Maintenance step simulation method based on model fusion

Abstract

The invention discloses a maintenance step simulation method based on model fusion, and relates to the technical field of virtual maintenance. According to the method, a high-fidelity multi-physics field simulation environment is constructed by deeply fusing a three-dimensional geometric model with physical attributes such as density and friction coefficient and assembly logic; an obstacle avoidance path is planned by using a fast expansion random tree algorithm, and real-time calculation of the deformation of the flexible pipeline and the statics equilibrium state of parts is realized in combination with a mass spring model and a polyhedral friction cone theory. On the basis, a multi-dimensional evaluation system containing space gaps, view shielding and operation torque weights is constructed, feasibility confidence indexes of overhaul steps are quantitatively evaluated, and a virtual guide potential field and directed acyclic graph logic verification are assisted, so that an overhaul guide file conforming to physical rules and process specifications is finally and automatically generated. The problems that traditional simulation lacks mechanical verification and flexible body simulation is distorted are effectively solved, and the reliability of maintenance process design is remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of virtual maintenance, and in particular to a method for simulating maintenance procedures based on model fusion. Background Technology

[0002] As industrial equipment becomes larger and more precise, its internal structure is becoming increasingly compact, significantly increasing the difficulty of inspection and maintenance. Traditional maintenance process design often relies on engineers' experience or two-dimensional drawings, lacking comprehensive consideration of physical factors such as the confinement of three-dimensional space, the weight of components, and contact friction. This can lead to disassembly paths being unenforceable in actual operation due to insufficient space, excessive torque, or interference collisions. Although existing virtual disassembly and assembly software can perform geometric collision detection, most treat components as rigid bodies, ignoring the deformation characteristics of flexible pipelines and complex mechanical equilibrium conditions. Furthermore, it lacks an automated verification mechanism for the disassembly sequence logic, resulting in significant deviations between simulation results and the real physical environment, making it difficult to generate highly reliable maintenance process guidance documents. Summary of the Invention

[0003] To address the aforementioned problems in existing technologies, the present invention aims to provide a maintenance procedure simulation method based on model fusion, the method comprising the following steps: Step S1: Obtain the three-dimensional geometric model and assembly process data of the equipment to be inspected, and map and fuse the three-dimensional geometric model with physical attribute information and motion constraint logic to construct a multi-physics field maintenance simulation environment.

[0004] Step S2: In the multiphysics maintenance simulation environment, the disassembly path of the component to be disassembled from the installation position to the safe area is planned based on the fast expanding random tree algorithm, and a discretized simulated maintenance action sequence is generated.

[0005] Step S3: Execute the simulated maintenance action sequence along the disassembly path, and detect the interference and mechanical balance between the component to be disassembled and the surrounding environment in real time.

[0006] Step S4: Extract the execution feature parameters of the simulated maintenance action sequence, and calculate the operational feasibility confidence index of the current maintenance step using a multidimensional weighted evaluation function.

[0007] Step S5: If the operational feasibility confidence index meets the preset pass threshold, the simulated maintenance action sequence is encapsulated into a standard maintenance viewpoint animation, and a maintenance process guidance document is generated.

[0008] Preferably, step S2 includes the following sub-steps: Step S201: Identify the bounding box boundary of the component to be disassembled, and set the degree-of-freedom constraint matrix of the component to be disassembled.

[0009] Step S202: Using the installation location as the starting node and the safe area as the target node, randomly sample and grow a path tree in the configuration space that avoids static obstacles.

[0010] Step S203: The successful paths in the path tree are smoothed using a spline interpolation algorithm to generate the disassembly path containing position coordinates and attitude quaternions.

[0011] Preferably, in step S4, the formula for calculating the operational feasibility confidence index is: ,in, This represents the confidence index of the feasibility of the operation; This represents the minimum critical gap value between the component to be disassembled and the surrounding obstacles; This indicates the preset standard reference gap value; This represents the visualization rating coefficient of the inspection viewpoint for the operating area, and its value ranges from zero to one. This indicates the maximum peak torque during the maintenance process; Indicates the maximum rated torque allowed by the robotic arm or operator; , , These are the spatial factor weight, the field of view factor weight, and the load factor weight, respectively.

[0012] Preferably, step S1 includes: The mesh data of the three-dimensional geometric model is parsed to extract vertex coordinates and face normal vectors; an attribute association table is established between the unique identifier of the component and the physical attribute information, including density, elastic modulus and friction coefficient; a kinematic pair constraint graph is constructed based on the assembly process data, the components are defined as graph nodes, and the revolute joints and prismatic joints are defined as weighted graph edges, thus completing the initialization of the multiphysics maintenance simulation environment.

