sensitive network. The upper-level monitoring system can be used to subscribe to real-time data published by the edge computing nodes.

[0056] Optionally, when transmitting data via a time-sensitive network, the periodic servo control stream transmitted between the main controller and each servo driver has the first priority, the image data stream has the second priority, the production information data stream transmitted between the edge computing node and the upper monitoring system has the third priority, and the equipment configuration and diagnostic data has the fourth priority. The fourth priority data is transmitted through the best-effort channel.

[0057] The first priority is the highest, followed by the second, then the third, and finally the fourth.

[0058] Based on system communication requirements, communication channels are pre-configured or automatically negotiated via protocols to allocate priorities for different types of data streams within the TSN network. For example, the highest priority can be assigned to periodic servo control streams, meaning motion control commands transmitted between the main controller and each servo driver can have the first priority. Fixed, periodic time windows can also be reserved in the gating scheduler. Within this window, the switch only forwards the control stream, ensuring uninterrupted, low-jitter transmission.

[0059] In addition, the image data stream can be a video image stream sent from an industrial camera to an edge computing node. It can be assigned a medium-to-high priority, i.e., the second priority, and bandwidth can be reserved for it, thereby effectively ensuring the real-time transmission of large-volume image data.

[0060] In addition, for production information data streams, a third priority can be assigned to them. These can be production information data sent from edge computing nodes to the upper-level monitoring system, or subscription and publication data, etc. Fixed bandwidth can be allocated to them to ensure the real-time performance and reliability of production data uploads.

[0061] Additionally, other non-real-time configuration and diagnostic information can be assigned a fourth priority and transmitted through the best-effort channel in the TSN network.

[0062] In this application, priority can be set for various data streams in any way, and no restrictions are imposed on this.

[0063] Therefore, in this embodiment of the application, when transmitting data through a time-sensitive network, different priorities and transmission resources can be allocated to different data streams, so that all kinds of data can be transmitted in an orderly manner in the same TSN network, avoiding data transmission chaos and conflicts, and improving the efficiency and accuracy of data transmission in the entire servo press control system.

[0064] Optionally, the main controller can be used to start the stamping task based on a global synchronization clock, and at the beginning of each control cycle, send motion control commands to each servo driver synchronously through a time-sensitive network. Each servo driver can be used to drive the corresponding servo motor to work based on the global synchronization clock after receiving the motion control command, and update the local status data in each control cycle, and then send the local status data to the main controller through the time-sensitive network.

[0065] Understandably, the main controller can calculate the target position, target speed, or target torque to be achieved in the next control cycle for each servo axis of the press, such as the slide drive axis, ejector axis, and pressure axis, based on the stamping process requirements, compensation instructions fed back from edge computing nodes, and a global synchronization clock. This calculation is then encapsulated as motion control instructions and synchronously sent to the corresponding servo driver via the TSN network. Upon receiving the instructions, the servo driver drives the servo motor to move the servo axis according to the target parameters, achieving precise closed-loop control of position, speed, or torque.

[0066] In addition, local status data can be real-time operating data of itself and its corresponding servo axis collected and generated by each servo drive within a control cycle. This data can include at least one of the following: actual position, actual speed, actual torque, and drive status. For example, during the operation of a servo motor, the servo drive collects the actual position, actual speed, and actual torque of the corresponding servo axis in real time through built-in detection elements such as motor encoders and current sensors. At the same time, it monitors its own operating status and determines whether there is an overload, a fault, or normal communication. This data can then be integrated into local status data and transmitted back to the main controller via the TSN network within a preset time window in each control cycle.

[0067] Optionally, the main controller can be used to synchronously send motion control commands to each servo drive through a time-sensitive network within the first time window of each control cycle. Each servo drive can be used to send local status data to the main controller through a time-sensitive network within the second time window of each control cycle. The main controller can also be used to receive local status data from each servo drive through a time-sensitive network within the third time window of each control cycle.

