Intelligent manufacturing workshop equipment collaborative management and control method and system

Abstract

The application relates to the technical field of device cooperative management and control, and discloses an intelligent manufacturing workshop device cooperative management and control method and system, which comprises the following steps: obtaining running state information and a state transition graph, constructing a cooperative process causal network and extracting a common confirmed causal frontier; taking the common confirmed causal frontier as a deduction starting point, combining the state transition graph to calculate a causal compatible state set and a cooperative recoverable state domain, and generating a state interaction tolerance time window; constructing a related device state envelope based on the state interaction tolerance time window, screening a candidate degradation action, obtaining a target degradation action and executing the target degradation action to generate a local event record; reconstructing a cross-device event causal partial order graph, generating a candidate consistent causal section, selecting a target recovery section, hierarchically eliminating bifurcation events after the target recovery section, and recovering the cooperative management and control task. The application realizes the safe alignment of each device to a business continuation node, and enables the smooth continuation of the workshop cooperative management and control task on the basis of reserving the completed workload.

Description

Technical Field

[0001] This invention relates to the field of equipment collaborative management and control technology, specifically to a method and system for collaborative management and control of equipment in an intelligent manufacturing workshop. Background Technology

[0002] In intelligent manufacturing workshops where machining, material handling, quality inspection, and assembly operations are seamlessly integrated, the completion of processing, material release, and resource deactivation of preceding equipment typically constitute the execution conditions for subsequent equipment to initiate material handling, inspection, or assembly actions. Existing equipment collaborative management methods mainly coordinate the action sequence and task progress of each piece of equipment through interlock signals between programmable controllers, equipment status messages, and task statuses issued by the manufacturing execution system. When status messages are missing, communication is delayed, or confirmation information is interrupted, a fixed communication timeout threshold is typically used to trigger equipment suspension, safety shutdown, or task reset. After the status interaction is restored, production tasks continue to be executed based on the state saved before the anomaly or the current feedback state of the equipment.

[0003] However, when an abnormal state interaction occurs, different devices may be in the initiation, execution, material holding, or process completion stages, making it difficult for existing control methods to adapt to the actual process of device operation. First, fixed timeout thresholds do not take into account the current stage of device operation, the remaining execution process, and the physical interlock position, which may lead to premature shutdown when the device still has safe waiting conditions, or delayed intervention when the device is about to enter the interlock area. Second, when the real-time status of associated devices cannot be obtained, selecting local control actions based on a single inferred status or a unified shutdown rule makes it difficult to ensure that the selected actions are applicable to other possible states of associated devices, resulting in out-of-order processes, resource occupation conflicts, or device movement interference. Finally, after the state interaction is restored, directly using the pre-abnormal state, the latest state of each device, or the unified reset state to continue the task may result in the repeated execution of completed processes or the cancellation of the preceding actions required to maintain clamping, bearing, and device positional relationships, thereby affecting the safe exit of the device and the connection of remaining tasks. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for collaborative management and control of equipment in an intelligent manufacturing workshop, comprising: Acquire the operating status information and state transition diagram of equipment in the intelligent manufacturing workshop, construct a collaborative process causal network based on the operating status information, and extract the common confirmed causal front. Taking the jointly confirmed causal frontier as the starting point of deduction, the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly are calculated in combination with the state transition diagram, and a state interaction tolerance window is generated when the causal compatible state set deviates from the cooperative recoverable state domain. Based on the state interaction tolerance window, construct the associated device state envelope, use the associated device state envelope to filter candidate degradation actions, obtain the target degradation action and execute it, and generate local event records. Based on the local event records, a cross-device event causal partial order graph is reconstructed, and candidate consistent causal aspects are generated. A target recovery aspect is selected through the candidate consistent causal aspects, and the branching events after the target recovery aspect are resolved in layers to restore the collaborative management and control task.

[0006] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management and control method of the present invention, the method of constructing a collaborative process causal network based on the operating status information and extracting a common confirmed causal front includes identifying collaborative task execution events and equipment status change events based on the operating status information, and setting a collaborative event node for each identified event. Based on the collaborative constraints between devices, collaborative events that constitute the premise for the occurrence of another collaborative event are marked as necessary prerequisite events. Directed causal edges are connected from the necessary prerequisite event nodes to the subsequent collaborative event nodes to construct a causal network of collaborative processes. Using the most recently confirmed collaborative event node of each intelligent manufacturing workshop equipment as the boundary, all necessary preceding event nodes are extracted in reverse along the directed causal edge to obtain the confirmed event node set of each equipment; the intersection of each confirmed event node set is processed, and the terminal collaborative event nodes in the intersection result are used to form a common confirmed causal front.

[0007] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management and control method of the present invention, wherein: taking the jointly confirmed causal frontier as the starting point of deduction, and combining the state transition diagram to calculate the causal compatible state set and collaborative recoverable state domain during the state interaction anomaly, the equipment task state corresponding to the jointly confirmed causal frontier is mapped to the state transition diagram to obtain the starting state node and current action stage of the intelligent manufacturing workshop equipment, and each state transition edge starting from the starting state node is divided into a continuous transition edge and an interrupted transition edge; Based on the current action stage, calculate the remaining reachable domain of each intelligent manufacturing workshop equipment, expand the candidate state branch from the starting state node along the continuous transition edge, and prune it using the remaining reachable domain to obtain the continuous candidate state branch; combine the continuous candidate state branches, and use the collaborative process causal network to perform causal compatibility screening to obtain the causal compatible state set. Based on the remaining collaborative tasks, the state transition diagram, and the inter-device collaborative constraints, a target set of recoverable state combinations for which the remaining collaborative tasks can continue to be executed is determined; a state transition path is searched along the state transition diagram from each state combination in the causal compatible state set to a state combination that satisfies the inter-device collaborative constraints, thus obtaining a collaboratively recoverable state domain; when the collaboratively recoverable state domain is insufficient to cover the target set of recoverable state combinations, the causal compatible state set and the collaboratively recoverable state domain are recalculated according to the interrupted transition edges; When a state combination that does not belong to the cooperative recoverable state domain appears in the causal compatible state set, it is determined that the causal compatible state set deviates from the cooperative recoverable state domain. The earliest remaining execution time of each candidate state branch after pruning is calculated to reach the physical interlock node along the state transition diagram. Based on the earliest remaining execution time, the maximum response time of the degradation action, and the safety margin, a state interaction tolerance window is generated.

