Automated control-based automobile disassembly line integration method and system

Abstract

The application discloses an automobile disassembly line integration method and system based on automatic control, and particularly relates to the field of control technology for disassembling scrapped automobiles, and is used for solving the problem that the cause-effect consistency between information decision and physical execution is destroyed when the remaining disassembly strategy is changed by online detection triggering; state data output by an online detection station is acquired to generate a current state index of a part, a dynamic threshold is determined according to a delayable time allowance of an associated disassembly action in a scheduled robot task queue to determine the remaining disassembly strategy change requirement, after it is determined that there is a change requirement, a robot task queue and a conveying device action sequence are extracted, a timing gap is identified according to a disassembly force transmission chain length and a beat deviation allowance, the changed disassembly strategy is decomposed into replacement task units and inserted into the timing gap, and a logistics track switch triggering time is synchronously updated, and finally, the task sequence is sent to a robot controller and a conveying device controller.

Description

Technical Field

[0001] This invention relates to the field of end-of-life vehicle dismantling control technology, and more specifically, to an integrated method and system for an automated vehicle dismantling line. Background Technology

[0002] Existing automated end-of-life vehicle dismantling lines generally employ integrated control schemes to uniformly schedule dismantling task allocation, conveyor operation, and robot work sequences. Typically, a static dismantling list is generated based on the vehicle model and initial surface damage assessment. Each workstation then performs dismantling sequentially according to the list, and material transport paths are arranged accordingly. During dismantling, some workstations are equipped with online detection systems to obtain the actual condition of internal parts after body panels or assemblies are disassembled, allowing for adjustments to subsequent processing flows.

[0003] However, when the status information obtained from online detection necessitates changes to the remaining dismantling strategy, the original integrated architecture struggles to respond effectively. The evaluation and decision-making updates of the dismantling plan are conducted at the information level, while vehicle transport, tooling preparation, and robot task queues have already been pre-locked according to previous decisions. This creates a time misalignment, preventing the dismantling depth adjustment and path change instructions made based on the new status from being adopted in a timely manner on the physical execution side. This disrupts the causal consistency between decision information and the physical process, resulting in the dismantling line being in a passive compromise state between high-value parts recovery rate and process efficiency for a long time. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides an integrated method and system for automobile dismantling lines based on automated control to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: An integrated method for automotive dismantling lines based on automated control includes the following steps: S1: Obtain the status data output by the online inspection station after assessing the status of the internal parts of the currently disassembled vehicle; S2: Generate the current status index of the part based on the status data, obtain the delay margin of the disassembly action associated with the part in the scheduled but not executed robot task queue, determine the dynamic threshold with the delay margin, and determine that there is a remaining disassembly strategy change requirement when the current status index of the part is lower than the dynamic threshold. S3: When it is determined that there is a remaining disassembly strategy change requirement, extract the robot task queue and the action sequence of the conveying device, obtain the disassembly force transmission chain length of the task node that conflicts with the changed disassembly strategy in the vehicle structure connection topology and the cycle deviation margin of the conveying section where the task node is located, and identify the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as the time gap. S4: Using the time gap as a constraint, decompose the changed dismantling strategy into alternative task units, insert the alternative task units into the time gap, and update the trigger time of the associated logistics turnout. S5: Send the task sequence after inserting the replacement task unit to the robot controller and the conveyor controller to control the robot and the conveyor to act at the updated logistics switch trigger time.

[0006] Furthermore, S1 includes: the online inspection station projects coded structured light onto the surface of the internal parts of the currently disassembled vehicle, collects the structured light pattern reflected by the surface of the internal parts of the currently disassembled vehicle, determines the spatial coordinates of each point on the surface of the internal parts of the currently disassembled vehicle based on the structured light pattern, and generates state data characterizing the three-dimensional morphology of the surface of the internal parts of the currently disassembled vehicle.

[0007] Furthermore, based on the structured light pattern, the spatial coordinates of each point on the surface of the currently disassembled vehicle's internal parts are determined, and state data characterizing the three-dimensional morphology of the surface of the currently disassembled vehicle's internal parts is generated. This includes: comparing the phase of the structured light pattern reflected by the surface of the currently disassembled vehicle's internal parts with the original projected structured light pattern to obtain the phase offset of the structured light pattern; calculating the depth distance from each point on the surface of the currently disassembled vehicle's internal parts to the structured light projection source based on the phase offset; and combining the depth distance of each point with the planar position of each point to form state data characterizing the three-dimensional morphology of the surface of the currently disassembled vehicle's internal parts.

[0008] Furthermore, S2 includes: extracting feature values ​​representing the degree of surface damage of the internal parts of the vehicle being disassembled as the current state index of the parts based on the state data; reading the planned start time of the disassembly action associated with the parts from the queue of scheduled but not yet executed robot tasks; calculating the time difference between the current time and the planned start time as the delayable time margin; determining a dynamic threshold based on the delayable time margin; the delayable time margin is negatively correlated with the dynamic threshold; and determining that there is a remaining need to change the disassembly strategy when the current state index of the parts is lower than the dynamic threshold.

[0009] Furthermore, S3 includes: in the vehicle structure connection topology, locating the fastener connections corresponding to the task nodes that conflict with the changed disassembly strategy, tracing back along the fastener connections to the boundary of the disassembled parts, and counting the number of fasteners included in the tracing path as the length of the disassembly force transmission chain; reading the real-time beat of the corresponding segment of the task node that conflicts with the changed disassembly strategy in the action sequence of the conveying device, and calculating the difference between the real-time beat and the rated beat as the beat deviation margin; from the task nodes that conflict with the changed disassembly strategy, selecting the task nodes whose disassembly force transmission chain length does not exceed the first reference and whose beat deviation margin is not lower than the second reference, and identifying the time interval between the planned start times of two adjacent task nodes as the timing gap.

[0010] Furthermore, the process involves tracing back along the fastener connections to the boundary of the disassembled part, and counting the number of fasteners included in the tracing path as the length of the disassembly force transmission chain. This includes: in the vehicle structure connection topology, starting from the fastener corresponding to the task node that conflicts with the changed disassembly strategy, tracing upstream along the adjacency relationship between fasteners until the fastener directly connected to the disassembled part is reached, and the number of fastener adjacency edges passed during the tracing process is taken as the length of the disassembly force transmission chain.

[0011] Further, S4 includes: decomposing the modified dismantling strategy into robot grasping actions, robot moving actions, and robot dismantling actions according to the dismantling process; forming alternative task units by combining the robot grasping actions, robot moving actions, and robot dismantling actions; comparing the estimated execution time of the alternative task unit with the duration of the timing gap, selecting a target timing gap from the timing gap whose duration is longer than the estimated execution time of the alternative task unit, inserting the alternative task unit into the target timing gap, and adjusting the planned start time of the affected task nodes in the robot task queue accordingly; recalculating the trigger time of the logistics switch according to the execution time period of the alternative task unit, and writing the recalculated trigger time of the logistics switch into the action sequence of the conveying device.