[0068] Understandably, within the same control cycle, the first, second, and third time windows are arranged sequentially, with the first time window preceding the second, and the second preceding the third. The duration of each time window is pre-configured based on the data transmission volume and fixed in the TSN switch's gating schedule table. This orderly division of time windows enables conflict-free transmission of motion control commands and local status data uploads, effectively avoiding data collisions and transmission chaos within the same communication link, and ensuring the timing coordination and deterministic data interaction of multi-axis servo control.

[0069] Optionally, edge computing nodes can be used to process image data to determine the deviation between the actual position and the theoretical position of the current sheet metal. Then, combined with the real-time motion status of the press, trajectory compensation instructions or slider trajectory fine-tuning instructions for the robot loading and unloading unit can be generated. The trajectory compensation instructions can then be sent to the robot controller via a time-sensitive network, or the slider trajectory fine-tuning instructions can be sent to the main controller via a time-sensitive network.

[0070] The edge computing node first receives image data transmitted by the visual monitoring unit via TSN. It then processes the image data, such as by using grayscale conversion and Gaussian filtering to eliminate ambient light interference. Next, it uses edge detection and feature matching algorithms to locate the sheet metal outline, extracting the sheet metal's actual position coordinates. These actual coordinates are then compared with preset theoretical position coordinates to calculate the positional deviations, such as X-axis offset, Y-axis offset, and rotation angle deviation. Furthermore, the edge computing node can synchronously acquire real-time motion status data of the press via the TSN network, such as the current stamping stage, the real-time position of the slider, and the servo axis running speed. Combined with the aforementioned positional deviations, it performs collaborative analysis. If the deviation originates from sheet metal loading and positioning errors, and the current stage is the pre-stamping processing stage, it can generate trajectory compensation instructions adapted to the industrial robot's motion trajectory. If the deviation originates from slider trajectory drift during stamping, and the current stage is the stamping execution stage, it can generate slider trajectory fine-tuning instructions to correct the servo axis motion parameters. Afterwards, the edge computing node can send trajectory compensation commands to the robot controller or send slider trajectory fine-tuning commands to the main controller through the TSN network.

[0071] Understandably, the aforementioned compensation or fine-tuning instructions all carry a global synchronization timestamp, which can effectively ensure precise synchronization between robot trajectory correction, slider motion adjustment, and the overall stamping sequence of the press, thereby achieving dynamic optimization of sheet metal positioning accuracy and stamping forming accuracy.

[0072] Optionally, edge computing nodes can be used to encapsulate and process the received trajectory compensation values, visual detection results, device status data, and raw motion control data, and publish them through the time-sensitive network using a publish-subscribe pattern.

[0073] Optionally, edge computing nodes can continuously receive and aggregate data from various units in the system via the TSN network. This data can include sheet metal position trajectory compensation values ​​calculated through visual analysis, visual inspection results of stamped workpieces (such as dimensional compliance judgments, surface defect types and locations), local status data uploaded by each servo drive (such as actual servo axis position, speed, torque, and drive fault alarm information), and raw motion control data issued by the main controller (such as target position commands, speed commands, and torque commands). This data can be organized and encapsulated into a structured information model with clear semantics, according to the OPC UA unified architecture standard. For example, the current value of servo drive A can be encapsulated as a variable node in the information model, defining its name, data type, engineering units, and other information.

[0074] After data encapsulation, edge computing nodes can uniformly publish the encapsulated fused data packets via the TSN network based on the publish / subscribe mode of the Open Platform Communications Unified Architecture (OPC UA). During the publication process, the fused data packets are transmitted according to a preset third-priority data stream and carry a globally synchronized timestamp to ensure data time-series consistency, allowing upper-level monitoring systems or other authorized network nodes to subscribe independently as needed. Upper-level monitoring systems, such as Manufacturing Execution Systems (MES) and Supervisory Control and Data Acquisition (SCADA) systems, as subscribers, can obtain real-time full data of the entire press system by subscribing to the aforementioned data packets, thus providing complete and accurate data source support for production process visualization monitoring, process parameter traceability and optimization, and equipment fault early warning.