[0008] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management and control method of the present invention, wherein: constructing a state envelope of associated equipment based on the state interaction tolerance window, and using the state envelope of associated equipment to screen candidate degradation actions includes: recording the duration of state interaction abnormality when a state interaction abnormality occurs between intelligent manufacturing workshop equipment, and extracting the state components of associated equipment from the causal compatible state set when the duration of the state interaction abnormality reaches the state interaction tolerance window. Based on the state transition diagram, spatial and logical projections are performed on the state components of the associated device, and state components with the same projection results are merged to form the state envelope of the associated device. Based on the current action stage and executable control method of the device, a set of candidate degradation actions is generated. The current device state after each candidate degradation action is executed is deduced along the state transition diagram. The action is then jointly verified with all associated device state combinations in the associated device state envelope. Candidate degradation actions that lack necessary preceding event nodes, do not satisfy inter-device coordination constraints, or violate physical interlock constraints are eliminated, thus forming a set of executable degradation actions.

[0009] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management and control method described in this invention, the step of obtaining and executing the target degradation action and generating a local event record includes: for the candidate degradation actions in the set of executable degradation actions, calculating the action commitment degree based on the number of irreversible state transitions added in the actual execution path of the equipment after the candidate degradation action is executed and the number of excluded legal recovery paths. Map the device state after the candidate degradation action is executed to the state transition graph, search for the recovery path from the device state to the collaborative recoverable state domain, calculate the recovery distance increment based on the number of state transition edges added to the recovery path relative to before the candidate degradation action was executed, and calculate the production continuity retention degree based on the number of collaborative task states that can be reached along the state transition graph after the candidate degradation action is executed. Candidate degradation actions are sorted in multiple levels according to the ascending order of action commitment, the ascending order of recovery distance increment, and the descending order of production continuity retention. The candidate degradation action at the top of the multiple levels is selected as the target degradation action. The target degradation action is executed, and the collaborative event nodes and action completion evidence that actually occurred in each smart manufacturing workshop equipment during the state interaction anomaly and degradation action execution are written into the local event record.

[0010] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management method described in this invention, the following steps are included: reconstructing a cross-device event causal partial order graph based on the local event records to generate candidate consistent causal sections: after the state interaction is restored, extracting collaborative event nodes and action completion evidence from the local event records of each intelligent manufacturing workshop equipment; connecting causal edges within the same intelligent manufacturing workshop equipment according to the occurrence order of collaborative event nodes; and connecting cross-device causal edges between corresponding collaborative event nodes of different intelligent manufacturing workshop equipment according to the directed causal edges in the collaborative process causal network to form a cross-device event causal partial order graph. From the cross-device event causal partial order graph, select boundary event nodes, and extract the necessary preceding event nodes corresponding to the boundary event nodes in reverse along the intra-device causal edge and cross-device causal edge to form candidate event cross-sections. Map the device task state corresponding to each candidate event aspect to the state transition diagram, verify whether each candidate event aspect contains the necessary preceding event nodes corresponding to each collaborative event node, whether the device task state satisfies the collaborative constraint relationship between devices, and whether the irreversible state transition has evidence of action completion, and take the candidate event aspect that passes all the verifications as the candidate consistent causal aspect.

[0011] As a preferred embodiment of the intelligent manufacturing workshop equipment collaborative management and control method of the present invention, wherein: a target recovery section is selected through the candidate consistent causal section, the bifurcation events after the target recovery section are resolved in layers, and the recovery collaborative management and control task includes: statistically analyzing the completed workload that can be inherited by the remaining collaborative tasks based on the action completion evidence in the candidate consistent causal section, obtaining the effective work retention of each candidate consistent causal section, and selecting the candidate consistent causal section with the largest effective work retention as the target recovery section; Extract the physical maintenance relationship formed by the bifurcation events after the target recovery section, identify reversible events in the bifurcation events based on the state transition graph, and extract the preceding reversible events that depend on maintaining the physical maintenance relationship or completing a safe exit along the cross-device event causal partial order graph. Mark the preceding reversible events as frozen events. Starting from the frozen events, traverse the bifurcation events horizontally along the cross-device event causal partial order graph, and mark other reversible events that are associated with the frozen events in terms of physical resource occupation as frozen events, thus obtaining a set of frozen events. For reversible events that are not marked as frozen events, perform state rollback; for each frozen event in the set of frozen events, select a recovery action segment from the set of recovery action segments whose entry state is consistent with the current device state and whose exit state can be mapped to the service continuation node in the state transition diagram, and execute it as a safe exit action; The device status after the safe exit action is executed is mapped to the business continuation node, and the equipment in each smart manufacturing workshop is controlled to continue to execute the remaining collaborative tasks from the business continuation node, thus restoring the collaborative management and control tasks.

[0012] A collaborative management and control system for equipment in an intelligent manufacturing workshop, wherein: The causal front module acquires the operating status information and state transition diagram of the equipment in the intelligent manufacturing workshop, constructs a collaborative process causal network based on the operating status information, and extracts the jointly confirmed causal front. The tolerance window module takes the jointly confirmed causal frontier as the starting point for deduction, combines the state transition diagram to calculate the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly, and generates a state interaction tolerance window when the causal compatible state set deviates from the cooperative recoverable state domain. The degradation decision module constructs a state envelope of associated devices based on the state interaction tolerance window, uses the state envelope of associated devices to filter candidate degradation actions, obtains the target degradation action and executes it, and generates a local event record. The collaborative management module reconstructs a cross-device event causal partial order graph based on the local event records, generates candidate consistent causal sections, selects a target recovery section through the candidate consistent causal sections, and performs hierarchical resolution on the branching events after the target recovery section to restore the collaborative management task.

[0013] A computer device includes: a memory and a processor; the memory stores a computer program, wherein: when the processor executes the computer program, it implements the steps of the method described in any one of the present invention.

[0014] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described in any one of the present invention.

[0015] The beneficial effects of this invention are as follows: By constructing a collaborative process causal network and extracting a commonly confirmed causal front, this invention converges the local task cognition under asynchronous operation of multiple devices into a unified global inference starting point, reducing inference bias caused by asymmetric cognition between devices. Through causal compatibility screening and calculation of the collaborative recoverable state domain, the boundary of state combinations with recoverable conditions is defined, and a state interaction tolerance window is generated using the earliest remaining arrival time of physically interlocked nodes, avoiding premature degradation responses or missed opportunities. By using spatial and logical projection to perform low-dimensional compression of the multi-device state combination space, joint verification is completed using the state envelope of associated devices, reducing the computational overhead of real-time verification. By constructing a set of frozen events and using layer-by-layer linkage marking associated with physical resource occupancy, reversible events that are not suitable for direct rollback due to physical maintenance relationships or safe exit requirements are distinguished from other reversible events that can be directly rolled back. A safe exit action with exit mapping guarantee is executed on frozen events, achieving safe alignment of each device to the business continuation node, enabling smooth continuation of workshop collaborative management tasks while retaining completed workload. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is an overall flowchart of a collaborative management and control method for equipment in an intelligent manufacturing workshop, provided as an embodiment of the present invention.