[0013] Preferably, step S3 includes: Construct a first directed bounding box for the component to be disassembled and a second directed bounding box for the surrounding objects; construct a set of separation axes, which includes the basis vectors of the two directed bounding boxes and their cross product vector; for each separation axis in the set of separation axes... Determine whether the following separation inequality is satisfied: ,in, This represents the separation axis vector currently being detected; The distance vector between the geometric centers of the two directed bounding boxes is denoted as . ,in and These are the geometric center coordinates of the first directed bounding box and the second directed bounding box, respectively; and Each of the two oriented bounding boxes represents the two oriented bounding boxes at the separation axis. The projection radius on the surface; the projection radius is calculated as follows: , indicating that the directed bounding box is in the th order. Half-length dimension along the local axial direction; This indicates that the directed bounding box is in the th... Unit basis vectors along the local axis; Represents the dot product operation of vectors; This represents absolute value operations; if the separation inequality is satisfied on any separation axis, then it is determined that no interference has occurred.

[0014] Preferably, the mechanical equilibrium state detection in step S3 includes: All external forces acting on the components to be disassembled are mapped as six-dimensional generalized force spinors; the normal and tangential vectors of the contact point set of the components are extracted to construct a polyhedral friction cone; the grasping matrix of the contact point set is calculated; it is determined whether the six-dimensional generalized force spinor is located within the convex hull space spanned by the grasping matrix and the polyhedral friction cone. If it is located within the convex hull space, it is determined that the static equilibrium condition is satisfied.

[0015] Preferably, step S5 includes: The assembly process data is converted into a directed acyclic graph structure, where nodes represent parts and directed edges represent physical occlusion or installation dependencies; the in-degree value of the node corresponding to the part to be disassembled is calculated; only when the in-degree value is zero is the part confirmed to be in a disassembleable state without dependency constraints, otherwise it is marked as a logical conflict and the conflict node path is output.

[0016] Preferably, step S1 further includes: The flexible pipeline is discretized into a sequence of mass points connected by virtual springs. In step S3, the resultant external force of gravity, spring force, and environmental damping force on each mass point is calculated. The acceleration of the mass point is calculated based on the resultant external force. The spatial coordinates of the mass point are updated based on the position of the current time step and the position of the previous time step, combined with the acceleration. The distance between adjacent mass points is iteratively corrected to converge to a preset physical rod length to maintain the topological length constraint of the flexible pipeline.

[0017] Preferably, the method further includes: A virtual guiding potential field is established around the disassembly path; displacement input of the operation point is obtained through a human-computer interaction device; the displacement input is decomposed into a feed component along the tangential direction of the path and a deviation component along the normal direction of the path; an elastic restoring force based on the virtual guiding potential field is applied to the deviation component to generate tactile feedback data or a corrected motion trajectory to assist the user in making fine adjustments along the planned path.

[0018] Preferably, after step S5, the method further includes: Based on the component postures of adjacent time steps in the simulated maintenance action sequence, a continuously moving swept volume is constructed; a continuous collision detection algorithm is used to detect whether there is a mesh intersection between the swept volume and the static environment model; if there is a mesh intersection, the intersection depth is calculated and marked as a dynamic penetration failure; if there is no mesh intersection, the continuity safety of the maintenance path in the time dimension is verified.

[0019] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention constructs a multiphysics simulation environment that integrates geometry, physical properties, and assembly logic. It not only realizes rigid body collision detection based on the separating axis theorem, but also innovatively introduces a mass spring model to perform discretized dynamic calculations of flexible pipelines, enabling maintenance simulations to realistically reflect the impact of pipeline deformation on the disassembly space. Combined with static equilibrium detection based on six-dimensional generalized force spinor and polyhedral friction cone, the system can accurately predict whether components will become unstable and slip due to gravity or external forces during disassembly, effectively compensating for the lack of mechanical verification in pure geometric simulation.

[0020] This invention proposes a quantitative evaluation and intelligent assistance mechanism. By calculating the operational feasibility confidence index, which includes the weights of spatial clearance, field of view obstruction, and operating torque, it achieves a scientific score of the difficulty of maintenance steps. Combined with force tactile feedback technology based on virtual guiding potential field, the system can assist users in automatically correcting displacements that deviate from the normal direction when fine-tuning the path. Finally, it uses a directed acyclic graph to verify the logical dependency of the disassembly sequence, thereby outputting standard maintenance viewpoint animations and guidance documents that conform to both physical laws and process logic. Attached Figure Description

[0021] Figure 1 This is an exemplary flowchart of the maintenance procedure simulation method of the present invention.