[0075] Optionally, industrial cameras can be used to acquire image data under the action of a global time synchronization trigger signal, and after preprocessing the image data through an image acquisition card, the data can be sent to edge computing nodes via a time-sensitive network in a streaming manner.

[0076] The image data can include at least one of the following: sheet metal positioning image data in the mold area, dynamic image data of the stamping process, and quality inspection image data of the stamped workpiece. Specifically, the sheet metal positioning image data in the mold area can be used to verify the accuracy of the robot's placement of the sheet metal after loading; the dynamic image data of the stamping process can be used for process monitoring or analysis of dynamic processes such as mold closing and material forming; and the quality inspection image data of the stamped workpiece can be used for preliminary defect detection of the finished product, such as scratches and deformation, thereby achieving online quality judgment.

[0077] Understandably, industrial cameras can acquire image data under the trigger of a global time synchronization signal. After acquisition, the image data can be transmitted to a matching image acquisition card, which will then perform preprocessing on the image data, such as noise reduction filtering, image enhancement, and format standardization conversion. This reduces the image processing load on subsequent edge computing nodes and improves data processing efficiency.

[0078] In addition, the pre-processed image data usually carries a global timestamp (TSN) corresponding to the acquisition time. It can be sent to the edge computing node via TSN in a streaming mode according to the preset second priority data stream rules, thereby effectively ensuring the security and reliability of data transmission.

[0079] Optionally, the main controller can be used to determine the multi-axis synchronization error value for the current cycle based on the actual position feedback of each servo axis in the current cycle. Then, based on the multi-axis synchronization error value, the system state-space model, the network delay measurement value, and the disturbance value, an objective function is constructed, and the speed correction amount for each servo axis is determined by minimizing the objective function.

[0080] Wherein, the actual position feedback of each servo axis in the current cycle can be p i (k), where i is a positive integer representing the servo axis number, and the synchronization error can be expressed as: e sync (k)=max(p i (k))-min(p i (k)), the network latency measurement can be expressed as External disturbances can be represented as .

[0081] Furthermore, the synchronization error state vector can be represented as: ;

[0082] in, This indicates the deviation of the servo axis from the ideal trajectory, i.e., the synchronization error state.

[0083] Furthermore, based on the control period T, the state update dynamic equation for the synchronization error can be expressed as:

[0084] ;

[0085] Among them, u i (k) represents the speed correction amount for the i-th servo axis, d i (k) represents the comprehensive disturbance term during the synchronization process of the i-th servo axis.

[0086] Furthermore, the comprehensive disturbance term can be further expressed as:

[0087] ;

[0088] Where T is the control period. For mechanical coupling systems, N represents the network latency factor affecting position, j represents the number of the nearest neighbor servo axis that is directly mechanically connected to the i-th servo axis, and N represents the position of the nearest neighbor servo axis. i Let i be the set of nearest neighbor axes directly mechanically connected to the i-th servo axis. This is a measurement of latency in the TSN network. External disturbances such as frictional changes and sudden load changes.

[0089] Then, the error state of the multi-axis servo system can be uniformly modeled as a state equation:

[0090] ;

[0091] Among them, I N It is an N-dimensional identity matrix, where N is the total number of servo axes. Let be the graph Laplace matrix, where,

[0092] ;

[0093] To simplify the description, we can define matrices / vectors:

[0094] ;

[0095] The state equation can then be simplified to:

[0096] .

[0097] Subsequently, regarding high-precision control of multi-axis synchronization errors, in this embodiment, synchronization errors can be suppressed by using the optimal control input within a preset calculation window. An optimization target can be constructed first, and the time-domain H can be predicted. p For example, you can choose 2-6 steps to predict the future H. p System state, control time domain H c If 2-3 steps are optional, only the first H is optimized. cThe control inputs for each step are as follows: u(k), u(k+1), ..., u(k+H) c -1), and the subsequent control inputs remain unchanged or are set to 0. The objective function can be as follows:

[0098] ;

[0099] Among them, u j (k+i-1|k) represents the velocity correction of the j-th axis at time k+i-1.