[0018] Figure 2 A flowchart for generating a state interaction tolerance window provided in an embodiment of the present invention.

[0019] Figure 3 A computer equipment diagram for a collaborative management and control method for intelligent manufacturing workshop equipment provided in an embodiment of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0021] Example 1, referring to Figures 1-3As an embodiment of the present invention, a method for collaborative management and control of equipment in an intelligent manufacturing workshop is provided, comprising: S1: Obtain the operating status information and state transition diagram of the equipment in the intelligent manufacturing workshop, construct a collaborative process causal network based on the operating status information, and extract the common confirmed causal front.

[0022] In this embodiment, the intelligent manufacturing workshop includes multiple collaboratively operating devices, such as robotic arms, conveyor belts, and welding units. Each device continuously generates two types of events while performing collaborative tasks: collaborative task execution events (e.g., the robotic arm completing a grasping action, the conveyor belt confirming its arrival) and device state change events (e.g., the welding unit switching from standby to working state). Due to the different control cycles of each device and the uncertain delays in network transmission, each device's perception of the overall progress of the workshop tasks is asynchronous. That is, an event recorded and confirmed by one device may not have been received and confirmed by another device. If the latest confirmed event of a certain device is directly used as the starting point for global inference, the event basis upon which the inference relies only reflects a partial view of the corresponding device. Events not yet confirmed by other devices will be incorrectly included in the inference scope, leading to subsequent anomaly handling decisions being based on incomplete cognition, thus causing cross-device collaborative errors. Therefore, it is necessary to introduce a collaborative process causal network and a jointly confirmed causal front mechanism to establish a global inference benchmark recognized by all devices.

[0023] Specifically, the process involves acquiring the operational status information and state transition diagrams of equipment in the intelligent manufacturing workshop. This includes: each piece of equipment continuously reporting its operational status information through its own controller during operation. The operational status information comprises two categories: firstly, real-time status data, including the current state node number of the equipment, the type of action currently being performed, the action start timestamp, and the action completion timestamp; secondly, event stream data, including event records generated by each equipment controller at the moment of action switching. Each event record includes an event type identifier (collaborative task execution event or equipment state change event), the equipment number that triggered the event, the event occurrence timestamp, and the preceding event number associated with the event. This operational status information is reported by each equipment controller to the workshop data acquisition layer at a fixed cycle. The acquisition cycle is determined based on the equipment control cycle. For example, for a robotic arm controller with a control cycle of 10 milliseconds, the acquisition cycle can be set to 10 milliseconds to ensure the completeness and timing accuracy of the event records.

[0024] Secondly, the state transition diagram is a directed graph pre-built for each piece of equipment in the smart manufacturing workshop. It describes all the state nodes that the equipment may go through and their transition relationships. It is jointly determined by the equipment control logic document and workshop process specifications provided by the equipment manufacturer and is a known input that is built offline and then used online. Each state node in the state transition diagram corresponds to a definite state of the equipment in a certain process stage. For example, the state nodes of a robotic arm include initial position standby, movement to gripping position, gripper closing, lifting workpiece, movement to placement position, gripper opening, and returning to initial position. Each state transition edge corresponds to the action of the equipment switching from one state node to another. The transition edge is marked with the estimated execution time of the transition (unit: milliseconds, jointly determined by the equipment factory test data and on-site debugging data) and whether the transition is irreversible.

[0025] Specifically, transition edges in the state transition diagram with reverse transition paths are marked as reversible transition edges; transition edges that cannot return to the previous state after reaching it through any control action are marked as irreversible transition edges. For example, after the welding unit completes the spot welding action, the workpiece is permanently connected, and the corresponding transition edge is marked as irreversible. In addition, physical interlock nodes must be pre-marked in the state transition diagram, that is, state nodes that will trigger irreversible physical constraints after reaching them, such as the state node corresponding to the robotic arm moving to the clamping limit position, and the state node corresponding to the conveyor belt reaching the end stop position. The position of the physical interlock nodes is determined by the physical structure parameters of the equipment and the hard limit settings in the control logic, and the marking is completed in the offline creation stage of the state transition diagram.

[0026] Furthermore, after acquiring the operating status information and state transition diagram of each device, the event flow data of each device provides the original input for the identification of subsequent collaborative event nodes, and the state transition diagram provides the structural basis for subsequent state deduction and path search. Together, they support the construction of the causal network of collaborative processes.

[0027] Specifically, constructing a collaborative process causal network based on the operational status information and extracting a common causal front includes the following steps: identifying collaborative task execution events and equipment status change events generated by each device based on the operational status information, and setting a corresponding collaborative event node for each identified event.

[0028] Based on the inter-equipment collaborative constraints, the dependency directions between each collaborative event node are determined. These constraints refer to the fact that, in the workshop process, the occurrence of one collaborative event is contingent upon the completion of another collaborative event; for example, the start of a welding unit requires the confirmation of the conveyor belt's arrival. Collaborative events that constitute the prerequisites for another collaborative event are marked as necessary precursor events. Directed causal edges are connected from necessary precursor event nodes to subsequent collaborative event nodes to construct a collaborative process causal network. This collaborative process causal network fully characterizes the causal dependencies between various equipment events within the workshop.

[0029] After the collaborative process causal network is constructed, the most recently confirmed collaborative event node of each intelligent manufacturing workshop equipment is used as the boundary. The system then traverses backward along the directed causal edge to extract the set of all necessary preceding event nodes that can be reached backward from the boundary node, thus obtaining the confirmed event node set of each equipment. The confirmed event node sets of each equipment are then intersected, and the terminal collaborative event nodes in the intersection result form a common confirmed causal front. The terminal collaborative event node refers to a node in the intersection result that has no subsequent event nodes belonging to the intersection, representing the common progress boundary that has been confirmed by all equipment and is closest to the current time in causal order.

[0030] For example, taking a three-device scenario as an example, suppose the robotic arm has confirmed the set of collaborative event nodes. The conveyor belt has been confirmed to be connected to the collaborative event node set. The welding unit has been confirmed to be part of the collaborative event node set. The intersection of the event node sets confirmed by the three devices is: The intersection operation uses the set intersection method to find nodes that appear simultaneously in each set; where Since there are no successor nodes in the intersection, it is the terminal node. This constitutes a common causal frontier.

