A flange production line multi-machine collaborative scheduling method and system based on digital twinning

Abstract

The present application belongs to the technical field of production line coordinated control, and particularly relates to a flange production line multi-machine collaborative scheduling method and system based on digital twinning, which comprises the following steps: acquiring the real-time driving current and vibration acceleration amplitude of the main shaft of a machine tool to construct a retardation coefficient; combining the number of flanges to be processed in the buffer zone of a conveying belt to acquire the blocking tendency index of the machine tool; based on the real-time image stream of the conveying belt area and the real-time running speed of the conveying belt, the grasping potential energy of a mechanical arm is acquired; according to the blocking tendency indexes of the current machine tool and the downstream machine tool, the comprehensive shunt impedance is calculated by using the grasping potential energy, and a grasping or release instruction is generated according to the comprehensive shunt impedance. The present application realizes the adaptive balance of the production line load under the premise of guaranteeing the physical safety of the equipment, effectively prevents logistics deadlock and improves the production efficiency.

Description

Technical Field

[0001] This invention relates to the field of production line coordination and control technology. More specifically, this invention relates to a method and system for multi-machine collaborative scheduling of a flange production line based on digital twins. Background Technology

[0002] In automated flange machining production lines, materials are typically transported to various processing units by conveyor belts, and loading and unloading operations are performed by robotic arms. The distribution control system rationally allocates materials based on the real-time status of each processing unit, preventing material accumulation at local workstations due to overload or equipment idling due to insufficient material supply, thereby achieving load balance and maximizing output across the entire production line.

[0003] In related technologies, for example, Chinese patent application CN120972815A discloses a collaborative control method for copper clad laminate production lines based on digital twins. This method includes: constructing a multi-level dataset with dynamic virtual-real mapping through full-process digital twin modeling, process data normalization, and multi-level feature labeling to achieve high consistency between physical production line conditions and the simulation system; combining autonomous scheduling intelligent agents to achieve real-time perception and joint feature mining of multi-dimensional data such as equipment load, health, and energy consumption, improving the production line's real-time adaptive scheduling capability for sudden loads, equipment aging, and other states; periodically and adaptively generating scheduling target weights to achieve optimal immediate balance of multi-objective benefits; employing evolutionary game theory and swarm intelligence optimization to obtain the optimal resource scheduling scheme across equipment under multi-objective constraints; and continuously feeding back the evolutionary model through real-time production data to form a scheduling... implement calibration A closed-loop mechanism for rescheduling.

[0004] However, while related technologies consider the health and load of the equipment, they mostly refer to statistical data or overall indicators, failing to directly detect the specific resistance generated instantaneously during processing. When the raw material suddenly hardens or the cutting tool suffers minor damage, the machine tool's actual processing speed slows down. However, because the scheduling system cannot detect this change in real time, it will still assign tasks to that machine tool, leading to hidden material accumulation. Secondly, since the conveyor belt is in continuous motion, the difficulty for the robotic arm to grasp materials varies depending on the distance and speed of the materials. Forcing operation under extreme conditions can easily lead to failure to grasp materials or even collision accidents. Summary of the Invention

[0005] To address the technical problems in the prior art, such as the inability to perceive the instantaneous physical blockage during machine tool cutting, leading to a disconnect between task allocation and actual capabilities, and the neglect of the cost of robotic arm gripping in high-speed assembly line environments, resulting in high equipment operation risks, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a multi-machine collaborative scheduling method for a flange production line based on digital twins, comprising: acquiring the real-time drive current and vibration acceleration amplitude of the machine tool spindle; constructing a slack coefficient based on the difference between the real-time drive current and the no-load reference current, and the difference between the vibration acceleration amplitude and the reference vibration threshold; acquiring the number of flanges to be processed in the conveyor belt buffer zone; constructing a buffer saturation based on the ratio of the number of flanges to be processed to the preset maximum capacity; and acquiring the machine tool's blocking tendency index in combination with the slack coefficient; acquiring the real-time image stream of the conveyor belt area and the real-time running speed of the conveyor belt; determining the real-time position coordinates of the flanges based on the real-time image stream; acquiring the grasping potential energy based on the distance deviation between the real-time position coordinates of the flanges and the robot arm base, and the real-time running speed of the conveyor belt and the maximum linear speed of the robot arm end effector; acquiring the average blocking tendency index of the downstream machine tools; obtaining a comprehensive shunt impedance based on the current machine tool's blocking tendency index, the average blocking tendency index of the downstream machine tools, and the grasping potential energy; and performing production line scheduling based on the comprehensive shunt impedance.