[0100] To control the increment, it can be u j (k+i-1|k) and u j The difference of (k+i-2|k), For synchronization error weights, To control energy weights, For smoothness weights.

[0101] The problem can then be transformed into a quadratic programming (QP) problem, solved cycle-by-cycle using a lightweight QP solver such as OSQP, and the output is the velocity correction u. i (k) is added to the instruction in the next cycle.

[0102] The above process will be briefly explained below using the example shown in Table 1.

[0103] Table 1. Parameter Table for Multi-Axis Servo Synchronous Control

[0104]

[0105] First, the Laplace matrix It can be represented as follows:

[0106] ;

[0107] The state transition matrix can be represented as:

[0108] ;

[0109] The control matrix can be represented as:

[0110] .

[0111] For the current state: x(k) = [0.0, 1.0, 2.0, 0.5];

[0112] Prediction can be performed without a control baseline, whereby,

[0113] The prediction x(1|0) can be expressed as:

[0114] ;

[0115] The prediction x(2|0) can be expressed as:

[0116] ;

[0117] The objective function can then be minimized:

[0118] ;

[0119] The constraint can be expressed as: -0.5≤u i (k)≤0.5, and then solve for it to obtain the speed correction amount for each servo axis:

[0120] ;

[0121] It can be seen that shafts 1 and 4 maintain their original speed, while shafts 2 and 3 slow down.

[0122] The optimized control values ​​can then be applied and calculations performed.

[0123] Calculate x(2|0):

[0124] ;

[0125] Calculate x(2|1):

[0126] .

[0127] The following table compares the error improvement effects before and after the correction.

[0128] Table 2 Comparison of the improvement effects of multi-axis synchronization error

[0129]

[0130] As illustrated in the above example, in determining the speed correction amount for each servo axis, the system state for multiple future steps can be predicted in advance through a preset prediction time domain. This allows for the output of control quantities for compensation before the error expands, suppressing error accumulation in multi-axis coupled scenarios from the source. Simultaneously, by optimizing the objective function through multiple objectives, indicators such as synchronization error, control energy, and control smoothness are incorporated into the solution. This not only stabilizes and suppresses multi-axis synchronization errors to ensure stamping accuracy but also limits the abrupt changes in control quantities, reducing impact loads on mechanical components and achieving a balance between accuracy, stability, and equipment lifespan. Furthermore, by explicitly embedding physical constraints on control quantities, such as the speed correction range, during the optimization process, equipment failures caused by control quantity overload can be effectively avoided, improving system operational safety, significantly enhancing anti-interference capabilities, and making it more suitable for complex working conditions in actual production.

[0131] It is understandable that the time-sensitive network-based servo press control system provided in this application can be divided into three levels in practical applications: enterprise management, data processing, and production execution.

[0132] like Figure 1 As shown, the enterprise management level can be an IT / supervisory monitoring system, which may include Manufacturing Execution System (MES), Enterprise Resource Planning (ERP), etc., connected to the core switch via a common Ethernet connection. The data processing level can be an edge computing node, which can connect IT and OT to achieve real-time data processing and closed-loop control. The production execution layer can be an OT / press system, which serves as the actual production site unit, connected to a TSN switch via a TSN connection, and is responsible for executing the stamping operations of the servo press. It may include a multi-axis servo control unit for driving the multi-axis motion of the press, the press body as the stamping actuator, a vision monitoring unit for acquiring images of sheet metal or workpieces, a robotic loading and unloading unit for automatic sheet metal handling, a PLC, and a human-machine interface for on-site operation and status display.

[0133] For example, the MES or ERP at the IT layer can issue production work orders. Then, the edge computing node receives images and equipment status data from the OT layer, performs visual analysis and error calculation, and generates compensation instructions. After that, the PLC and servo units at the OT layer execute stamping and loading / unloading operations based on the TSN synchronous clock, and receive compensation instructions from the edge computing to correct the actions. The edge computing node then encapsulates the production data and uploads it to the IT layer via ordinary Ethernet to realize the visual monitoring and data traceability of the production process.