[0031] It should be noted that, in order to address the problem that inconsistent perceptions of task progress among devices due to asynchronous operation and transmission delays, and the issue that using a single device's local view as the starting point for inference will lead to cross-device collaboration errors, the problem is that by taking the intersection end node of the set of confirmed event nodes of each device in the collaborative process causal network, a common confirmed causal front recognized by all devices is formed, thereby reducing the inference bias caused by the misalignment of perceptions among devices.

[0032] S2: Taking the jointly confirmed causal frontier as the starting point of deduction, and combining the state transition diagram, calculate the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly, and generate a state interaction tolerance window when the causal compatible state set deviates from the cooperative recoverable state domain.

[0033] After determining the common causal front in step S1, abnormal state interactions may occur between workshop equipment due to communication failures, process cycle deviations, etc. At this time, it is necessary to clarify the following two points: First, among the possible state combinations that each equipment may experience during the abnormality period, which are compatible with the confirmed causal history and which cannot be restored to normal coordination; Second, how much time is left before the equipment triggers the physical interlock node so that the degradation response can be completed within this time. The lack of quantitative characterization of these two points may lead to premature production interruption or delay in degradation, resulting in the physical interlock being triggered and the task recovery path being completely closed. It is necessary to calculate the causal compatible state set, the cooperative recoverable state domain, and generate the state interaction tolerance window in sequence based on the common causal front and the state transition diagram.

[0034] Specifically, such as Figure 2 The diagram shows the process for generating a state interaction tolerance window. Starting with the jointly confirmed causal frontier, and combining it with the state transition diagram, it calculates the causally compatible state set and the cooperatively recoverable state domain during state interaction anomalies. The process includes the following steps: mapping the task states of each device corresponding to the jointly confirmed causal frontier to the state transition diagram of each device, obtaining the initial state node and current action stage of each smart manufacturing workshop device. The state transition diagram is a directed graph pre-established for each device, describing all possible state nodes and their transition relationships. Each state transition edge originating from the initial state node of each device is divided into continuous transition edges and interrupted transition edges; a continuous transition edge refers to a transition edge that continues along the current action stage; an interrupted transition edge refers to a transition edge that deviates from the current action stage.

[0035] Furthermore, based on the current action stage, the remaining reachable domain of each intelligent manufacturing workshop equipment is calculated using the directed graph reachability analysis method. That is, the set of state nodes that can be reached from the starting state node under the constraints of the current action stage along the state transition graph. The specific calculation process is as follows: Starting from the starting state node, gradually expand along the transition edges in the state transition graph that belong to the current action stage, and record all reachable state nodes until no new nodes can be added.

[0036] The candidate state branches are expanded from the starting state node along the continuous transition edge, and the remaining reachable domain is used for pruning to remove state nodes that are not in the remaining reachable domain, thus obtaining the continuous candidate state branches of each smart manufacturing workshop equipment.

[0037] The continuous candidate state branches of each intelligent manufacturing workshop equipment are combined to form a multi-equipment state combination set. For each state combination, causal compatibility screening is performed using the collaborative process causal network. Specifically, if the task progress corresponding to a certain state combination implies that a necessary prerequisite event has not yet occurred, but the jointly confirmed causal frontier has recorded that the corresponding necessary prerequisite event has been completed, then the state combination is determined to be inconsistent with the causal history and is eliminated. The state combinations that pass the causal compatibility screening are retained to obtain the causal compatible state set. .

[0038] Further, based on the remaining collaborative tasks, the state transition diagram, and the inter-device collaborative constraints, a set of target recovery state combinations for continuing the remaining collaborative tasks is determined. This set of target recovery state combinations refers to the set of state combinations that satisfy the inter-device collaborative constraints in the state transition diagram and, starting from the corresponding state combination, can continue executing all remaining collaborative task steps. Along the state transition diagram, for the causally compatible state set... For each state combination in the algorithm, a directed graph reachability analysis method is used to search for state transition paths that can be reached from the state combination and satisfy the inter-device cooperative constraints. All initial state combinations that can reach the state combinations satisfying the inter-device cooperative constraints through these paths are then identified, yielding the cooperatively recoverable state domain. When the cooperative recoverable state domain If the target recovery state combination set is insufficient to cover the target recovery state combination set, the candidate state branches are expanded again according to the interruption-type transition edges to obtain interruption-type candidate state branches, and the causal compatible state set and cooperative recoverable state domain are recalculated according to the same process as above to include more candidate recovery paths.

[0039] In one alternative embodiment, if the collaborative recoverable state domain is still insufficient to cover the target recovery state combination set after recalculation according to the interrupted transfer edge, the current calculation result is retained and the generation of subsequent tolerance windows continues.

[0040] Furthermore, when the set of causally compatible states The state does not belong to the cooperative recoverable state domain. When the state combination is determined to deviate from the cooperative recoverable state domain, it indicates that under the current abnormal continuous situation, some state combinations cannot recover to normal cooperation on their own and need to complete the degradation response before the physical interlock node is triggered.

[0041] Specifically, the earliest remaining execution time for each candidate state branch after pruning to reach the physical interlock node along the state transition diagram is calculated. A physical interlock node is a state node that will trigger irreversible physical constraints upon arrival, such as a robotic arm moving to its clamping limit position or a conveyor belt reaching its end stop. Its position is pre-marked and determined in the state transition diagram. The earliest remaining execution time for each candidate state branch to reach the physical interlock node is solved using the shortest path calculation method on the state transition diagram. The specific calculation process is as follows: taking the current end node of the candidate state branch as the starting point and the physical interlock node as the ending point, assign a corresponding estimated execution time (unit: milliseconds, pre-defined by equipment control cycle and process cycle data) to each transition edge in the state transition diagram. Calculate the shortest weighted path length from the current end node to the nearest physical interlock node, which is the earliest remaining execution time of the candidate state branch. .

[0042] The earliest remaining execution time to reach the physical interlock node among all candidate state branches. Based on this, and combined with the preset maximum response time of degradation actions and safety margin, a state interaction tolerance window is generated. The calculation formula is as follows: ; in, It is the set of all candidate state branches after pruning, including the continuation candidate state branches and all newly added interrupted candidate state branches after being expanded according to the interrupted transition edges; Branch for candidate states The earliest remaining execution time, in milliseconds, required to reach the nearest physical interlock node along the state transition graph; The maximum response time for a degradation action is the maximum time required from triggering the degradation action selection to completing the execution of the target degradation action, measured in milliseconds. It is preset by the device controller's command response latency and the size of the candidate degradation action set. For safety margin, the unit is milliseconds, determined based on the statistical upper bound of workshop network communication latency, for example, taking 1.5 times the maximum historical communication latency. The state interaction tolerance window represents the maximum allowable tolerance time for state interaction anomalies, in milliseconds, provided that the degradation action can be completed before the physical interlock node is triggered.