[0007] This invention directly reflects the physical load state and stability during the cutting process by collecting current and vibration data during machine tool operation. This allows for real-time detection of decreased machining capacity due to workpiece material fluctuations or tool wear, without relying on external manual monitoring. The invention maps the spindle state to a resistance coefficient and combines it with the saturation status of the conveyor belt buffer. This accurately assesses the machine tool's actual capacity to accept new tasks, avoiding overload caused by assigning tasks solely based on the buffer's available space. Furthermore, by combining the robotic arm's grasping potential energy with the conveyor belt's speed, the invention assesses the cost and risk of performing grasping actions. By comparing the congestion levels of the current machine tool and downstream machines, a comprehensive shunt impedance is generated. This allows the invention to proactively lower the interception threshold to alleviate production line pressure when the current machine tool has a certain load but downstream congestion is more severe. Simultaneously, it forcibly abandons grasping when the robotic arm is in a dangerous position. While ensuring the safe operation of individual machines, this achieves adaptive allocation of production line tasks, reducing production line downtime caused by localized material blockages and improving the overall operating efficiency and stability of automated production lines.

[0008] Preferably, the hysteresis coefficient satisfies the following relationship: In the formula, For the first The damping coefficient of the machine tool. For the first Effective overload current of the machine tool spindle. For the first The machine tool's preset no-load reference current, For the first The amplitude of vibration acceleration of the machine tool spindle. For the first The preset reference vibration threshold of the machine tool, This is the vibration sensitivity coefficient.

[0009] This invention achieves the ability to anticipate and reflect the instability of the cutting process in advance when an abnormal increase in spindle vibration is detected, even if the current load has not yet increased significantly. This reduces the probability that the scheduling system will assign tasks to the production node with potential failure risk, thus protecting the machine tool spindle and avoiding the production of defective products.

[0010] Preferably, the blocking tendency index satisfies the following relationship: In the formula, For the first Blocking tendency index of Taiwanese machine tools For the first The damping coefficient of the machine tool. The quantity of flanges to be processed. This is the maximum capacity.

[0011] This invention constructs a blockage tendency index based on the ratio of the number of flanges to be processed to the maximum capacity. When the material accumulation in the buffer is close to the limit, the blockage tendency index increases rapidly, thereby amplifying the signal strength of the high-risk blockage state. This allows a sufficiently strong impedance signal to be generated before the buffer overflows to suppress new grabbing behavior, prevent the conveyor belt from stacking or jamming, and maintain the tracking of the slack coefficient under low inventory conditions, accurately reflecting the machine tool's true processing capacity.

[0012] Preferably, the grasping potential energy satisfies the following relationship: In the formula, For the first A robotic arm uses its grasping potential energy to target a specific flange. For the flange center and the first The absolute value of the distance deviation of the center of the robotic arm base in the direction of the conveyor belt movement. For the first The maximum preset working radius of each robotic arm For the real-time operating speed of the conveyor belt, This represents the maximum linear speed of the robotic arm's end effector.

[0013] This invention assesses the cost of the robotic arm deviating from its optimal working area by calculating the distance deviation between the flange center and the robotic arm base center in the direction of movement. As the flange gradually moves away from the center point, the grasping potential energy increases, reflecting the increased risk of posture singularity when the robotic arm approaches the boundary of the workspace. This invention introduces the ratio of the real-time running speed of the conveyor belt to the maximum tracking speed of the robotic arm, further amplifying the potential energy value when the conveyor belt is running at high speed, thereby suppressing the system from performing high-risk grasping and preventing grasping failures or equipment collisions caused by the lag in the dynamic response of the robotic arm.