[0134] It should be noted that the above examples are merely illustrative and the number or function of each unit can be adjusted according to actual needs. They should not be taken as limitations on the units in the servo press control system of this application.

[0135] In this embodiment, the servo press control system based on time-sensitive networking can include a time-sensitive network, a multi-axis servo control unit, a robot loading and unloading unit, a vision monitoring unit, an edge computing node, and a host monitoring system. Each unit is connected to the time-sensitive network, thereby enabling global time synchronization and deterministic data transmission among the units. This improves the control range, data transmission and interaction convenience of the servo press control system, and also enhances the accuracy and reliability of the control.

[0136] like Figure 2 As shown, the servo press control method based on time-sensitive networks may include the following steps:

[0137] Step 201: Perform global time synchronization based on the global synchronization clock and establish a deterministic communication channel.

[0138] Global time synchronization can be achieved through time-sensitive networking. The TSN switch can act as the master clock source, broadcasting clock synchronization signals to multi-axis servo control units, robot loading and unloading units, vision monitoring units, edge computing nodes, and upper-level monitoring systems, so that the local clocks of all access nodes are synchronized with the master clock, forming a unified global synchronization clock.

[0139] In addition, based on preset data flow priority rules, a gating scheduling table can be configured in the TSN switch to divide non-overlapping dedicated time windows and communication bandwidths for different types of data flows.

[0140] Specifically, a first-priority dedicated time window is allocated to the periodic control flow of the multi-axis servo control unit, a second-priority bandwidth is allocated to the image data flow of the vision monitoring unit, a third-priority channel is allocated to the production information data flow, and a fourth-priority best-effort channel is allocated to the equipment configuration diagnostic data. This configuration can effectively ensure that various types of data are transmitted without conflict and with low latency, thus building a communication link for deterministic transmission.

[0141] Step 202: Based on the global synchronization clock, the main controller controls the multi-axis servo control unit and the robot loading and unloading unit to perform stamping operations through a time-sensitive network, and triggers the vision monitoring unit to acquire images.

[0142] The main controller determines the start time of a single stamping cycle based on a global synchronization clock. Then, through the first priority communication channel of the TSN network, it synchronously sends motion control commands to each servo driver of the multi-axis servo control unit. These commands include the target position, speed, and torque parameters for each servo axis. Simultaneously, it sends loading / unloading control commands to the robot controller of the robot loading / unloading unit, including parameters such as the sheet metal gripping position, transfer path, and placement accuracy requirements. Upon receiving the commands, the robot controller drives the industrial robot to perform sheet metal picking and placing operations, accurately transferring the sheet metal to be stamped to the mold area and completing its positioning. Simultaneously, the main controller sends a global time synchronization trigger signal to the industrial camera of the vision monitoring unit, triggering the camera to start image acquisition at a preset frame rate. The acquired image data covers the sheet metal positioning image in the mold area, providing a visual data source for subsequent position deviation analysis.

[0143] Step 203: The edge computing node receives and processes the image data, and feeds back the generated compensation instructions to the main controller or robot controller through a time-sensitive network.

[0144] The industrial camera in the vision monitoring unit transmits the acquired image data to a matching image acquisition card. After preprocessing such as noise reduction filtering and image enhancement, the data is sent to the edge computing node via the second priority data stream channel of the TSN network. Upon receiving the image data, the edge computing node performs edge detection and feature matching on the sheet metal positioning image, calculating the deviation between the actual and theoretical positions of the sheet metal. Simultaneously, it performs collaborative analysis by combining this data with real-time press motion data synchronously acquired from the main controller. If the deviation originates from positioning errors during robot loading and unloading, the edge computing node generates a robot trajectory compensation command, which is sent to the robot controller via the TSN network to correct the robot's subsequent loading and unloading trajectories. If the deviation originates from servo axis motion trajectory drift, the edge computing node generates a slider trajectory fine-tuning command, which is sent to the main controller via the TSN network, providing a correction basis for subsequent multi-axis synchronous control.