[0012] Further, S5 includes: extracting the planned start time of the adjusted task node from the robot task queue after inserting the replacement task unit; extracting the updated logistics switch trigger time from the conveyor action sequence after inserting the replacement task unit; combining the planned start time of the adjusted task node with the corresponding robot action instruction to form a robot control message; combining the updated logistics switch trigger time with the corresponding conveyor action instruction to form a conveyor control message; sending the robot control message to the robot controller; and sending the conveyor control message to the conveyor controller, so that the robot controller and the conveyor controller execute actions at the corresponding times.

[0013] Furthermore, robot control messages are sent to the robot controller, and conveyor control messages are sent to the conveyor controller, so that the robot controller and the conveyor controller perform actions at corresponding times, including: inserting a unified timestamp into the robot control message and inserting a unified timestamp into the conveyor control message. The unified timestamp is generated according to the planned start time of the first task node in the robot task queue after inserting the replacement task unit. The robot controller starts the operation synchronously according to the unified timestamp, and the conveyor controller starts the operation synchronously according to the unified timestamp.

[0014] On the other hand, the present invention provides an integrated system for an automotive dismantling line based on automated control, comprising the following modules: The data acquisition module is used to acquire the status data output by the online inspection station after it performs a status assessment on the internal parts of the currently disassembled vehicle. The demand determination module is used to generate the current status index of the part based on the status data, obtain the delay margin of the disassembly action associated with the part in the scheduled but not executed robot task queue, determine the dynamic threshold with the delay margin, and determine that there is a remaining demand for disassembly strategy change when the current status index of the part is lower than the dynamic threshold. The gap identification module is used to extract the robot task queue and the action sequence of the conveying device when it is determined that there is a remaining disassembly strategy change requirement. It obtains the length of the disassembly force transmission chain of the task node that conflicts with the changed disassembly strategy in the vehicle structure connection topology and the cycle deviation margin of the conveying section where the task node is located. It identifies the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as a time gap. The insertion update module is used to decompose the changed dismantling strategy into alternative task units with time slots as constraints, insert the alternative task units into the time slots, and update the trigger time of the associated logistics turnouts. The instruction issuing module is used to issue the task sequence after inserting the replacement task unit to the robot controller and the conveying device controller, so as to control the robot and the conveying device to act at the time of the updated logistics switch trigger.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By introducing a delay margin into the decision-making process for change requirements, the decision threshold on the information side is calibrated by the actual elasticity of the physical execution side. This ensures that changes to the remaining disassembly strategy are only triggered when there is still room for adjustment in the robot's task queue. This avoids the problem of change instructions being unable to be executed after they are generated from the decision-making source. When identifying timing gaps, both the length of the disassembly force transmission chain and the cycle deviation margin are considered. Time intervals with small mechanical involvement and sufficient delivery elasticity are selected as insertion windows. Replacement task units are limited to execution within these windows, thereby completing the dynamic reconstruction of the disassembly strategy without impacting the original task sequence or causing hidden damage to adjacent parts.

[0016] 2. The revised dismantling strategy is decomposed into alternative task units consisting of robot grasping actions, robot moving actions, and robot dismantling actions. After inserting timing gaps, the trigger time of the logistics switch is updated synchronously, and the updated task sequence is sent to the robot controller and the conveying device controller through a unified timestamp. This keeps the dismantling actions and material flow synchronized on the time axis, achieving causal consistency between information decision-making and physical execution in the operation of the dismantling line. This allows high-value parts discovered by online detection to be recovered in a timely manner, avoiding the loss of part value due to decision lag and maintaining the continuity of production rhythm. Attached Figure Description

[0017] Figure 1 This is a flowchart of an integrated method for an automotive dismantling line based on automated control, according to the present invention. Figure 2 This is a schematic diagram of the structure of an integrated system for an auto dismantling line based on automated control according to the present invention. Detailed Implementation

[0018] 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 The present invention provides an integrated method for automotive dismantling lines based on automated control, which includes the following steps: S1: Obtain the status data output by the online inspection station after assessing the status of the internal parts of the currently disassembled vehicle; S2: Generate the current status index of the part based on the status data, obtain the delay margin of the disassembly action associated with the part in the scheduled but not executed robot task queue, determine the dynamic threshold with the delay margin, and determine that there is a remaining disassembly strategy change requirement when the current status index of the part is lower than the dynamic threshold. S3: When it is determined that there is a remaining disassembly strategy change requirement, extract the robot task queue and the action sequence of the conveying device, obtain the disassembly force transmission chain length of the task node that conflicts with the changed disassembly strategy in the vehicle structure connection topology and the cycle deviation margin of the conveying section where the task node is located, and identify the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as the time gap. S4: Using the time gap as a constraint, decompose the changed dismantling strategy into alternative task units, insert the alternative task units into the time gap, and update the trigger time of the associated logistics turnout. S5: Send the task sequence after inserting the replacement task unit to the robot controller and the conveyor controller to control the robot and the conveyor to act at the updated logistics switch trigger time.

[0020] In the implementation of S1, after the end-of-life vehicle undergoes a pretreatment process to remove residual fluids such as fuel and coolant, it is conveyed into the dismantling line. At the upstream station of the dismantling line, some external covers have been removed, exposing the internal parts of the vehicle being dismantled. To obtain the actual condition of the internal parts of the vehicle being dismantled, an online inspection station is set up after the cover removal station and before the deep dismantling station. The online inspection station can be an inspection station equipped with industrial cameras and structured light projectors. When the pallet or lifting device carrying the vehicle being dismantled transports it to the online inspection station, the conveying device pauses to allow for condition assessment.

[0021] When the online inspection station assesses the condition of internal parts of a disassembled vehicle, a structured light projector first projects coded structured light onto the surface of the part. The coded structured light is a set of light stripe patterns with known spatial frequencies and phase distributions. Each stripe in the pattern is assigned a unique code before projection, and the patterns are projected sequentially according to a preset time sequence. For example, four sinusoidal light stripe patterns with phase differences of 90 degrees can be projected, each stripe in each pattern having a unique code. When the light stripe patterns reach the surface of the internal parts, the surface undulations, such as cracks, dents, wear grooves, or material detachment, modulate the projected light stripe patterns, causing them to bend and deform.