[0014] Preferably, the combined shunt impedance satisfies the following relationship: In the formula, For the first A robotic arm is used to target the overall shunt impedance of the current flange. For the first Each robotic arm corresponds to a machine tool's blocking tendency index. For the position located at the The average of the blocking tendency index of all machine tools downstream of a robotic arm. For the first The grasping potential energy of a robotic arm.

[0015] This invention utilizes the relative relationship between the congestion tendency index of the current machine tool and downstream machine tools, along with the grasping potential energy, to construct a comprehensive shunt impedance. When the load on the current machine tool increases, the comprehensive shunt impedance is increased to favor the release decision. Conversely, when downstream machine tools are generally congested, the comprehensive shunt impedance is attenuated to lower the grasping threshold, prompting the current machine tool to actively overload within a safe range to intercept materials and share the downstream pressure. This allows each node on the production line to adaptively adjust its behavior pattern according to the overall congestion situation, preventing excessive accumulation of materials at the end and achieving balanced and coordinated control of the production line load.

[0016] Preferably, before obtaining the real-time drive current and vibration acceleration amplitude of the machine tool spindle, the method further includes: during the self-test phase before the machine tool starts, controlling the spindle to rotate at the rated speed under no-load conditions, calculating the average value of the drive current within a preset time period as the no-load reference current, and calculating the average value of the vibration acceleration amplitude as the reference vibration threshold.

[0017] Preferably, determining the real-time position coordinates of the flange based on the real-time image stream includes: extracting the contour edge of the flange to be processed in the real-time image stream using the Canny edge detection operator, determining the pixel coordinates of the flange in the image coordinate system based on the center of the contour; calling the pre-calibrated camera intrinsic and extrinsic parameter matrices, and using a perspective transformation algorithm to map the pixel coordinates to the real-time position coordinates in the physical coordinate system of the conveyor belt.

[0018] Preferably, obtaining the number of flanges to be processed in the conveyor belt buffer includes: obtaining the number of flanges to be processed in the current buffer by using an infrared counting sensor installed at the machine tool feed inlet.

[0019] Preferably, the production line scheduling based on the comprehensive shunt impedance includes: in response to the comprehensive shunt impedance being less than a preset interception threshold, generating a grabbing command to control the robotic arm to grab the flange from the conveyor belt to the machine tool for processing; otherwise, generating a release command to allow the flange to flow downstream to the workstation with the conveyor belt.

[0020] Secondly, the present invention provides a multi-machine collaborative scheduling system for a flange production line based on digital twins, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned multi-machine collaborative scheduling method for a flange production line based on digital twins is implemented.

[0021] By adopting the above technical solution, a computer program is generated for the above-mentioned multi-machine collaborative scheduling method for flange production lines based on digital twins, and stored in the memory so that it can be loaded and executed by the processor. In this way, terminal equipment can be made according to the memory and the processor for convenient use.

[0022] The beneficial effects of this invention are as follows: By collecting data on the drive current and vibration acceleration of the machine tool spindle during machining, this invention reflects the changes in physical load caused by uneven material or tool wear during cutting. This allows for the assessment of the actual machining status of the machine tool without interrupting production, identifying nodes that appear to have available space but have reduced actual throughput capacity, thus preventing tasks from being assigned to equipment with potential hazards. This invention utilizes visual information to locate flange positions and calculates the grasping potential energy of the robotic arm based on the conveyor belt speed, evaluating the operational difficulty of the robotic arm under different spatial positions and dynamic speeds. This invention adjusts the robotic arm's grasping or releasing strategy by comparing the congestion levels of the current machine tool and downstream machine tools. When downstream congestion occurs, the upstream machine tool can proactively undertake more processing tasks, alleviating downstream pressure. When the robotic arm needs to perform large-span tracking or is at its movement limit, the system can select to release materials, avoiding high-risk operations by the robotic arm. This enables the production line to automatically balance the load based on the actual health and operational capacity of each node, reducing downtime and maintenance time caused by overload or collisions, and improving the collaborative efficiency and equipment safety of the entire production line in continuous operation environments. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a multi-machine collaborative scheduling method for a flange production line based on digital twins according to the present invention; Figure 2 This is a schematic diagram illustrating the integrated shunt impedance and the capture and release command in this invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] This invention discloses a multi-machine collaborative scheduling method for a flange production line based on digital twins, referring to... Figure 1 This includes steps S1-S4: S1. Construct the hysteresis coefficient based on current and vibration amplitude.