[0145] Step 204: The main controller, in conjunction with the compensation command, sends motion control commands to each servo driver to drive the servo motor to move the slider to perform the stamping operation, and corrects the multi-axis synchronization error in each control cycle.

[0146] In this process, after receiving the slider trajectory fine-tuning command from the edge computing node, the main controller can update the target parameters in the motion control command based on the preset stamping process parameters. Then, based on the global synchronization clock, within the first time window of each control cycle, it synchronously sends the updated motion control command to each servo driver via the TSN network. Each servo driver executes the command based on the time reference of the global synchronization clock, driving the corresponding servo motor to drive the press slider, top cylinder, and other actuators to complete actions such as mold closing and stamping according to the preset trajectory. Simultaneously, within each control cycle, the servo driver collects its own and the servo axis's local state data and sends it back to the main controller. The main controller uses a multi-axis coupled disturbance model based on the Graph Laplace matrix and combines it with a model predictive control algorithm to solve for the optimal speed correction, which is then superimposed on the motion control command of the next cycle to correct multi-axis synchronization errors in real time, ensuring the coordinated accuracy of each servo axis during the stamping process.

[0147] The steps and processes for solving the optimal speed correction amount can be referred to the descriptions of the various embodiments of this application, and will not be repeated here.

[0148] Step 205: After stamping is completed, the visual monitoring unit collects image data, and the edge computing node processes the image data and encapsulates and publishes the detection results and equipment status data.

[0149] After the press completes the stamping action and the slide returns to a safe position, the main controller can send a global time synchronization trigger signal to the vision monitoring unit again, triggering the industrial camera to acquire quality inspection image data of the stamped workpiece. This data can include key information such as the workpiece's appearance dimensions and surface condition. After receiving the quality inspection image data, the edge computing node identifies scratches, deformations, and other defects on the workpiece surface using a defect detection algorithm, generating quality inspection results. It can also integrate various types of information, such as trajectory compensation values, raw motion control data, and equipment status data uploaded by each servo drive, from the current stamping cycle. These are standardized and encapsulated according to a preset communication protocol, and based on a publish-subscribe model, the encapsulated fused data packet is published to the TSN network through a third-priority communication channel for authorized nodes to subscribe to and obtain.

[0150] Step 206: The supervisory control system subscribes to and updates the monitoring data to complete a single stamping cycle.

[0151] The upper-level monitoring system can subscribe to the fusion data packets published by the edge computing nodes in the TSN network, and then parse and store the data to update the content displayed on the production monitoring interface in real time. The displayed information includes the workpiece quality judgment results of a single stamping, multi-axis synchronization error curves, equipment operating status parameters, etc.

[0152] Optionally, the supervisory control system can archive all data from this stamping cycle to a database for subsequent production process optimization, equipment fault warning, and production data traceability. After completing a single stamping cycle from sheet metal loading, stamping, quality inspection, and data archiving, the system waits for the next global synchronization clock trigger signal to enter the next round of stamping operations.

[0153] In this embodiment, global time synchronization can be performed first based on a global synchronization clock, and a deterministic communication channel can be established. The main controller, based on the global synchronization clock, controls the multi-axis servo control unit and the robot loading and unloading unit to perform stamping operations through a time-sensitive network, and triggers the vision monitoring unit to acquire images. The edge computing node receives and processes the image data, and feeds back the generated compensation instructions to the main controller or robot controller through the time-sensitive network. The main controller, in combination with the compensation instructions, sends motion control instructions to each servo driver to drive the servo motor to drive the slider to perform stamping operations. In each control cycle, the multi-axis synchronization error is corrected. After stamping is completed, the vision monitoring unit acquires image data, the edge computing node processes it, encapsulates and publishes the detection results and equipment status data, and the upper-level monitoring system subscribes to and updates the monitoring data to complete a single stamping cycle. Therefore, a global time synchronization and hierarchical priority deterministic communication channel is first established through a time-sensitive network. Then, the main controller links the multi-axis servo and robot loading and unloading units to start the stamping operation and trigger visual acquisition. The image data is processed by the edge computing node to generate compensation instructions to complete the correction. Finally, visual quality inspection and data release are used to achieve full-process control of production, thereby improving the control range of the servo pressure control system, the convenience of data transmission and interaction, and thus improving the accuracy and reliability of control.