[0022] In the online inspection station, an industrial camera positioned at a fixed point relative to the structured light projector synchronously captures the structured light patterns reflected from the surfaces of the currently disassembled vehicle's interior parts. The light stripes in the structured light patterns reflected from the surfaces of the currently disassembled vehicle's interior parts undergo positional shifts and shape changes relative to the light stripes in the original projected structured light patterns. These positional shifts and shape changes carry information about the three-dimensional morphology of the surfaces of the currently disassembled vehicle's interior parts.

[0023] The process of determining the spatial coordinates of each point on the surface of the currently disassembled vehicle's internal parts based on structured light patterns involves comparing the phase of the structured light pattern reflected from the surface of the currently disassembled vehicle's internal parts with the original projected structured light pattern to obtain the phase offset of the structured light pattern. Specifically, the phase comparison involves extracting the actual phase value of each light stripe from the structured light pattern reflected from the surface of the currently disassembled vehicle's internal parts, extracting the initial phase value of the corresponding light stripe from the original projected structured light pattern, and subtracting the actual phase value from the initial phase value point by point to obtain the phase offset of the structured light pattern at each pixel location. For example, for four phase-shifted images, the actual phase value at each pixel location is calculated from the grayscale values ​​of the aforementioned pixel locations in the four images using a phase-shifting algorithm, while the initial phase value is directly given by the known encoding of the projected light stripe pattern.

[0024] The method for calculating the depth distance from each point on the surface of the currently disassembled vehicle interior part to the structured light projection source based on the phase offset is as follows: there is a triangulation relationship between the phase offset of the structured light pattern and the depth distance. The depth distance is calculated as: Z = (B × f) / (Δφ × d); where Z represents the depth distance from a point on the surface of the currently disassembled vehicle interior part to the structured light projection source, B represents the baseline distance between the structured light projector and the industrial camera, f represents the focal length of the industrial camera, Δφ represents the phase offset of the structured light pattern, and d represents the spatial frequency of the light stripe pattern. For each pixel position, the phase offset of the structured light pattern at the corresponding pixel position is substituted into the above triangulation relationship to calculate the depth distance from the aforementioned pixel position on the surface of the currently disassembled vehicle interior part to the structured light projection source. The structured light projection source refers to the center of the exit pupil of the structured light projector.

[0025] The process of combining the depth distance and planar position of each point to form state data representing the three-dimensional morphology of the surface of the currently disassembled vehicle interior parts is as follows: The imaging plane of the industrial camera itself provides planar position coordinates for each pixel location. The planar position coordinates are determined by the column and row numbers of the pixel and the intrinsic parameter matrix obtained from the industrial camera calibration. The depth distance corresponding to each pixel location is used as the depth coordinate perpendicular to the imaging plane. This is combined with the aforementioned planar position coordinates of the pixel location to form the spatial coordinates of each point on the surface of the currently disassembled vehicle interior parts. The set of all spatial coordinates constitutes the state data representing the three-dimensional morphology of the surface of the currently disassembled vehicle interior parts. This state data includes the macroscopic contour of the surface of the currently disassembled vehicle interior parts, as well as quantitative descriptions of microscopic damage morphology such as crack depth, pit volume, and protrusion height, providing a direct basis for subsequent steps to determine the current state of the parts.

[0026] In practice, the structured light projector and industrial camera at the online inspection station undergo system calibration after installation. The calibration results include the baseline distance between the structured light projector and the industrial camera, the focal length of the industrial camera, the spatial frequency of the light stripe pattern, and the intrinsic and extrinsic parameter matrices of the industrial camera. After calibration, these parameters remain unchanged and are directly used during each inspection. During periods when the conveyor is paused, the structured light projector completes the projection of all light stripe patterns, and the industrial camera completes all data acquisition and status data generation. The entire process is completed within the allowable time of a single conveyor cycle, and the status data is transmitted to subsequent steps for later use.

[0027] In the specific implementation of S2, the state data generated in S1, representing the three-dimensional morphology of the surface of the currently disassembled vehicle interior parts, is transmitted to the controller that executes subsequent steps. The process of generating the current state index of the parts based on the state data is as follows: a feature value is extracted from the state data as the current state index of the parts, and the aforementioned feature value is used to quantify the degree of damage to the surface of the currently disassembled vehicle interior parts.

[0028] The feature extraction method involves traversing the spatial coordinates of each point in the state data representing the three-dimensional morphology of the surface of the currently disassembled vehicle's internal parts. The point cloud data formed by these spatial coordinates is then fitted with a plane to obtain an ideal, undamaged plane on the surface of the currently disassembled vehicle's internal parts, which serves as a reference plane. The plane fitting uses the least squares method, with the criterion being the minimum sum of the squared perpendicular distances from the spatial coordinates of each point to the fitted plane. The normal vector and intercept of the fitted plane are then calculated, and the resulting plane becomes the reference plane. The perpendicular deviation distance from the spatial coordinates of each point to the reference plane is calculated point-to-plane using the formula for the distance between the spatial coordinates of each point and the reference plane. The root mean square (RMS) value of all perpendicular deviation distances is then calculated. This RMS value is obtained by squaring each perpendicular deviation distance, taking the arithmetic mean, and then taking the square root. The obtained RMS value is used as an indicator of the current state of the part, representing the degree of surface damage to the currently disassembled vehicle's internal parts. A larger root mean square (RMS) value indicates a deeper crack, a deeper pit, or a higher protrusion on the surface of the disassembled vehicle's internal parts, signifying more severe damage. For example, on an undamaged, smooth surface, the vertical deviation distance between points is close to zero, resulting in a near-zero RMS value. However, on a surface with deep cracks, the vertical deviation distance at the crack points increases significantly, leading to a larger RMS value. The calculation of the RMS value does not rely on additional sensors; it can be completed directly using the spatial coordinates from the state data, ensuring tight data integration and feasibility between steps.

[0029] The process of obtaining the delayable time margin for disassembly actions associated with parts in the scheduled but unexecuted robot task queue is as follows: The scheduled but unexecuted robot task queue is generated and maintained by the integrated control system when a disassembly task is issued. The robot task queue contains several task nodes, each recording the disassembly action type, the disassembly action object, and the planned start time of the disassembly action. The planned start time of the disassembly action refers to the execution start time allocated by the integrated control system for each disassembly action when executing the current disassembly task, accumulated in seconds from the moment the disassembly line starts running. The task node corresponding to the disassembly action associated with the currently disassembled internal vehicle part is retrieved from the robot task queue, and the planned start time of the disassembly action recorded in the aforementioned task node is read. The current time is provided by the real-time clock of the integrated control system, which records the number of seconds elapsed since the disassembly line started running. The time difference between the current time and the planned start time of the disassembly action is calculated as the delayable time margin, with the unit of the delayable time margin being seconds. For example, if the current time is the 1000th second of the dismantling line operation, and the planned start time for the dismantling action associated with the currently dismantling internal vehicle parts is the 1100th second, then the delay margin is 100 seconds. The delay margin reflects the tolerance of the physical execution side for decision changes. The larger the delay margin, the more time there is before the actual execution of the dismantling action, and the smaller the disturbance that decision changes will cause to the already scheduled task queue. Conversely, the smaller the delay margin, the more likely the dismantling action is to be executed, and the greater the cost of decision changes.