[0027] It should be noted that during flange cutting, the hardness fluctuation of the flange material and the wear condition of the cutting tool are important variables affecting the actual machining time. When uneven material hardness leads to increased cutting force, the spindle drive current will increase significantly; when the cutting tool experiences chipping or severe wear, the high-frequency vibration accompanying the cutting process will intensify. Both of these phenomena can cause the machine tool to fail to operate at the preset feed rate, resulting in machining delays. Therefore, this invention constructs a hysteresis coefficient based on the real-time drive current of the machine tool spindle and the amplitude of vibration acceleration.

[0028] Specifically, real-time drive current data and vibration acceleration amplitude data of the spindle are acquired by a current sensor deployed on the machine tool spindle drive and a vibration acceleration sensor mounted on the spindle housing, respectively. Pre-calibrated no-load reference current and reference vibration threshold are retrieved. The difference between the real-time drive current data and the no-load reference current is calculated. If the difference is less than or equal to zero, the effective overload current of the spindle is set to zero; if the difference is greater than zero, the difference is used as the effective overload current of the spindle. The deviation of the vibration acceleration amplitude data from the reference vibration threshold is calculated to construct a cutting stability correction factor. The machine tool's hysteresis coefficient is constructed based on the effective overload current, the no-load reference current, and the cutting stability correction factor.

[0029] For example, during the self-test phase before the machine tool starts, the spindle is controlled to rotate at the rated speed under no-load conditions, and the average value of the drive current within 10 minutes is used as the no-load reference current, and the average value of the vibration acceleration amplitude is used as the reference vibration threshold.

[0030] Specifically, the hindrance coefficient satisfies the following relationship: ; In the formula, For the first The damping coefficient of the machine tool. For the first Effective overload current of the machine tool spindle. For the first The machine tool's preset no-load reference current, For the first The amplitude of vibration acceleration of the machine tool spindle. For the first The preset reference vibration threshold of the machine tool, The vibration sensitivity coefficient, The empirical value range is [0.3, 0.6]. In this embodiment... The value is 0.5. The implementer can adjust this value according to the rigidity of the machine tool. For example, when using a high-rigidity heavy-duty machine tool, the value can be appropriately reduced. This reduces sensitivity to minute vibrations.

[0031] in, This represents the percentage increase in pure cutting resistance relative to the no-load state. A larger value indicates that the flange material being machined is too hard or the depth of cut is too large, leading to a significant increase in cutting resistance faced by the spindle, thus increasing the resistance coefficient and serving as a warning. Conversely, when the real-time current does not exceed the reference current, the effective overload current is... Zero, resulting in The value is zero, ensuring the non-negative boundary of the hindrance coefficient. As a cutting stability correction factor, when the vibration acceleration amplitude As the value increases, the cutting stability correction factor also increases, thus affecting... The vibration is amplified so that, under the same current load, the condition with greater vibration is judged as having a higher degree of stagnation, thus alerting the dispatching system to reduce the number of orders dispatched.

[0032] S2. Obtain the blocking tendency index based on the number of flanges and the resistance coefficient.

[0033] It should be noted that the risk of machine tool congestion is determined by both the conveyor belt's storage space and processing capacity. The saturation of the conveyor belt's storage space reflects the current material accumulation, while the resistance coefficient predicts the machine tool's material throughput efficiency per unit time. When the resistance coefficient increases due to operating conditions, it means the machine tool cannot quickly free up buffer space, causing the current inventory to stagnate. Even if the buffer's physical space is not full, a low throughput rate will still obstruct subsequent material flow. Therefore, this invention obtains a congestion tendency index based on the number of flanges to be processed in the conveyor belt buffer and the preset maximum capacity, combined with the resistance coefficient.

[0034] Specifically, the number of flanges to be processed in the current buffer is obtained by an infrared counting sensor installed at the machine tool's feed inlet, and the maximum capacity preset in the digital twin model is also obtained. The ratio of the number of flanges to be processed to the maximum capacity is calculated to construct the buffer saturation, and the machine tool's blocking tendency index is constructed by combining it with the hysteresis coefficient.