[0154] It should be understood that the specific features, operations, and details described herein with respect to the methods of this application can also be similarly applied to the apparatus and system of this application, or vice versa. Furthermore, each step of the methods of this application described above can be performed by a corresponding component or unit of the apparatus or system of this application.

[0155] It should be understood that the various modules / units of the device of this application can be implemented wholly or partially through software, hardware, firmware, or a combination thereof. Each module / unit can be embedded in the processor of the electronic device in hardware or firmware form or independent of the processor, or it can be stored in the memory of the electronic device in software form for the processor to call to execute the operation of each module / unit. Each module / unit can be implemented as an independent component or module, or two or more modules / units can be implemented as a single component or module.

[0156] like Figure 3As shown, this application provides an electronic device 300, which includes a processor 301 and a memory 302 storing computer program instructions. When the processor 301 executes the computer program instructions, it implements the steps of the aforementioned time-sensitive network-based servo press control method. This electronic device 300 can be broadly categorized as a server, terminal, or any other electronic device with the necessary computing and / or processing capabilities.

[0157] In one embodiment, the electronic device 300 may include a processor, memory, network interface, communication interface, etc., connected via a system bus. The processor of the electronic device 300 can be used to provide necessary computing, processing, and / or control capabilities. The memory of the electronic device 300 may include non-volatile storage media and internal memory. The non-volatile storage media may store an operating system, computer programs, etc. The internal memory can provide an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface and communication interface of the electronic device 300 can be used to connect and communicate with external devices via a network. When the computer program is executed by the processor, it performs the steps of the method of this application.

[0158] This application provides a computer-readable storage medium storing computer program instructions, which, when executed by a processor, implement the aforementioned time-sensitive network-based servo press control method.

[0159] Those skilled in the art will understand that the method steps of this application can be performed by a computer program instructing related hardware, such as electronic device 300 or a processor. The computer program can be stored in a non-transitory computer-readable storage medium, and its execution causes the steps of this application to be performed. Depending on the context, any reference herein to memory, storage, or other media may include non-volatile or volatile memory. Examples of non-volatile memory include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, magnetic tape, floppy disk, magneto-optical data storage device, optical data storage device, hard disk, solid-state drive, etc. Examples of volatile memory include random access memory (RAM), external cache memory, etc.

[0160] The technical features described above can be combined arbitrarily. Although not all possible combinations of these technical features are described, any combination of these technical features should be considered to be covered by this specification, provided that such combination does not contain contradictions.

[0161] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A servo press control system based on time-sensitive networks, characterized in that, include: Time-sensitive network, multi-axis servo control unit, robot loading and unloading unit, vision monitoring unit, edge computing node and upper-level monitoring system; The time-sensitive network is connected to the multi-axis servo control unit, the robot loading and unloading unit, the vision monitoring unit, the edge computing node, and the upper-level monitoring system, respectively, to provide global time synchronization and deterministic data transmission. The multi-axis servo control unit includes a main controller and multiple servo drivers. The main controller is used to send motion control commands to each of the servo drivers and to send loading and unloading control commands to the robot loading and unloading unit through the time-sensitive network based on the global time synchronization. The robot loading and unloading unit includes an industrial robot and a robot controller. The robot controller is used to drive the industrial robot to perform sheet metal picking and placing operations under the action of the loading and unloading control command. The visual monitoring unit includes an industrial camera connected to the time-sensitive network. The industrial camera is used to acquire images under the action of a global time synchronization trigger signal and transmit the acquired image data to the edge computing node through the time-sensitive network. The edge computing node is used to receive and process the image data, generate compensation instructions, and feed the compensation instructions back to the main controller or the robot controller through the time-sensitive network. The upper-level monitoring system is used to subscribe to the real-time data published by the edge computing nodes.