[0030] The method for determining the dynamic threshold based on the delayable time margin follows a negative correlation between the delayable time margin and the dynamic threshold. The specific process for setting the dynamic threshold is as follows: First, a baseline threshold is preset in the integrated control system. The baseline threshold is obtained by calibrating as follows: Statistical values ​​of the current status indicators of parts are collected during the historical operation of the dismantling line under different delayable time margin conditions when the remaining dismantling strategy change requirement is determined and the dismantling strategy adjustment is successfully implemented. The maximum value among the statistical values ​​of the current status indicators of parts that makes the change success rate higher than the preset minimum success rate is selected as the baseline threshold, and the baseline threshold is written into the integrated control system. Simultaneously, a nominal delayable time margin is preset. The nominal delayable time margin refers to the average time interval between two adjacent task nodes under standard operating cycle time of the dismantling line, determined by the rated speed of the conveyor and the station spacing. The nominal delayable time margin is also written into the integrated control system. The dynamic threshold is calculated as: D = D0 × (T0 / T); where D represents the dynamic threshold, D0 represents the baseline threshold, T0 represents the nominal delayable time margin, and T represents the delayable time margin. As can be seen from the above calculation formula, when the delay margin T increases, the dynamic threshold D decreases, and when the delay margin T decreases, the dynamic threshold D increases. To address the boundary case where the delay margin T approaches zero, a minimum delay margin is set before calculation, for example, a minimum delay margin of 1 second. When the delay margin T is less than the minimum delay margin, the delay margin T is set to the minimum delay margin to avoid overflow in the dynamic threshold D calculation.

[0031] The logic for comparing the current state indicators of a part with a dynamic threshold is as follows: When the current state indicator is lower than the dynamic threshold, it indicates that the damage to the surface of the part inside the vehicle being disassembled is so severe that it is not worthwhile to continue disassembling it according to the original disassembly strategy. Continuing disassembly would waste robot and conveyor resources, and the value of the recovered part would be lower than the disassembly cost. In this case, it is determined that there is a need to change the remaining disassembly strategy, and the integrated control system triggers subsequent steps to reconstruct the remaining disassembly strategy. When the current state indicator is not lower than the dynamic threshold, it indicates that the damage to the part inside the vehicle being disassembled is still within an acceptable range, and disassembly according to the original disassembly strategy still has recovery value. In this case, it is determined that there is no need to change the remaining disassembly strategy, and the queue of scheduled but unexecuted robot tasks remains unchanged. By introducing a delay margin into the dynamic adjustment of the judgment threshold, the judgment logic takes into account both the actual damage level of the part and the adjustment flexibility of the physical execution side, avoiding the situation in the prior art where decisions are effective but cannot be executed due to making decisions solely based on the part's state.

[0032] In the specific implementation of S3, after S2 determines that there is a remaining need to change the dismantling strategy, the integrated control system triggers the execution of S3. S3 first extracts the queue of scheduled but not yet executed robot tasks read in S2, and simultaneously extracts the action sequence of the conveyor device. The action sequence of the conveyor device is generated and maintained synchronously by the integrated control system with the queue of scheduled but not yet executed robot tasks when the dismantling task is issued. The action sequence of the conveyor device includes the start time, stop time, running speed, and material switch trigger time of the conveyor device in each conveying section. The aforementioned conveying section refers to the path interval traversed by the conveyor device to transport the current dismantling vehicle from the current workstation to the next workstation. Each conveying section corresponds to a set of action records in the action sequence of the conveyor device.

[0033] The revised disassembly strategy refers to the redefined disassembly plan based on the remaining disassembly strategy change requirements determined in S2. The revised strategy includes instructions to stop the original disassembly action for currently disassembled vehicle interior parts, perform a new disassembly action, or skip the original disassembly action. The process for determining task nodes that conflict with the revised disassembly strategy involves comparing each task node in the scheduled but unexecuted robot task queue with the new disassembly action or the skipped original disassembly action involved in the revised strategy. Task nodes in the scheduled but unexecuted robot task queue whose disassembly action object or disassembly action type contradicts the revised disassembly strategy are marked as conflicting task nodes. For example, if the revised disassembly strategy requires skipping the disassembly of a currently disassembled vehicle interior part, and a task node in the scheduled but unexecuted robot task queue already has that disassembly action object as the currently disassembled vehicle interior part, then that task node is marked as a conflicting task node with the revised disassembly strategy.

[0034] The vehicle structural connection topology refers to the fastener connection diagram constructed by the integrated control system based on the vehicle model repair manual and disassembly process documents before the vehicle is disassembled. The vehicle structural connection topology uses fasteners as nodes and the connections between fasteners as edges, recording the assembly hierarchy relationships between various parts of the vehicle being disassembled through fasteners such as bolts, clips, and welds. The vehicle structural connection topology is generated and stored in the integrated control system before the disassembly task begins. Each edge in the vehicle structural connection topology has a directional attribute, pointing from downstream fasteners to upstream fasteners. Downstream fasteners are those farther from the disassembled parts on the vehicle, while upstream fasteners are those closer to the disassembled parts. In the vehicle structural connection topology, the boundary of disassembled parts refers to the interface between parts that have been disassembled and those that have not yet been disassembled. The boundary of disassembled parts is updated in real time by the integrated control system according to the disassembly progress. When all fastener connections of a part have been released, that part is placed outside the boundary of disassembled parts.

[0035] The process of obtaining the disassembly force transmission chain length is as follows: In the vehicle structure connection topology, find the fastener corresponding to the task node that conflicts with the changed disassembly strategy, and use this fastener as the tracing starting point. The fastener corresponding to the task node that conflicts with the changed disassembly strategy refers to the fastener that needs to be removed by the disassembly action of the task node that conflicts with the changed disassembly strategy. Starting from the tracing starting point, trace upstream level by level along the adjacency relationship between fasteners. The adjacency relationship refers to the existence of an edge with a directional attribute between two fastener nodes in the vehicle structure connection topology. The tracing direction is from the downstream fastener along the directional attribute of the edge to the upstream fastener. The tracing terminates when the tracing reaches the fastener that is directly connected to the disassembled part. The fastener that is directly connected to the disassembled part means that the part fixed by the fastener has been marked outside the boundary of the disassembled part. The number of fastener adjacency edges traversed during the tracing process is taken as the disassembly force transmission chain length. The unit of the disassembly force transmission chain length is "number", which represents the number of fastener connection relationships traversed by the tracing path. The length of the disassembly force transmission chain reflects the degree of mechanical involvement of surrounding undisassembled parts when disassembling a fastener corresponding to a task node that conflicts with the changed disassembly strategy. The shorter the disassembly force transmission chain, the shorter the connection path between the fastener and the disassembled parts, and the less interference to surrounding undisassembled parts during disassembly. Conversely, the longer the disassembly force transmission chain, the deeper the fastener is in the assembly chain, and disassembly may cause mechanical involvement to multiple surrounding undisassembled parts.