[0022] Figure 2 An exemplary flowchart illustrating the steps involved in generating a maintenance action sequence according to the present invention. Implementation

[0023] The present invention will be further described below with reference to embodiments.

[0024] like Figure 1As shown, this invention provides a method for simulating maintenance procedures based on model fusion. The method includes the following steps: Step S1 involves acquiring the 3D geometric model and assembly process data of the equipment to be repaired, and mapping and fusing the 3D geometric model with physical attribute information and motion constraint logic to construct a multiphysics maintenance simulation environment. In practice, the 3D geometric model is typically parsed from industrial common exchange formats such as STEP AP242 or JT, containing precise boundary representation data. The assembly process data is parsed from PPR structured documents, defining the timing logic of disassembly operations.

[0025] Step S1 includes: The system analyzes the mesh data of the 3D geometric model to extract vertex coordinates and facet normals; it establishes an attribute association table between unique component identifiers and physical property information, including density, elastic modulus, and friction coefficient; and it constructs a kinematic pair constraint graph based on assembly process data, defining components as graph nodes and revolute and prismatic joints as weighted graph edges, thus initializing the multiphysics maintenance simulation environment. For example, for a flange connection structure fastened by bolts, the system identifies its kinematic pair type as a fixed pair and assigns a Coulomb friction coefficient to the contact surface; for a hydraulic cylinder assembly, it identifies it as a cylindrical pair and sets its axial stroke limit and rotation angle limit.

[0026] Step S1 also includes: The flexible pipeline is discretized into a sequence of mass points connected by virtual springs. In step S3, the resultant external force of gravity, spring force, and environmental damping force on each mass point is calculated. The acceleration of the mass point is calculated based on the resultant external force. The spatial coordinates of the mass point are updated based on the position of the current time step and the position of the previous time step, combined with the acceleration. The distance between adjacent mass points is iteratively corrected to converge to a preset physical rod length to maintain the topological length constraint of the flexible pipeline. This step essentially constructs a flexible body simulation model based on position dynamics, which can accurately simulate the bending stiffness and self-collision behavior of high-pressure oil pipes or wiring harnesses during disassembly.

[0027] Step S2: In a multiphysics maintenance simulation environment, the disassembly path of the component to be disassembled from its installation location to a safe area is planned based on the fast expanding random tree algorithm, and a discretized sequence of simulated maintenance actions is generated. The fast expanding random tree algorithm can effectively solve the path planning problem of high-dimensional multi-degree-of-freedom systems by randomly sampling in the configuration space and constructing a search tree.

[0028] like Figure 2 As shown, step S2 in this embodiment includes the following sub-steps: Step S201: Identify the bounding box boundary of the component to be disassembled and set the degree-of-freedom constraint matrix of the component to be disassembled; for example, the degree-of-freedom constraint matrix is ​​a six-dimensional Boolean vector, corresponding to translation along the three coordinate axes and rotation around the three coordinate axes.

[0029] Step S202: Using the installation location as the starting node and the safe area as the target node, randomly sample and grow a path tree in the configuration space that avoids static obstacles.

[0030] Step S203: The successful paths in the path tree are smoothed using a spline interpolation algorithm to generate disassembly paths containing position coordinates and attitude quaternions. The spline interpolation algorithm can use B-splines or non-uniform rational B-spline interpolation.

[0031] Step S3: Execute the simulated maintenance action sequence along the disassembly path, and detect the interference and mechanical balance between the parts to be disassembled and the surrounding environment in real time.

[0032] Step S3 includes: Construct the first directed bounding box of the component to be disassembled and the second directed bounding box of the environment; construct a set of separation axes, which contains the basis vectors of the two directed bounding boxes and their cross product vector; for each separation axis in the set of separation axes... Determine whether the following separation inequality is satisfied: ,in, This represents the separation axis vector currently being detected; This represents the distance vector between the geometric centers of two directed bounding boxes, i.e. ,in and These are the geometric center coordinates of the first and second directed bounding boxes, respectively. and These represent two oriented bounding boxes at the separation axis. The projection radius on the surface; the calculation method for the projection radius is as follows: , indicating that the directed bounding box is in the th order. Half-length dimension along the local axial direction; This indicates that the directed bounding box is in the th... Unit basis vectors along the local axis; Represents the dot product operation of vectors; This represents absolute value operations; if the separation inequality is satisfied on any separation axis, then it is determined that no interference has occurred. This algorithm, known as the separation axis theorem, is a classic algorithm in rigid body collision detection, capable of determining whether two convex polyhedra intersect in constant time complexity.