2. The control system as described in claim 1, characterized in that, When data is transmitted through the time-sensitive network, the periodic servo control stream transmitted between the main controller and each of the servo drives has the first priority, the image data stream has the second priority, the production information data stream transmitted between the edge computing node and the upper monitoring system has the third priority, and the configuration and diagnostic data has the fourth priority. The fourth priority data is transmitted through the best-effort channel.

3. The control system as described in claim 1, characterized in that, The main controller is used to start the stamping task based on a global synchronization clock, and at the beginning of each control cycle, it synchronously sends motion control commands to each of the servo drives through the time-sensitive network. The motion control commands include at least one of the following: position command, speed command, and torque command. Each servo driver is used to drive the corresponding servo motor to work based on a global synchronization clock after receiving the motion control command, update local status data in each control cycle, and send the local status data to the main controller through the time-sensitive network. The local status data includes at least one of the following: actual position, actual speed, actual torque, and driver status.

4. The control system as described in claim 3, characterized in that, The main controller is used to synchronously send motion control commands to each of the servo drives through the time-sensitive network within the first time window of each control cycle. Each of the servo drives is used to send the local status data to the main controller via the time-sensitive network within a second time window of each control cycle; The main controller is also configured to receive local status data of each servo drive via the time-sensitive network within a third time window of each control cycle.

5. The control system as described in claim 1, characterized in that, The edge computing node is used to process the image data to determine the deviation between the actual position and the theoretical position of the current sheet material. Combined with the real-time motion state of the press, it generates trajectory compensation instructions or slider trajectory fine-tuning instructions for the robot loading and unloading unit. The trajectory compensation instructions are sent to the robot controller through the time-sensitive network, or the slider trajectory fine-tuning instructions are sent to the main controller through the time-sensitive network.

6. The control system as described in claim 1, characterized in that, The edge computing node is used to encapsulate the received trajectory compensation values, visual detection results, device status data, and raw motion control data, and publish them through the time-sensitive network using a publish-subscribe pattern.

7. The control system as described in claim 1, characterized in that, The visual monitoring unit also includes an image acquisition card; The industrial camera is used to acquire image data under the action of the global time synchronization trigger signal, and after preprocessing the image data through the image acquisition card, it is sent to the edge computing node through the time-sensitive network in a streaming mode. The image data includes at least one of the following: sheet metal positioning image data of the mold area, dynamic image data of the stamping process, and workpiece quality inspection image data after stamping.

8. The control system as described in claim 1, characterized in that, The main controller is used to determine the multi-axis synchronization error value of the current cycle based on the actual position feedback of each servo axis in the current cycle. Based on the multi-axis synchronization error value, the system state space model, the network delay measurement value and the disturbance value, an objective function is constructed, and the speed correction amount of each servo axis is determined by minimizing the objective function.

9. A servo press control method based on time-sensitive networks, characterized in that, The servo press control system based on time-sensitive networks as described in any one of claims 1-8 includes: Global time synchronization is performed based on a global synchronization clock, and a deterministic communication channel is established; The main controller, based on a global synchronous clock, controls the multi-axis servo control unit and the robot loading and unloading unit to perform stamping operations through a time-sensitive network, and triggers the vision monitoring unit to acquire images. The edge computing node receives and processes image data, and feeds back the generated compensation instructions to the main controller or robot controller through a time-sensitive network. The main controller, in conjunction with the compensation command, sends motion control commands to each servo driver to drive the servo motor to move the slider to perform the stamping operation, and corrects the multi-axis synchronization error in each control cycle. After stamping is completed, the visual monitoring unit collects image data, and the edge computing node processes the data and encapsulates and publishes the detection results and equipment status data. The supervisory control system subscribes to and updates monitoring data to complete a single stamping cycle.

10. An electronic device, characterized in that, The electronic device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, it implements the servo press control method based on time-sensitive networks as described in claim 9.

Images