[0036] The process of obtaining the cycle time deviation margin is as follows: In the conveyor action sequence, find the segment corresponding to the task node that conflicts with the modified dismantling strategy. The segment corresponding to the task node that conflicts with the modified dismantling strategy refers to the conveying segment in the conveyor action sequence when the conveyor transports the currently dismantled vehicle to the workstation of the task node that conflicts with the modified dismantling strategy. The real-time cycle time of the aforementioned segment corresponding to the task node that conflicts with the modified dismantling strategy is read from the conveyor action sequence. The real-time cycle time refers to the time interval between two adjacent workstations actually traversed by the conveyor in the aforementioned segment corresponding to the task node that conflicts with the modified dismantling strategy, measured by the encoder or displacement sensor of the conveyor and recorded in the integrated control system. The rated cycle time refers to the designed time interval between two adjacent workstations traversed by the conveyor under normal operating conditions, determined by the rated speed of the conveyor and the workstation spacing. The rated cycle time is pre-stored in the integrated control system. The difference between the real-time cycle time and the rated cycle time is calculated as the cycle time deviation margin, with the unit of the cycle time deviation margin being seconds. A positive cycle deviation margin indicates that the actual operating speed of the conveying device is slower than the design speed, and that there is unused time flexibility in the conveying section. A negative cycle deviation margin indicates that the actual operating speed of the conveying device has reached or exceeded the design speed, and that the time flexibility in the conveying section has been exhausted or does not exist.

[0037] After obtaining the disassembly force transmission chain length and cycle deviation margin for all task nodes that conflict with the changed disassembly strategy, a two-condition screening is performed. The first benchmark is the upper limit threshold of the disassembly force transmission chain length, which is set as follows: Disassembly line management personnel statistically analyze historical cases where the disassembly of a fastener caused damage to adjacent parts during disassembly. The statistical value of the disassembly force transmission chain length in these cases is extracted, and the disassembly force transmission chain length corresponding to a significant increase in the probability of causing damage to adjacent parts is used as the first benchmark. The typical value range of the first benchmark is 2 to 4 fastener adjacent edges. The second benchmark is the lower limit threshold of the cycle deviation margin, which is set as follows: Disassembly line management personnel statistically analyze the frequency of successful insertion of new tasks by the conveyor device under different cycle deviation margin conditions. The cycle deviation margin corresponding to a decrease in the frequency of successful insertion of new tasks to a preset minimum frequency is used as the second benchmark. The typical value range of the second benchmark is 0.5 seconds to 1.0 seconds. Both the first and second benchmarks are calibrated and written into the integrated control system during the dismantling line commissioning phase, and can be updated based on the actual operating results during the dismantling line operation.

[0038] From the task nodes that conflict with the modified disassembly strategy, those whose disassembly force transmission chain length does not exceed the first benchmark and whose cycle deviation margin is not lower than the second benchmark are selected. Task nodes that meet these two conditions are designated as candidate task nodes, while those that do not are excluded. Among the candidate task nodes, the time interval between the planned start times of two adjacent task nodes is identified as a timing gap. Two adjacent task nodes refer to two task nodes that are adjacent in the scheduled but not yet executed robot task queue and both belong to the candidate task nodes. The time interval between planned start times refers to the time span between the planned start time of the preceding task node and the planned start time of the following task node, starting at the planned start time of the preceding task node and ending at the planned start time of the following task node. A timing gap represents a time window within which a new disassembly action can be safely inserted under the current physical execution constraints. The timing gap is not affected by the hidden damage risk caused by the disassembly force transmission chain and has sufficient transport elasticity to accommodate the inserted new disassembly action.

[0039] By identifying timing gaps through dual-condition screening, unlike methods that only rely on time intervals to find insertion windows, this scheme considers both the mechanical influence range of the disassembly action and the dynamic operating margin of the conveying device. Only when neither of these two dimensions constitutes a constraint is the time interval between the corresponding two task nodes considered a usable timing gap. This avoids mistakenly using time intervals with hidden damage risks or insufficient conveying flexibility as insertion windows, thereby ensuring the safety of inserting replacement task units in subsequent steps.

[0040] In the specific implementation of S4, the timing gaps identified in S3 are transmitted to the controller executing S4. The modified disassembly strategy is generated after S2 determines that there is a remaining need to modify the disassembly strategy. The modified disassembly strategy includes instructions to stop the original disassembly action of the currently disassembled internal parts of the vehicle, to perform a new disassembly action, or to skip the original disassembly action. The process of decomposing the modified disassembly strategy into robot gripping action, robot moving action, and robot disassembly action according to the disassembly process is as follows: The disassembly process defines the sequence of actions required to complete the disassembly of a part. The robot gripping action refers to the path action of the robot end effector to grab the corresponding disassembly tool from the tool library. The robot moving action refers to the path action of the robot carrying the disassembly tool to move from the current pose to the disassembly operation pose. The robot disassembly action refers to the action of the robot performing any one of the disassembly operations such as loosening bolts, removing clips, or cutting welds in the disassembly operation pose. The decomposed robot grasping action, robot moving action, and robot disassembly action are combined into a substitute task unit. The substitute task unit contains the sequential constraints and connection relationships between the robot grasping action, robot moving action, and robot disassembly action. The sequential constraint means that the robot moving action can only be executed after the robot grasping action is completed, and the robot disassembly action can only be executed after the robot moving action is completed.

[0041] The estimated execution time of the alternative task unit is calculated by the integrated control system based on the preset execution time of each action. The preset execution time is generated in advance by the robot's offline programming software based on path length, joint speed, and disassembly operation type, and stored in the integrated control system. The execution time of the robot's grasping action is calculated based on the distance between the tool library and the current workstation, as well as the robot's joint rated speed. For example, if the distance between the tool library and the current workstation is 2 meters and the robot's joint rated speed is 1 meter per second, the execution time of the robot's grasping action is 2 seconds. The execution time of the robot's movement action is calculated based on the distance between the robot's current position and the disassembly operation pose, as well as the robot's joint rated speed. The execution time of the robot's disassembly action is determined based on the fastener type. When the fastener type is a bolt, the execution time of the robot's disassembly action is the bolt removal time; when the fastener type is a clip, the execution time of the robot's disassembly action is the clip extraction time; and when the fastener type is a weld point, the execution time of the robot's disassembly action is the weld point shearing time. The bolt removal time, clip extraction time, and weld point shearing time are all pre-calibrated process parameters stored in the integrated control system. The estimated execution time of the alternative task unit is obtained by summing the execution time of the robot's grasping action, the execution time of the robot's moving action, and the execution time of the robot's disassembly action. The estimated execution time of the alternative task unit is in seconds.