[0035] Specifically, the blocking tendency index satisfies the following relationship: ; In the formula, For the first Blocking tendency index of Taiwanese machine tools For the first The damping coefficient of the machine tool. The quantity of flanges to be processed. This is the maximum capacity.

[0036] in, This is an inventory pressure amplification factor. When the number of flanges to be processed... Approaching maximum capacity When this occurs, it indicates that the buffer space of the conveyor belt is about to be exhausted, thus amplifying the congestion tendency index and marking a high-risk congestion state; The smaller the value, the more likely the machine tool buffer is to be in a low-load state, and the smaller the contribution of inventory pressure to the risk of blockage. At this time, the blockage tendency index is mainly dominated by the stagnation coefficient, reflecting the actual throughput capacity of the machine tool in the current state.

[0037] S3. Grab potential energy based on conveyor belt image and speed.

[0038] It should be noted that in assembly line operations, the flange moves continuously with the conveyor belt. For a fixed robotic arm, there exists an optimal gripping point where the arm's lever arm is shortest, rigidity is best, and large-scale lateral following movements are unnecessary. When the flange is at this point, the gripping cost is lowest. As the flange gradually deviates from this center point, the robotic arm needs to extend to its limit or perform significant dynamic compensation, increasing the gripping difficulty and physical risk. Therefore, this invention determines the real-time position coordinates of the flange and acquires the gripping potential energy based on the real-time image stream of the conveyor belt area and the real-time operating speed of the conveyor belt.

[0039] Specifically, a real-time image stream of the conveyor belt area is acquired using an industrial vision camera mounted directly above the conveyor belt. The Canny edge detection operator is used to extract the contour edges of the flange to be processed from the image, and the pixel coordinates of the flange in the image coordinate system are determined based on the geometric center of the contour. Pre-calibrated camera intrinsic and extrinsic parameter matrices are retrieved, and a perspective transformation algorithm is used to map the pixel coordinates of the flange to the real-time position coordinates in the physical coordinate system of the conveyor belt. Simultaneously, the real-time operating speed of the conveyor belt is obtained by reading the encoder feedback value of the conveyor belt servo motor. Based on the real-time position coordinates of the flange and the preset installation position of the robotic arm base in the physical coordinate system of the conveyor belt, the real-time distance deviation between the center of the flange and the center of the robotic arm base is calculated.

[0040] Furthermore, the maximum working radius parameter preset by the robotic arm and the maximum linear speed parameter of the end effector are retrieved. Based on the real-time distance deviation between the flange center and the robotic arm base center, a spatial position potential energy term reflecting the risk of robotic arm posture singularity is constructed. Based on the real-time running speed of the conveyor belt and the maximum linear speed of the robotic arm end effector, a velocity coupling potential energy term reflecting the difficulty of dynamic tracking is constructed. The spatial position potential energy term and the velocity coupling potential energy term are combined to obtain the robotic arm's grasping potential energy for the current flange.

[0041] Specifically, we capture the relationship that the potential energy satisfies: ; In the formula, For the first A robotic arm uses its grasping potential energy to target a specific flange. For the flange center and the first The absolute value of the distance deviation of the center of the robotic arm base in the direction of the conveyor belt movement. For the first The maximum preset working radius of each robotic arm For the real-time operating speed of the conveyor belt, This represents the maximum linear speed of the robotic arm's end effector.

[0042] In the formula, This represents the degree of deviation of the flange position from the optimal gripping point. When the flange is exactly at the... When the robotic arm is directly in front of it, the distance deviation Approaching 0, A value close to 1 indicates that the robotic arm is in its optimal working position, requiring minimal attitude adjustment and exhibiting the lowest gripping resistance, without requiring significant posture adjustments. As the conveyor belt moves the flange away from the center position, the distance deviation... Gradually increase As it increases, it leads to As the deviation increases, the robotic arm approaches the boundary of its workspace, the risk of abnormal configuration increases sharply, the cost of the grasping operation increases, and the grasping potential energy increases.