[0033] Step S4: Extract the execution feature parameters of the simulated maintenance action sequence, and use a multidimensional weighted evaluation function to calculate the operational feasibility confidence index of the current maintenance step.

[0034] In step S4, the formula for calculating the operational feasibility confidence index is: ,in, Indicates the confidence index of operational feasibility; This represents the minimum critical clearance value between the component to be disassembled and surrounding obstacles. This indicates a preset standard reference gap value, such as the hand operation space allowance set based on human body size data; This represents the visualization score coefficient of the inspection viewpoint for the operating area. Its value ranges from zero to one and can be obtained by calculating the proportion of occluded pixels using a ray casting algorithm. This indicates the maximum peak torque during the maintenance process; Indicates the maximum rated torque allowed by the robotic arm or operator; , , These are the weights for spatial factors, visual visibility factors, and workload factors. This formula integrates three dimensions—spatial accessibility, visual visibility, and operational workload—to comprehensively evaluate the ergonomic performance of maintenance operations.

[0035] Step S5: If the operational feasibility confidence index meets the preset pass threshold, the simulated maintenance action sequence is encapsulated into a standard maintenance viewpoint animation, and a maintenance process guidance document is generated.

[0036] Step S5 includes: The assembly process data is converted into a directed acyclic graph structure, where nodes represent parts and directed edges represent physical occlusion or installation dependencies. The in-degree value of the node corresponding to the part to be disassembled is calculated. Only when the in-degree value is zero is the part confirmed to be in a disassembleable state without dependency constraints; otherwise, it is marked as a logical conflict and the conflict node path is output.

[0037] The method in this embodiment also includes: A virtual guiding potential field is established around the disassembly path; the displacement input of the operation point is obtained through a human-computer interaction device; the displacement input is decomposed into a feed component along the tangential direction of the path and a deviation component along the normal direction of the path; an elastic restoring force based on the virtual guiding potential field is applied to the deviation component to generate tactile feedback data or a corrected motion trajectory to assist the user in making fine adjustments along the planned path.

[0038] Step S5 is followed by: Based on the component postures of adjacent time steps in the simulated maintenance action sequence, a continuously moving swept volume is constructed. A continuous collision detection algorithm is used to detect whether there is a mesh intersection between the swept volume and the static environment model. If there is a mesh intersection, the intersection depth is calculated and marked as a dynamic penetration failure. If there is no mesh intersection, the continuity safety of the maintenance path in the time dimension is verified.

[0039] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A method for simulating maintenance procedures based on model fusion, characterized in that, The method includes the following steps: Step S1: Obtain the three-dimensional geometric model and assembly process data of the equipment to be repaired, and map and fuse the three-dimensional geometric model with physical attribute information and motion constraint logic to construct a multi-physics field maintenance simulation environment; Step S2: In the multiphysics maintenance simulation environment, the disassembly path of the component to be disassembled from the installation position to the safe area is planned based on the fast expanding random tree algorithm, and a discretized simulated maintenance action sequence is generated. Step S3: Execute the simulated maintenance action sequence along the disassembly path, and detect the interference and mechanical balance between the component to be disassembled and the surrounding environment in real time. Step S4: Extract the execution feature parameters of the simulated maintenance action sequence, and calculate the operational feasibility confidence index of the current maintenance step using a multidimensional weighted evaluation function; Step S5: If the operational feasibility confidence index meets the preset pass threshold, the simulated maintenance action sequence is encapsulated into a standard maintenance viewpoint animation, and a maintenance process guidance document is generated.

2. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, Step S2 includes the following sub-steps: Step S201: Identify the bounding box boundary of the component to be disassembled, and set the degree-of-freedom constraint matrix of the component to be disassembled; Step S202: Using the installation location as the starting node and the safe area as the target node, randomly sample and grow a path tree in the configuration space that avoids static obstacles; Step S203: The successful paths in the path tree are smoothed using a spline interpolation algorithm to generate the disassembly path containing position coordinates and attitude quaternions.

3. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, In step S4, the formula for calculating the operational feasibility confidence index is as follows: ,in, This represents the confidence index of the feasibility of the operation; This represents the minimum critical gap value between the component to be disassembled and the surrounding obstacles; This indicates the preset standard reference gap value; This represents the visualization rating coefficient of the inspection viewpoint for the operating area, and its value ranges from zero to one. This indicates the maximum peak torque during the maintenance process; Indicates the maximum rated torque allowed by the robotic arm or operator; , , These are the spatial factor weight, the field of view factor weight, and the load factor weight, respectively.

4. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, Step S1 includes: The mesh data of the three-dimensional geometric model is parsed to extract vertex coordinates and face normal vectors; an attribute association table is established between the unique identifier of the component and the physical attribute information, including density, elastic modulus and friction coefficient; a kinematic pair constraint graph is constructed based on the assembly process data, the components are defined as graph nodes, and the revolute joints and prismatic joints are defined as weighted graph edges, thus completing the initialization of the multiphysics maintenance simulation environment.

5. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, Step S3 includes: Construct a first directed bounding box for the component to be disassembled and a second directed bounding box for the surrounding objects; construct a set of separation axes, which includes the basis vectors of the two directed bounding boxes and their cross product vector; for each separation axis in the set of separation axes... Determine whether the following separation inequality is satisfied: ,in, This represents the separation axis vector currently being detected; The distance vector between the geometric centers of the two directed bounding boxes is denoted as . ,in and These are the geometric center coordinates of the first directed bounding box and the second directed bounding box, respectively; and Each of the two oriented bounding boxes represents the two oriented bounding boxes at the separation axis. The projection radius on the surface; the projection radius is calculated as follows: , indicating that the directed bounding box is in the th order. Half-length dimension along the local axial direction; This indicates that the directed bounding box is in the th... Unit basis vectors along the local axis; Represents the dot product operation of vectors; This represents absolute value operations; if the separation inequality is satisfied on any separation axis, then it is determined that no interference has occurred.

6. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, The mechanical equilibrium state detection in step S3 includes: All external forces acting on the components to be disassembled are mapped as six-dimensional generalized force spinors; the normal and tangential vectors of the contact point set of the components are extracted to construct a polyhedral friction cone; the grasping matrix of the contact point set is calculated; it is determined whether the six-dimensional generalized force spinor is located within the convex hull space spanned by the grasping matrix and the polyhedral friction cone. If it is located within the convex hull space, it is determined that the static equilibrium condition is satisfied.

7. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, Step S5 includes: The assembly process data is converted into a directed acyclic graph structure, where nodes represent parts and directed edges represent physical occlusion or installation dependencies; the in-degree value of the node corresponding to the part to be disassembled is calculated; only when the in-degree value is zero is the part confirmed to be in a disassembleable state without dependency constraints, otherwise it is marked as a logical conflict and the conflict node path is output.

8. The method for simulating maintenance steps based on model fusion according to claim 1, characterized in that, Step S1 further includes: The flexible pipeline is discretized into a sequence of mass points connected by virtual springs. In step S3, the resultant external force of gravity, spring force, and environmental damping force on each mass point is calculated. The acceleration of the mass point is calculated based on the resultant external force. The spatial coordinates of the mass point are updated based on the position of the current time step and the position of the previous time step, combined with the acceleration. The distance between adjacent mass points is iteratively corrected to converge to a preset physical rod length to maintain the topological length constraint of the flexible pipeline.

9. The maintenance procedure simulation method based on model fusion according to claim 1, characterized in that, The method further includes: A virtual guiding potential field is established around the disassembly path; displacement input of the operation point is obtained through a human-computer interaction device; the displacement input is decomposed into a feed component along the tangential direction of the path and a deviation component along the normal direction of the path; an elastic restoring force based on the virtual guiding potential field is applied to the deviation component to generate tactile feedback data or a corrected motion trajectory to assist the user in making fine adjustments along the planned path.

10. The method for simulating maintenance steps based on model fusion according to claim 1, characterized in that, Following step S5, the following is also included: Based on the component postures of adjacent time steps in the simulated maintenance action sequence, a continuously moving swept volume is constructed; a continuous collision detection algorithm is used to detect whether there is a mesh intersection between the swept volume and the static environment model; if there is a mesh intersection, the intersection depth is calculated and marked as a dynamic penetration failure; if there is no mesh intersection, the continuity safety of the maintenance path in the time dimension is verified.

Images