[0042] The process of comparing the estimated execution time of the alternative task unit with the duration of the time slot is as follows: All time slots identified in S3 are traversed. Each time slot has a duration, which is the difference between the start and end points of the corresponding time interval, in seconds. For each time slot, the estimated execution time of the alternative task unit is compared with the duration of the time slot. When selecting a target time slot whose duration is greater than the estimated execution time of the alternative task unit, if multiple time slots have durations greater than the estimated execution time of the alternative task unit, the time slot with the smallest difference between its duration and the estimated execution time of the alternative task unit is selected as the target time slot. This minimizes the time offset of affected task nodes in the scheduled but unexecuted robot task queue. If the duration of any time gap is greater than the estimated execution time of the alternative task unit, then the insertion of the alternative task unit is abandoned, and the queue of scheduled but unexecuted robot tasks remains unchanged, waiting to enter the next vehicle dismantling scheduling cycle for further processing, in order to avoid task node overlap and conflict in the robot task queue due to forced insertion.

[0043] The process of inserting a substitute task unit into the target timing gap is as follows: In the scheduled but not yet executed robot task queue, the starting point of the target timing gap is determined as the insertion time of the substitute task unit. The substitute task unit is written into the scheduled but not yet executed robot task queue as a task node. The robot grasping action, robot movement action, and robot disassembly action of the substitute task unit are set as three consecutive task nodes. The planned start time of the task node for the robot grasping action is set as the insertion time. The planned start times of the task nodes for the robot movement action and the robot disassembly action are sequentially postponed. The planned start times of the affected task nodes in the robot task queue are adjusted accordingly. Affected task nodes refer to all task nodes in the scheduled but not yet executed robot task queue located after the target timing gap. The planned start times of all affected task nodes are postponed to the estimated execution time of the substitute task unit to ensure that the order and interval of each task node in the robot task queue are not disrupted after the insertion of the substitute task unit.

[0044] The process of updating the trigger time of the associated logistics turnout based on the execution period of the alternative task unit and writing the recalculated trigger time of the logistics turnout into the action sequence of the conveying device is as follows: The execution period of the alternative task unit refers to the time interval from the planned start time of the task node of the robot grasping action of the alternative task unit to the planned start time of the task node of the robot disassembly action of the alternative task unit plus the execution time of the robot disassembly action of the alternative task unit. A logistics turnout refers to a steering device on the conveying device that guides the currently disassembled vehicle into different branches at a branch point. An associated logistics turnout refers to a logistics turnout related to the subsequent conveying path of the currently disassembled vehicle containing the internal parts involved in the alternative task unit. The method for recalculating the trigger time of the logistics switch is as follows: if the execution period of the substitute task unit includes the originally planned trigger time of the logistics switch, the originally planned trigger time of the logistics switch is postponed by the estimated execution time of the substitute task unit. If the execution period of the substitute task unit does not include the originally planned trigger time of the logistics switch, and the execution period of the substitute task unit is after the originally planned trigger time of the logistics switch, the originally planned trigger time of the logistics switch is advanced by the estimated execution time of the substitute task unit. This ensures that the dismantling vehicle can still flow to the subsequent workstation along the correct path after executing the substitute task unit. The recalculated trigger time of the logistics switch is overwritten into the corresponding logistics switch trigger time field in the conveyor action sequence. After overwriting, the conveyor action sequence and the updated scheduled but not executed robot task queue are synchronized in time, avoiding deviations between the physical position of the current dismantling vehicle on the conveyor and the information status in the robot task queue.

[0045] In the specific implementation of S5, after S4 completes the insertion of the replacement task unit and the update of the logistics switch trigger time, the queue of scheduled but not yet executed robot tasks includes three consecutive task nodes corresponding to the replacement task unit, and the action sequence of the conveying device includes the recalculated trigger time of the logistics switch. The integrated control system starts S5 and uniformly distributes all updated task sequences to the execution layer.

[0046] The process of extracting the adjusted planned start time of task nodes from the robot task queue after inserting the substitute task unit is as follows: The robot task queue after inserting the substitute task unit refers to the complete robot task queue obtained after writing the substitute task unit in S4 and shifting the planned start time of the affected task nodes backward. Adjusted task nodes refer to all task nodes in the robot task queue after inserting the substitute task unit whose planned start time has changed compared to the original planned start time read in S2, including the three consecutive task nodes corresponding to the substitute task unit and the affected task nodes whose estimated execution time of the substitute task unit has been postponed. For each adjusted task node, its planned start time and disassembly action type are extracted. The disassembly action type refers to any one of the following: robot grasping action, robot moving action, or robot disassembly action.

[0047] The process of combining the planned start time of the adjusted task node with the corresponding robot motion instructions to form a robot control message is as follows: Robot motion instructions are the translation of disassembly action types into motion command codes executable by the robot controller. These motion command codes can include any one of the following: target joint angle, target end-effector pose, or target torque. The integrated control system stores a mapping table between disassembly action types and motion command codes. This mapping table is generated by the robot's offline programming software and written into the integrated control system during the disassembly line debugging phase. The mapping table establishes a one-to-one correspondence between each disassembly action type and its corresponding motion command code. For example, a robot grasping action corresponds to the tool grasping motion command code in the mapping table; a robot movement action corresponds to the point-to-point motion command code; and a robot disassembly action corresponds to the bolt loosening or buckle removal motion command code. For each adjusted task node, the corresponding motion command code is retrieved from the mapping table between disassembly action type and motion command code based on the disassembly action type. The planned start time of the adjusted task node is then appended as a timestamp before the corresponding motion command code, forming a segment in the robot control message. All the segment messages of the adjusted task nodes are arranged in chronological order of their planned start times to form a complete robot control message.

[0048] The process of extracting the updated logistics turnout trigger times from the conveyor action sequence after inserting the replacement task unit is as follows: The conveyor action sequence after inserting the replacement task unit refers to the complete conveyor action sequence obtained after overwriting the recalculated logistics turnout trigger times in S4. The updated logistics turnout trigger times refer to the values ​​of the trigger time fields of each logistics turnout in the conveyor action sequence after the recalculated logistics turnout trigger times are overwritten. The updated logistics turnout trigger times are extracted item by item from the conveyor action sequence after inserting the replacement task unit. During extraction, the logistics turnout identifier and branch direction instruction associated with each updated logistics turnout trigger time are recorded simultaneously. The logistics turnout identifier is used to uniquely identify a specific logistics turnout on the conveyor, and the branch direction instruction is used to specify whether the aforementioned logistics turnout switches to the left branch or the right branch.