[0043] It should be noted that the lack of quantitative characterization of the anomaly tolerance time leads to the problem of downgrading too early or too late. By using causal compatibility screening to clarify the range of state combinations that are consistent with the confirmed causal history during the anomaly period, and by calculating the collaborative recoverable state domain to delineate the boundary of state combinations that still have the conditions to recover to normal collaborative conditions, and by generating a state interaction tolerance time window based on the earliest remaining arrival time of the physical interlock node, combined with the maximum response time of the downgrading action and the safety margin, it is ensured that the downgrading action is completed before any candidate state branch triggers the physical interlock node.

[0044] S3: Construct the associated device state envelope based on the state interaction tolerance window, use the associated device state envelope to filter candidate degradation actions, obtain the target degradation action and execute it, and generate local event records.

[0045] Furthermore, step S2 has determined the maximum tolerance duration for abnormal state interactions. When the duration of a recorded abnormal state interaction exceeds the maximum tolerance duration, a degradation action needs to be selected from the currently executable control actions and executed to prevent the physical interlocking node from being triggered. However, the state combination space of multiple smart manufacturing workshop devices expands with the increase in the number of devices and states. If each candidate degradation action is verified item by item in the complete multi-device state combination space, the computational overhead is difficult to meet the real-time response requirements. In addition, when multiple candidate degradation actions pass the compliance verification, the best one needs to be selected from them according to clear evaluation criteria so that the degradation decision takes into account security, subsequent recovery costs, and production losses.

[0046] Specifically, constructing the associated device state envelope based on the state interaction tolerance window includes the following steps: recording the duration of state interaction anomalies when they occur between devices in the smart manufacturing workshop. When the state interaction abnormal duration Reaching the state interaction tolerance window At that time, the set of causal compatible states obtained from step S2 Extract the state components of associated devices, where associated devices refer to other devices that have a collaborative constraint relationship with the device currently experiencing a state interaction anomaly.

[0047] Based on the state transition diagram, spatial projection and logical projection are performed on the state components of the associated devices respectively: Spatial projection refers to expanding the state components along the physical resource occupancy dimension and extracting the physical resource occupancy corresponding to each state component, such as the work area number occupied by the robotic arm and the workstation number currently occupied by the conveyor belt. The specific calculation process is as follows: extract the physical resource occupancy field corresponding to each dimension in the state component one by one to form the spatial projection vector of the state component. The specific formula is as follows: ; in, For the state components of the associated device; State components The corresponding number The occupancy identifier for physical resources can be either the resource number or an empty occupancy identifier. Physical resources The quantity is determined by the type of workshop resources.

[0048] Logical projection refers to expanding state components along the task phase dimension and extracting the task execution phase number corresponding to each state component. The specific calculation process is as follows: read the task phase field from the state component to form the logical projection scalar of the state component. The specific formula is as follows: ; in, State components The corresponding task execution stage number is determined by the encoding value of the current action stage in the device control logic.

[0049] For those with the same spatial projection vector and the same logical projection scalar The state components are merged to form the associated device state envelope. Among them, the associated device state envelope is a low-dimensional compressed representation of the associated device states in the causal compatible state set, which reduces the number of verification objects while covering all valid associated device state boundaries.

[0050] Further, the process of filtering candidate degradation actions using the associated device state envelope includes the following steps: First, enumerating and generating a set of candidate degradation actions based on the current action stage of the device and the executable control methods, wherein the executable control methods are determined by the set of instructions allowed to be issued by the device controller in the current state. Second, deducing the current device state after each candidate degradation action is executed along the state transition diagram to obtain the post-execution device state corresponding to each candidate degradation action. Third, associating each post-execution device state with the associated device state envelope. The status of all associated devices is jointly verified item by item, and candidate degradation actions that meet any of the following conditions are eliminated in turn: the device status after execution lacks the necessary preceding event nodes specified in the causal network of the collaborative process; the combination of the device status after execution and the status of associated devices does not satisfy the collaborative constraint relationship between devices; or the device status after execution or the status of associated devices violates the physical interlock constraint. The candidate degradation actions retained after the above three verifications form the set of executable degradation actions. .

[0051] Further, the target degradation action is obtained and executed, generating a local event log, including the following steps: for the set of executable degradation actions... Candidate downgrade actions Evaluation indicators are calculated from the following three dimensions. All three indicators are dimensionless quantities and can be directly used for multi-level ranking comparisons.

[0052] First, commitment to action By based on candidate downgrade actions The number of irreversible state transitions added to the actual execution path of the device after execution, and the number of candidate degradation actions performed. The number of excluded legitimate recovery paths is used to calculate the action commitment degree, using the following formula: ; in, The number of irreversible state transitions added in the actual execution path of the device (unit: entries) is compared with candidate degradation actions. The difference between the set of irreversible state transitions that have occurred before execution and the set of irreversible state transitions that have occurred cumulatively in the actual execution path of the device after execution is obtained by counting the number of elements in the set. The number of excluded legitimate recovery paths (in terms of paths) is calculated by determining the candidate degradation actions. Able to reach the cooperative recoverable state domain before execution The difference between the number of valid paths and the number of valid paths that can reach the cooperative recoverable state domain after execution is obtained; For action commitment, The smaller the value, the more likely it is to indicate a candidate for downgrading. The less the space for subsequent decision-making is narrowed.

[0053] Second, restore distance increment Downgrade candidate actions The executed device state is mapped to a state transition graph, and the shortest path calculation method is used to search for the path from the corresponding device state to the cooperative recoverable state domain. The shortest recovery path, whose number of state transition edges is denoted as . (Unit: Items); Candidate Downgrade Actions The number of state transition edges for the shortest recovery path to enter the cooperative recoverable state domain before execution is denoted as . (Unit: Items); The recovery distance increment is calculated based on the number of state transition edges added to the recovery path relative to the number of candidate degradation actions executed before the action was performed. The specific formula is as follows: ; in, To restore distance increments, the unit is bars; The smaller the value, the more likely it is that a candidate downgrade action will be performed. The lower the additional cost required to restore normal coordination.

[0054] Third, the degree of production continuity retention The production continuity retention level is calculated based on the number of collaborative task states reachable along the state transition graph after the candidate degradation action is executed. This includes: traversing the state transition graph from the post-execution equipment state nodes using a directed graph reachability analysis method to obtain candidate degradation actions. The number of collaborative task states that can be reached after the execution of the state transition graph, i.e., the degree of production continuity retention. ; The larger the value, the more likely it is to trigger a candidate downgrade action. The higher the continuity of the subsequent production process.