[0043] This reflects the difficulty of dynamic tracking and the real-time operating speed of the conveyor belt. The faster the speed, the higher the end-effector acceleration required for the robotic arm to achieve speed synchronization at the moment of grasping, which poses a greater load challenge to the robotic arm's drive system. Therefore, when When it increases, The greater the grasping potential energy, the more it inhibits the system from forcibly grasping at edge extreme positions in high-speed flowing conditions, preventing grasping failure or collision accidents caused by the lag in the dynamic response of the robotic arm.

[0044] S4. Combine blocking tendency and grabbing potential energy to generate diversion instructions.

[0045] It should be noted that if the upstream machine tool prioritizes material capture, it may reach saturation prematurely, while the downstream machine tool remains idle due to a lack of material input. Conversely, if the upstream machine tool completely abandons material capture due to its own load, the material may not be captured by any intermediate node during transmission, ultimately flowing to the end of the conveyor belt and causing process omissions. Therefore, this invention constructs a comprehensive shunt impedance to distribute downstream pressure while ensuring the safe operation of the machine tool, achieving a dynamic balance in the overall task allocation of the production line.

[0046] Specifically, the average congestion tendency index of all machine tools downstream of the current robotic arm is obtained through the production line communication network. The combined shunt impedance of the robotic arm to the flange is obtained using the congestion tendency index of the machine tool corresponding to the robotic arm, the average congestion tendency index of all machine tools downstream of the robotic arm, and the grasping potential energy.

[0047] Specifically, the combined shunt impedance satisfies the following relationship: ; In the formula, For the first A robotic arm is used to target the overall shunt impedance of the current flange. For the first Each robotic arm corresponds to a machine tool's blocking tendency index. For the position located at the The average of the blocking tendency index of all machine tools downstream of a robotic arm. For the first The grasping potential energy of a robotic arm.

[0048] in, This represents the pressure balance factor, when the machine tool's blockage tendency index... When it increases, An increase in the average congestion tendency index leads to a rise in the overall shunt impedance, causing the system to tend to make a release decision to avoid further overloading of the machine tool in an already congested state. When downstream machine tools are generally congested, this leads to an increase in the average congestion tendency index. When it increases, Reduce, for the overall shunt impedance This attenuation effect allows the upstream machine tool to forcibly lower the overall shunt impedance even if it already has a certain load, in case the downstream production line becomes more congested. This forces the upstream machine tool to share the pressure from the downstream and actively overload to intercept materials, preventing the entire production line from shutting down due to downstream deadlock. When the flange is about to overflow the grabbing range or the conveyor belt is running too fast, the grabbing potential energy... The overall shunt impedance will be amplified to prioritize the safety of the robotic arm's operation.

[0049] Furthermore, in response to the overall shunt impedance being less than a preset interception threshold, a grabbing command is generated to control the robotic arm to grab the flange from the conveyor belt and transfer it to the machine tool for processing; otherwise, a release command is generated to allow the flange to flow downstream to the workstation with the conveyor belt.

[0050] For example, the interception threshold is 2.5, which can be set by the implementer based on the overall saturation desired by the production line.

[0051] For example, Figure 2 This diagram illustrates the integrated shunt impedance and grab release command in this invention. As can be seen from the diagram, the integrated shunt impedance fluctuates dynamically with the production process. When the integrated shunt impedance value is less than the interception threshold, it indicates that the current machine tool's load pressure and grabbing potential energy are at a low level, or that downstream machine tools are congested and need to share the pressure. The system then generates a grab command to control the robotic arm to grab the material. Conversely, when the integrated shunt impedance value exceeds the interception threshold, it indicates that the machine tool is approaching saturation or the grabbing cost is too high. The system then generates a release command, allowing the material to flow downstream, thereby achieving adaptive allocation of production line tasks and load balancing.

[0052] This invention also discloses a multi-machine collaborative scheduling system for a flange production line based on digital twins, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, a multi-machine collaborative scheduling method for a flange production line based on digital twins according to this invention is implemented.