[0049] The process of combining the updated logistics turnout trigger time with the corresponding conveyor action command to form a conveyor control message is as follows: the conveyor action command translates the branch direction command into a switch signal or pulse signal executable by the conveyor controller. A high level indicates the logistics turnout is switching to the left branch, and a low level indicates it is switching to the right branch. The pulse signal uses pulse width to represent the duration of the logistics turnout switching. For each extracted logistics turnout, its updated trigger time is appended as a timestamp before the corresponding conveyor action command, forming a segment in the conveyor control message. All logistics turnout segment segments are arranged in chronological order of their updated trigger times to form a complete conveyor control message.

[0050] The process of sending robot control messages to the robot controller and conveyor control messages to the conveyor controller is as follows: The integrated control system establishes communication connections with the robot controller and conveyor controller via industrial Ethernet or fieldbus. The communication protocol uses TCP / IP-based socket communication or CAN bus-based frame communication. A unified timestamp is inserted into both the robot control message and the conveyor control message before sending them. The unified timestamp is generated based on the planned start time of the first task node in the robot task queue after the insertion of the replacement task unit. The first task node in the robot task queue after the insertion of the replacement task unit refers to the task node with the earliest planned start time in the robot task queue after the insertion of the replacement task unit. The planned start time of the first task node is used as the value of the unified timestamp, and the unit of the unified timestamp is seconds. For example, if the planned start time of the earliest task node in the robot task queue after the insertion of the replacement task unit is 1200 seconds, then the value of the unified timestamp is 1200 seconds.

[0051] The robot control message containing a unified timestamp is sent to the robot controller, and the conveyor control message containing a unified timestamp is sent to the conveyor controller. Upon receiving the robot control message, the robot controller first parses the unified timestamp from the message and uses it as the reference time for starting the operation. The robot controller aligns its own clock with the unified timestamp. When its own clock reaches the unified timestamp, the robot controller begins to execute the corresponding motion command codes sequentially according to the timestamp of each message segment in the robot control message. Similarly, upon receiving the conveyor control message, the conveyor controller first parses the unified timestamp from the message and uses it as the reference time for starting the operation. The conveyor controller aligns its own clock with the unified timestamp. When its own clock reaches the unified timestamp, the conveyor controller begins to output the corresponding switch signals or pulse signals sequentially according to the timestamp of each message segment in the conveyor control message.

[0052] By synchronizing the start of operations with a unified timestamp, the starting points of the robot controller and the conveyor controller are kept consistent in time. This avoids timing misalignment caused by communication delays or controller clock deviations, where the robot has already started executing a new task node while the conveyor has not yet completed the logistics switchover. This ensures the causal consistency of information and physical processes.

[0053] Example 2: Figure 2 A schematic diagram of an integrated system for an automated automotive dismantling line is provided according to the present invention. This system includes the following modules: The data acquisition module is used to acquire the status data output by the online inspection station after it performs a status assessment on the internal parts of the currently disassembled vehicle. The demand determination module is used to generate the current status index of the part based on the status data, obtain the delay margin of the disassembly action associated with the part in the scheduled but not executed robot task queue, determine the dynamic threshold with the delay margin, and determine that there is a remaining demand for disassembly strategy change when the current status index of the part is lower than the dynamic threshold. The gap identification module is used to extract the robot task queue and the action sequence of the conveying device when it is determined that there is a remaining disassembly strategy change requirement. It obtains the length of the disassembly force transmission chain of the task node that conflicts with the changed disassembly strategy in the vehicle structure connection topology and the cycle deviation margin of the conveying section where the task node is located. It identifies the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as a time gap. The insertion update module is used to decompose the changed dismantling strategy into alternative task units with time slots as constraints, insert the alternative task units into the time slots, and update the trigger time of the associated logistics turnouts. The instruction issuing module is used to issue the task sequence after inserting the replacement task unit to the robot controller and the conveying device controller, so as to control the robot and the conveying device to act at the time of the updated logistics switch trigger.

[0054] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0055] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0056] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and inventive constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

Claims

1. An integrated method for automotive dismantling lines based on automated control, characterized in that, Includes the following steps: S1: Obtain the status data output by the online inspection station after assessing the status of the internal parts of the currently disassembled vehicle; S2: Generate the current state index of the part based on the state data, obtain the delay margin of the disassembly actions associated with the part in the scheduled but not yet executed robot task queue, determine the dynamic threshold based on the delay margin, and determine that there is a remaining disassembly strategy change requirement when the current state index of the part is lower than the dynamic threshold, including: Based on the state data, feature values ​​representing the surface damage level of the internal parts of the vehicle being disassembled are extracted as the current state index of the parts. The planned start time of the disassembly action associated with the parts is read from the queue of scheduled but not executed robot tasks. The time difference between the current time and the planned start time is calculated as the delay margin. The dynamic threshold is determined based on the delay margin. The delay margin is negatively correlated with the dynamic threshold. When the current state index of the parts is lower than the dynamic threshold, it is determined that there is a remaining need to change the disassembly strategy. S3: When it is determined that there is a remaining need to change the disassembly strategy, extract the robot task queue and the action sequence of the conveying device, obtain the length of the disassembly force transmission chain of the task node that conflicts with the changed disassembly strategy in the vehicle structure connection topology, and the cycle deviation margin of the conveying section where the task node is located. Identify the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as time gaps, including: In the vehicle structure connection topology, the fastener connections corresponding to the task nodes that conflict with the changed disassembly strategy are located. The fastener connections are traced back step by step to the boundary of the disassembled parts, and the number of fasteners included in the backtracking path is counted as the length of the disassembly force transmission chain. The real-time cycle time of the corresponding section of the task node that conflicts with the changed disassembly strategy in the action sequence of the conveying device is read, and the difference between the real-time cycle time and the rated cycle time is calculated as the cycle time deviation margin. From the task nodes that conflict with the changed disassembly strategy, the task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle time deviation margin is not lower than the second reference are selected, and the time interval between the planned start time of two adjacent task nodes is identified as the timing gap. S4: Using time slots as constraints, decompose the modified dismantling strategy into alternative task units, insert the alternative task units into the time slots, and update the trigger times of the associated logistics switches, including: The revised dismantling strategy is decomposed into robot grasping actions, robot moving actions, and robot dismantling actions according to the dismantling process. These actions are then combined to form alternative task units. The estimated execution time of the alternative task units is compared with the duration of the time slots. A target time slot with a duration longer than the estimated execution time of the alternative task units is selected from the time slots. The alternative task units are then inserted into the target time slots, and the planned start times of the affected task nodes in the robot task queue are adjusted accordingly. Based on the execution time of the alternative task units, the trigger time of the logistics turnout is recalculated, and the recalculated trigger time of the logistics turnout is written into the action sequence of the conveying device. S5: Send the task sequence after inserting the replacement task unit to the robot controller and the conveyor controller to control the robot and the conveyor to act at the updated logistics switch trigger time.