[0055] Furthermore, according to the degree of action commitment Ascending order, restore distance increment Ascending order and degree of production continuity retention Descending order, for the set of executable degradation actions. All candidate degradation actions are ranked in a multi-level order, and the top-ranked candidate degradation action is selected as the target degradation action. The target degradation action is then executed, and the abnormal state interaction and the actual collaborative event nodes and action completion evidence that occur in each smart manufacturing workshop device during the degradation action execution are continuously written into the local event record of each device. The action completion evidence refers to the state confirmation signal generated by the device controller after the target degradation action is completed, which includes the action type, completion timestamp, and execution result identifier, used to verify the legality of irreversible state transitions.

[0056] The priority order of the three-dimensional multi-level arrangement is determined according to the decision-making logic of the industrial safety scenario: First, the action commitment is the first ranking criterion, and the action with the lowest narrowing of the subsequent decision space is selected first; second, the recovery distance increment is the second ranking criterion, and the action with the lowest subsequent recovery cost is selected first when the commitment is the same; finally, the degree of production continuity retention is the third ranking criterion, and more sustainable production states are retained first when the first two indicators are the same.

[0057] It should be noted that, in response to the problem of high real-time verification costs for candidate degradation actions due to the excessively large multi-device state combination space, and the lack of selection criteria when multiple candidate actions coexist, the multi-device state combination space is compressed to a low dimension through spatial projection and logical projection. Only state components with the same judgment result for three types of verification—completeness of necessary pre-event nodes, inter-device collaborative constraints, and physical interlock constraints—are merged into the same envelope term. The joint verification is completed by replacing the full enumeration of state combinations with the envelope of associated device states, thereby reducing computational overhead while ensuring verification completeness. Under the premise of satisfying safety constraints, the three-dimensional multi-level permutation evaluation system orderly distinguishes candidate degradation actions from three dimensions: irreversible commitment, recovery cost, and production loss, so that degradation decisions have clear interpretability.

[0058] S4: Reconstruct the cross-device event causal partial order graph based on the local event records, generate candidate consistent causal sections, select the target recovery section through the candidate consistent causal sections, perform hierarchical resolution on the branching events after the target recovery section, and restore the collaborative management and control task.

[0059] In this embodiment, after step S3 completes the execution of the target degradation action and generates local event records for each smart manufacturing workshop device, each smart manufacturing workshop device saves the event sequence from its own perspective during the degradation period. However, due to the different event paths experienced by each device during the degradation period, there are discrepancies in the perception of task progress among the devices: if task recovery is initiated directly from the local event view of a certain device, the events that have not yet been synchronized by other devices will lead to inconsistencies in cross-device states. The recovered collaborative task will continue to be executed on the basis of the inconsistent state, causing subsequent collaborative errors. In addition, some devices may have completed irreversible physical actions during the degradation period, such as the robotic arm completing workpiece clamping. Simple state rollback will destroy the consistency of physical states and may even cause equipment damage or safety accidents. Therefore, safe and consistent collaborative task recovery is required.

[0060] Specifically, a cross-device event causal partial order graph is reconstructed based on the local event records to generate candidate consistent causal aspects. The specific steps are as follows: After the state interaction is restored, collaborative event nodes and action completion evidence are extracted from the local event records of each smart manufacturing workshop device; causal edges within the same smart manufacturing workshop device are connected according to the occurrence order of collaborative event nodes within the same smart manufacturing workshop device; based on the directed causal edges in the collaborative process causal network constructed in step S1, cross-device causal edges are connected between corresponding collaborative event nodes in different smart manufacturing workshop devices to form a cross-device event causal partial order graph. The cross-device event causal partial order graph integrates the local event records of each smart manufacturing workshop device into a unified global event structure, where the partial order relationship is jointly determined by the causal edges within the device and the cross-device causal edges, reflecting the sequential causal constraints between events.

[0061] The selection of boundary event nodes from the cross-device event causal partial order graph is specifically as follows: traverse all cooperating event nodes in the cross-device event causal partial order graph, take each node as a candidate boundary event node, traverse in reverse along the intra-device causal edge and cross-device causal edge, extract all necessary preceding event nodes that the candidate boundary event node depends on, and form a candidate event section together with the candidate boundary event node and all the corresponding necessary preceding event nodes.

[0062] The device task states corresponding to each candidate event segment are mapped to the state transition diagram. Three checks are performed sequentially on each candidate event segment: First, whether the candidate event segment contains all the necessary preceding event nodes specified in the causal network of the collaborative process for each collaborative event node in the candidate event segment; second, whether the device task states of each smart manufacturing workshop corresponding to the candidate event segment satisfy the collaborative constraint relationship between devices; and third, whether each irreversible state transition involved in the candidate event segment has evidence of completion of the action written to the local event record in step S3. Candidate event segments that pass all of the above checks are designated as candidate consistent causal segments.

[0063] Specifically, selecting the target recovery aspect through the candidate consistent causal aspects includes the following steps: based on the action completion evidence in the candidate consistent causal aspects, calculating the amount of completed work that can be directly inherited by the remaining collaborative tasks, and obtaining the effective work retention of each candidate consistent causal aspect. The effective work retention is calculated as follows: The number of collaborative task steps corresponding to irreversible state transitions with evidence of action completion in candidate consistent causal sections is counted. The specific formula is as follows: ; in, For candidate consistent causal cross-sections The set of all irreversible state transitions involved in the process; Irreversible state transition The boolean value indicating whether there is evidence of action completion is determined by the existence of the corresponding entry in the local event log; The number of elements in the set; The effective working retention for candidate consistent causal sections, in steps. The larger the value, the more consistent the causal aspect from the candidate. The less redo work is required to initiate recovery, the more likely the candidate consistent causal section with the largest effective work retention will be selected as the target recovery section, so as to minimize task redo and reduce production losses.

[0064] Further, the bifurcation events after the target recovery section are resolved hierarchically to restore the collaborative control task, including the following steps: Extracting the physical maintenance relationships formed by all bifurcation events occurring after the target recovery section; the physical maintenance relationship refers to the continued maintenance of the physical state corresponding to certain bifurcation events after the target recovery section is determined, such as a robotic arm gripper maintaining a clamping state after the gripping action is completed. Identifying reversible events in the bifurcation events based on the state transition graph, i.e., events with reverse transition paths in the state transition graph that can be undone by control actions; extracting the following two types of preceding reversible events in reverse along the cross-device event causal partial order graph: first, preceding reversible events on which maintaining the physical maintenance relationship depends; second, preceding reversible events on which completing a safe exit depends; marking the above two types of preceding reversible events as frozen events.