[0053] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A multi-machine collaborative scheduling method for a flange production line based on digital twins, characterized in that, include: The real-time drive current and vibration acceleration amplitude of the machine tool spindle are obtained. The hysteresis coefficient is constructed based on the difference between the real-time drive current and the no-load reference current, and the difference between the vibration acceleration amplitude and the reference vibration threshold. The number of flanges to be processed in the conveyor belt buffer is obtained. The buffer saturation is constructed based on the ratio of the number of flanges to be processed to the preset maximum capacity. The machine tool blocking tendency index is obtained in combination with the hysteresis coefficient. The system acquires real-time image streams of the conveyor belt area and the real-time running speed of the conveyor belt. Based on the real-time image streams, it determines the real-time position coordinates of the flange. It obtains the grasping potential energy based on the distance deviation between the real-time position coordinates of the flange and the robotic arm base, as well as the real-time running speed of the conveyor belt and the maximum linear speed of the robotic arm end effector. It also acquires the average blocking tendency index of the downstream machine tool and obtains the comprehensive shunt impedance based on the current machine tool's blocking tendency index, the downstream machine tool's average blocking tendency index, and the grasping potential energy. Production line scheduling is based on the overall shunt impedance.

2. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 1, characterized in that, The hysteresis coefficient satisfies the following relationship: ； In the formula, For the first The damping coefficient of the machine tool. For the first Effective overload current of the machine tool spindle. For the first The machine tool's preset no-load reference current, For the first The amplitude of vibration acceleration of the machine tool spindle. For the first The preset reference vibration threshold of the machine tool, This is the vibration sensitivity coefficient.

3. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 2, characterized in that, The blocking tendency index satisfies the following relationship: ； In the formula, For the first Blocking tendency index of Taiwanese machine tools For the first The damping coefficient of the machine tool. The quantity of flanges to be processed. This is the maximum capacity.

4. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 1, characterized in that, The grasping potential energy satisfies the following relationship: ； In the formula, For the first A robotic arm uses its grasping potential energy to target a specific flange. For the flange center and the first The absolute value of the distance deviation of the center of the robotic arm base in the direction of the conveyor belt movement. For the first The maximum preset working radius of each robotic arm For the real-time operating speed of the conveyor belt, This represents the maximum linear speed of the robotic arm's end effector.

5. A multi-machine collaborative scheduling method for a flange production line based on digital twins according to claim 3, characterized in that, The combined shunt impedance satisfies the following relationship: ； In the formula, For the first A robotic arm is used to target the overall shunt impedance of the current flange. For the first Each robotic arm corresponds to a machine tool's blocking tendency index. For the position located at the The average of the blocking tendency index of all machine tools downstream of a robotic arm. For the first The grasping potential energy of a robotic arm.

6. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 1, characterized in that, Before acquiring the real-time drive current and vibration acceleration amplitude of the machine tool spindle, the method further includes: during the self-test phase before the machine tool starts, controlling the spindle to rotate at the rated speed under no-load conditions, calculating the average value of the drive current within a preset time period as the no-load reference current, and calculating the average value of the vibration acceleration amplitude as the reference vibration threshold.

7. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 1, characterized in that, The method of determining the real-time position coordinates of the flange based on the real-time image stream includes: extracting the contour edge of the flange to be processed in the real-time image stream using the Canny edge detection operator, determining the pixel coordinates of the flange in the image coordinate system based on the center of the contour; calling the pre-calibrated camera intrinsic and extrinsic parameter matrices, and using the perspective transformation algorithm to map the pixel coordinates to the real-time position coordinates in the physical coordinate system of the conveyor belt.

8. The method for multi-machine collaborative scheduling of a flange production line based on digital twins according to claim 1, characterized in that, The method of obtaining the number of flanges to be processed in the conveyor belt buffer includes: obtaining the number of flanges to be processed in the current buffer through an infrared counting sensor installed at the machine tool feed port.

9. A multi-machine collaborative scheduling method for a flange production line based on digital twins according to claim 1, characterized in that, The production line scheduling based on the comprehensive shunt impedance includes: in response to the comprehensive shunt impedance being less than a preset interception threshold, generating a grabbing command to control the robotic arm to grab the flange from the conveyor belt to the machine tool for processing; otherwise, generating a release command to allow the flange to flow downstream to the workstation with the conveyor belt.

10. A multi-machine collaborative scheduling system for a flange production line based on digital twins, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement a multi-machine collaborative scheduling method for a flange production line based on digital twins, as described in any one of claims 1-9.

Images