2. The method for integrating an automated automotive dismantling line according to claim 1, characterized in that, S1 includes: The online inspection station projects coded structured light onto the surface of the internal parts of the vehicle being disassembled, collects the structured light pattern reflected from the surface of the internal parts, determines the spatial coordinates of each point on the surface of the internal parts based on the structured light pattern, and generates state data characterizing the three-dimensional morphology of the surface of the internal parts.

3. The method for integrating an automated automotive dismantling line according to claim 2, characterized in that, Based on the structured light pattern, the spatial coordinates of each point on the surface of the currently disassembled vehicle's internal parts are determined, and state data representing the three-dimensional morphology of the surface of the currently disassembled vehicle's internal parts is generated. This includes: comparing the phase of the structured light pattern reflected by the surface of the currently disassembled vehicle's internal parts with the original projected structured light pattern to obtain the phase offset of the structured light pattern; calculating the depth distance from each point on the surface of the currently disassembled vehicle's internal parts to the structured light projection source based on the phase offset; and combining the depth distance of each point with the planar position of each point to form state data representing the three-dimensional morphology of the surface of the currently disassembled vehicle's internal parts.

4. The method for integrating an automated automotive dismantling line according to claim 1, characterized in that, The process involves tracing back along the fastener connections to the boundary of the disassembled part, and counting the number of fasteners included in the tracing path as the length of the disassembly force transmission chain. This includes: in the vehicle structure connection topology, starting from the fastener corresponding to the task node that conflicts with the changed disassembly strategy, tracing upstream along the adjacency relationship between fasteners until the fastener directly connected to the disassembled part is reached, and the number of fastener adjacency edges passed during the tracing process is taken as the length of the disassembly force transmission chain.

5. The method for integrating an automated automotive dismantling line according to claim 1, characterized in that, S5 includes: extracting the planned start time of the adjusted task node from the robot task queue after inserting the replacement task unit; extracting the updated logistics switch trigger time from the conveyor action sequence after inserting the replacement task unit; combining the planned start time of the adjusted task node with the corresponding robot action instruction to form a robot control message; combining the updated logistics switch trigger time with the corresponding conveyor action instruction to form a conveyor control message; sending the robot control message to the robot controller; sending the conveyor control message to the conveyor controller; and causing the robot controller and the conveyor controller to execute actions at the corresponding times.

6. The method for integrating an automated automotive dismantling line according to claim 5, characterized in that, The robot control message is sent to the robot controller, and the conveyor control message is sent to the conveyor controller, so that the robot controller and the conveyor controller perform actions at corresponding times, including: inserting a unified timestamp into the robot control message and inserting a unified timestamp into the conveyor control message. The unified timestamp is generated according to the planned start time of the first task node in the robot task queue after the insertion of the replacement task unit. The robot controller starts the operation synchronously according to the unified timestamp, and the conveyor controller starts the operation synchronously according to the unified timestamp.

7. An integrated system for an automated vehicle dismantling line, used to implement the automated vehicle dismantling line integration method according to any one of claims 1-6, characterized in that, Includes the following modules: The data acquisition module is used to acquire the status data output by the online inspection station after it performs a status assessment on the internal parts of the currently disassembled vehicle. The demand determination module is used to generate the current state index of the part based on the state data, obtain the delay margin of the disassembly actions associated with the part in the scheduled but not executed robot task queue, determine the dynamic threshold based on the delay margin, and determine that there is a remaining demand for disassembly strategy change when the current state index of the part is lower than the dynamic threshold, including: Based on the state data, feature values ​​representing the surface damage level of the internal parts of the vehicle being disassembled are extracted as the current state index of the parts. The planned start time of the disassembly action associated with the parts is read from the queue of scheduled but not executed robot tasks. The time difference between the current time and the planned start time is calculated as the delay margin. The dynamic threshold is determined based on the delay margin. The delay margin is negatively correlated with the dynamic threshold. When the current state index of the parts is lower than the dynamic threshold, it is determined that there is a remaining need to change the disassembly strategy. The gap identification module is used to extract the robot task queue and the action sequence of the conveying device when it is determined that there is a remaining need to change the disassembly strategy. It obtains the length of the disassembly force transmission chain of the task node conflicting with the changed disassembly strategy in the vehicle structure connection topology, as well as the cycle deviation margin of the conveying section where the task node is located. It identifies the time interval between adjacent task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle deviation margin is not lower than the second reference as a time gap, including: In the vehicle structure connection topology, the fastener connections corresponding to the task nodes that conflict with the changed disassembly strategy are located. The fastener connections are traced back step by step to the boundary of the disassembled parts, and the number of fasteners included in the backtracking path is counted as the length of the disassembly force transmission chain. The real-time cycle time of the corresponding section of the task node that conflicts with the changed disassembly strategy in the action sequence of the conveying device is read, and the difference between the real-time cycle time and the rated cycle time is calculated as the cycle time deviation margin. From the task nodes that conflict with the changed disassembly strategy, the task nodes whose disassembly force transmission chain length does not exceed the first reference and whose cycle time deviation margin is not lower than the second reference are selected, and the time interval between the planned start time of two adjacent task nodes is identified as the timing gap. The insertion and update module is used to decompose the changed dismantling strategy into alternative task units based on timing gaps, insert the alternative task units into the timing gaps, and update the trigger times of the associated logistics switches, including: The revised dismantling strategy is decomposed into robot grasping actions, robot moving actions, and robot dismantling actions according to the dismantling process. These actions are then combined to form alternative task units. The estimated execution time of the alternative task units is compared with the duration of the time slots. A target time slot with a duration longer than the estimated execution time of the alternative task units is selected from the time slots. The alternative task units are then inserted into the target time slots, and the planned start times of the affected task nodes in the robot task queue are adjusted accordingly. Based on the execution time of the alternative task units, the trigger time of the logistics turnout is recalculated, and the recalculated trigger time of the logistics turnout is written into the action sequence of the conveying device. The instruction issuing module is used to issue the task sequence after inserting the replacement task unit to the robot controller and the conveying device controller, so as to control the robot and the conveying device to act at the time of the updated logistics switch trigger.

Images