[0065] After initial freezing event marking is completed, starting from all marked freezing events, the branching events are traversed laterally along the cross-device event causal partial order graph: other reversible event nodes that are associated with the marked freezing events in terms of physical resource occupation are retrieved, where physical resource occupation association means that the two events involve resource occupation of the same physical device component or the same working area; the retrieved other reversible event nodes are linked and marked as freezing events; the above linked marking process is repeated until no new reversible event nodes are marked, finally obtaining the set of freezing events. In this embodiment, the horizontal traversal is implemented by expanding layer by layer: first, reversible events that are directly associated with the initial freeze event in terms of physical resource occupation are retrieved, marked, and then the retrieval continues from the newly marked freeze event as the starting point, expanding outward layer by layer to ensure that the transitivity of physical resource occupation associations is completely covered.

[0066] Perform a state rollback on all reversible events not marked as frozen events, that is, restore the corresponding device state to the state at the target recovery plane along the reverse transition path in the state transition diagram. This applies to the set of frozen events. For each freeze event, a recovery action segment is searched from a pre-established set of recovery action segments that meets the following two conditions: the entry state is consistent with the current device state; the exit state can be mapped to the service continuation node in the state transition diagram; and the recovery action segment that meets the conditions is executed as a safe exit action. Here, the service continuation node refers to a state node in the state transition diagram that can legally continue executing the remaining collaborative tasks from the target recovery aspect. It is determined by the node that satisfies the inter-device collaborative constraint relationship among the subsequent reachable nodes of the corresponding state in the target recovery aspect of the state transition diagram.

[0067] In one optional embodiment, when there are multiple recovery action segments that meet the conditions, they can be sorted in ascending order according to the number of state transition steps required to reach the service continuation node after each recovery action segment is executed, and the recovery action segment with the shortest recovery path is selected first to shorten the time required for safe exit.

[0068] Finally, the device status after all safe exit actions are completed is mapped to their respective business continuation nodes, and the devices in each smart manufacturing workshop are controlled to continue executing the remaining collaborative tasks from the business continuation nodes, thus restoring the collaborative management and control tasks.

[0069] It should be noted that, regarding the discrepancies in local event records of various intelligent manufacturing workshop equipment during the downgrade period, directly initiating recovery from a local view would lead to inconsistencies in cross-device states. This is addressed by reconstructing the cross-device event causal partial order graph to integrate local events of each device into a globally unified structure. Three checks are then performed to select candidate consistent causal aspects that satisfy causal integrity, collaborative constraints, and the legality of irreversible transfers. To address the issue that the irreversibility of physical processes may cause safety incidents due to simple state rollbacks, the construction of the frozen event set uses a layer-by-layer linkage marker associated with physical resource occupancy to distinguish between irreversible events requiring safe handling and reversible events that can be directly rolled back. Each frozen event in the frozen event set is then subjected to a safe exit action with exit mapping guarantees, rather than a forced rollback. While preserving the target recovery aspects of completed workload to the greatest extent possible, the state of each device is safely aligned with the business continuation node, enabling smooth continuation of collaborative management tasks under the premise of consistent physical states and satisfied collaborative constraints.

[0070] On the other hand, this embodiment also provides an intelligent manufacturing workshop equipment collaborative management and control system, which includes: The causal front module acquires the operating status information and state transition diagram of the equipment in the intelligent manufacturing workshop, constructs a collaborative process causal network based on the operating status information, and extracts the jointly confirmed causal front.

[0071] The tolerance window module takes the jointly confirmed causal frontier as the starting point for deduction, combines the state transition diagram to calculate the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly, and generates a state interaction tolerance window when the causal compatible state set deviates from the cooperative recoverable state domain.

[0072] The degradation decision module constructs a state envelope of associated devices based on the state interaction tolerance window, uses the state envelope of associated devices to filter candidate degradation actions, obtains the target degradation action and executes it, and generates local event records.

[0073] The collaborative management module reconstructs a cross-device event causal partial order graph based on the local event records, generates candidate consistent causal sections, selects a target recovery section through the candidate consistent causal sections, and performs hierarchical resolution on the branching events after the target recovery section to restore the collaborative management task.

[0074] like Figure 3 As shown, if the above functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0075] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0076] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0077] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0078] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for collaborative management and control of equipment in an intelligent manufacturing workshop, characterized in that, include: Acquire the operating status information and state transition diagram of equipment in the intelligent manufacturing workshop, construct a collaborative process causal network based on the operating status information, and extract the common confirmed causal front. Taking the jointly confirmed causal frontier as the starting point of deduction, the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly are calculated in combination with the state transition diagram, and a state interaction tolerance window is generated when the causal compatible state set deviates from the cooperative recoverable state domain. Based on the state interaction tolerance window, construct the associated device state envelope, use the associated device state envelope to filter candidate degradation actions, obtain the target degradation action and execute it, and generate local event records. Based on the local event records, a cross-device event causal partial order graph is reconstructed, and candidate consistent causal aspects are generated. A target recovery aspect is selected through the candidate consistent causal aspects, and the branching events after the target recovery aspect are resolved in layers to restore the collaborative management and control task.

2. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 1, characterized in that: Constructing a collaborative process causal network based on the operational status information and extracting a jointly confirmed causal front includes identifying collaborative task execution events and equipment status change events based on the operational status information, and setting a collaborative event node for each identified event; Based on the collaborative constraints between devices, collaborative events that constitute the premise for the occurrence of another collaborative event are marked as necessary prerequisite events. Directed causal edges are connected from the necessary prerequisite event nodes to the subsequent collaborative event nodes to construct a causal network of collaborative processes. Using the most recently confirmed collaborative event node of each intelligent manufacturing workshop equipment as the boundary, all necessary preceding event nodes are extracted in reverse along the directed causal edge to obtain the confirmed event node set of each equipment; the intersection of each confirmed event node set is processed, and the terminal collaborative event nodes in the intersection result are used to form a common confirmed causal front.

3. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 2, characterized in that: Taking the jointly confirmed causal frontier as the starting point of deduction, and combining the state transition diagram to calculate the causal compatible state set and cooperative recoverable state domain during the state interaction anomaly, the device task state corresponding to the jointly confirmed causal frontier is mapped to the state transition diagram to obtain the starting state node and current action stage of the intelligent manufacturing workshop equipment, and each state transition edge starting from the starting state node is divided into continuous transition edge and interrupted transition edge. Based on the current action stage, calculate the remaining reachable domain of each intelligent manufacturing workshop equipment, expand the candidate state branch from the starting state node along the continuous transition edge, and prune it using the remaining reachable domain to obtain the continuous candidate state branch; combine the continuous candidate state branches, and use the collaborative process causal network to perform causal compatibility screening to obtain the causal compatible state set. Based on the remaining collaborative tasks, the state transition diagram, and the inter-device collaborative constraints, a target set of recoverable state combinations for which the remaining collaborative tasks can continue to be executed is determined; a state transition path is searched along the state transition diagram from each state combination in the causal compatible state set to a state combination that satisfies the inter-device collaborative constraints, thus obtaining a collaboratively recoverable state domain; when the collaboratively recoverable state domain is insufficient to cover the target set of recoverable state combinations, the causal compatible state set and the collaboratively recoverable state domain are recalculated according to the interrupted transition edges; When a state combination that does not belong to the cooperative recoverable state domain appears in the causal compatible state set, it is determined that the causal compatible state set deviates from the cooperative recoverable state domain. The earliest remaining execution time of each candidate state branch after pruning is calculated to reach the physical interlock node along the state transition diagram. Based on the earliest remaining execution time, the maximum response time of the degradation action, and the safety margin, a state interaction tolerance window is generated.

4. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 3, characterized in that: Based on the state interaction tolerance window, a state envelope of associated devices is constructed. The process of filtering candidate degradation actions using the state envelope of associated devices includes recording the duration of state interaction anomalies when they occur between devices in the smart manufacturing workshop. When the duration of the state interaction anomaly reaches the state interaction tolerance window, the state components of associated devices are extracted from the causal compatible state set. Based on the state transition diagram, spatial and logical projections are performed on the state components of the associated device, and state components with the same projection results are merged to form the state envelope of the associated device. Based on the current action stage and executable control method of the device, a set of candidate degradation actions is generated. The current device state after each candidate degradation action is executed is deduced along the state transition diagram. The action is then jointly verified with all associated device state combinations in the associated device state envelope. Candidate degradation actions that lack necessary preceding event nodes, do not satisfy inter-device coordination constraints, or violate physical interlock constraints are eliminated, thus forming a set of executable degradation actions.

5. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 4, characterized in that: The process of obtaining and executing the target degradation action and generating a local event record includes, for the candidate degradation actions in the set of executable degradation actions, calculating the action commitment based on the number of irreversible state transitions added to the actual execution path of the device after the candidate degradation action is executed and the number of excluded legal recovery paths. Map the device state after the candidate degradation action is executed to the state transition graph, search for the recovery path from the device state to the collaborative recoverable state domain, calculate the recovery distance increment based on the number of state transition edges added to the recovery path relative to before the candidate degradation action was executed, and calculate the production continuity retention degree based on the number of collaborative task states that can be reached along the state transition graph after the candidate degradation action is executed. Candidate degradation actions are sorted in multiple levels according to the ascending order of action commitment, the ascending order of recovery distance increment, and the descending order of production continuity retention. The candidate degradation action at the top of the multiple levels is selected as the target degradation action. The target degradation action is executed, and the collaborative event nodes and action completion evidence that actually occurred in each smart manufacturing workshop equipment during the state interaction anomaly and degradation action execution are written into the local event record.

6. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 5, characterized in that: Reconstructing a cross-device event causal partial order graph based on the local event records and generating candidate consistent causal sections includes: after the state interaction is restored, extracting collaborative event nodes and action completion evidence from the local event records of each smart manufacturing workshop device; connecting causal edges within the device according to the occurrence order of collaborative event nodes within the same smart manufacturing workshop device; and connecting cross-device causal edges between corresponding collaborative event nodes of different smart manufacturing workshop devices according to the directed causal edges in the collaborative process causal network to form a cross-device event causal partial order graph. From the cross-device event causal partial order graph, select boundary event nodes, and extract the necessary preceding event nodes corresponding to the boundary event nodes in reverse along the intra-device causal edge and cross-device causal edge to form candidate event cross-sections. Map the device task state corresponding to each candidate event aspect to the state transition diagram, verify whether each candidate event aspect contains the necessary preceding event nodes corresponding to each collaborative event node, whether the device task state satisfies the collaborative constraint relationship between devices, and whether the irreversible state transition has evidence of action completion, and take the candidate event aspect that passes all the verifications as the candidate consistent causal aspect.

7. The intelligent manufacturing workshop equipment collaborative management and control method as described in claim 6, characterized in that: The target recovery section is selected by the candidate consistent causal section, and the bifurcation events after the target recovery section are resolved in layers. The recovery collaborative management task includes: statistically analyzing the completed workload that can be inherited by the remaining collaborative tasks based on the action completion evidence in the candidate consistent causal section, obtaining the effective work retention of each candidate consistent causal section, and selecting the candidate consistent causal section with the largest effective work retention as the target recovery section. Extract the physical maintenance relationship formed by the bifurcation events after the target recovery section, identify the reversible events in the bifurcation events based on the state transition diagram, extract the preceding reversible events on which the physical maintenance relationship is maintained or the safe exit is completed along the cross-device event causal partial order graph, and mark the preceding reversible events as freeze events. Starting from the freezing event, the branching events are traversed horizontally along the cross-device event causal partial order graph. Other reversible events that are associated with the freezing event in terms of physical resource occupation are linked and marked as freezing events to obtain a set of freezing events. For reversible events that are not marked as frozen events, perform state rollback; for each frozen event in the set of frozen events, select a recovery action segment from the set of recovery action segments whose entry state is consistent with the current device state and whose exit state can be mapped to the service continuation node in the state transition diagram, and execute it as a safe exit action; The device status after the safe exit action is executed is mapped to the business continuation node, and the equipment in each smart manufacturing workshop is controlled to continue to execute the remaining collaborative tasks from the business continuation node, thus restoring the collaborative management and control tasks.

8. A collaborative management and control system for intelligent manufacturing workshop equipment using the method described in any one of claims 1-7, characterized in that: The causal front module acquires the operating status information and state transition diagram of the equipment in the intelligent manufacturing workshop, constructs a collaborative process causal network based on the operating status information, and extracts the jointly confirmed causal front. The tolerance window module takes the jointly confirmed causal frontier as the starting point for deduction, combines the state transition diagram to calculate the causal compatible state set and the cooperative recoverable state domain during the state interaction anomaly, and generates a state interaction tolerance window when the causal compatible state set deviates from the cooperative recoverable state domain. The degradation decision module constructs a state envelope of associated devices based on the state interaction tolerance window, uses the state envelope of associated devices to filter candidate degradation actions, obtains the target degradation action and executes it, and generates a local event record. The collaborative management module reconstructs a cross-device event causal partial order graph based on the local event records, generates candidate consistent causal sections, selects a target recovery section through the candidate consistent causal sections, and performs hierarchical resolution on the branching events after the target recovery section to restore the collaborative management task.

9. A computer device, comprising: A memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, it implements the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-7.

Images