Ship outfitting path planning and simulation modeling method, system, equipment and medium

Abstract

The invention discloses a ship outfitting path planning and simulation modeling method, system and device and a medium, and belongs to the technical field of ship digital assembly. According to the method, the problems of low path planning efficiency, insufficient virtual-real fusion precision and lack of dynamic collision early warning in a complex cabin environment are solved, the cognitive load of workers is remarkably reduced, the assembly qualification rate and the collision early warning accuracy are improved, and the method is suitable for digital outfitting operation of complex system equipment such as ship pipelines and shaft-rudder systems and has a wide application prospect. And a key technical support is provided for digital shipbuilding.

Description

Technical Field

[0001] This application relates to the field of ship digital assembly, and in particular to a method, system, device and medium for ship outfitting path planning and simulation modeling. Background Art

[0002] In recent years, with the development trend of ship enlargement and lightweighting, the hull size and weight have gradually increased, the stiffness has relatively weakened, and the demand for increasingly refined ship design has gradually increased the requirements for ship system equipment outfitting. The outfitting operation of ship system equipment accounts for more than 45% of the entire shipbuilding cycle and plays an important role in the whole ship manufacturing process. At present, during the outfitting operation of ship system equipment, there are often problems such as low path planning efficiency in narrow channels, insufficient virtual-real fusion accuracy due to relying on manual calibration and process information, and lack of a dynamic collision warning and identification mechanism under multiple constraints. Therefore, unreasonable outfitting operations often lead to irreversible large structural deformations and safety accidents, greatly affecting the improvement of shipbuilding efficiency. Summary of the Invention

[0003] The purpose of this application is to provide a method, system, device and medium for ship outfitting path planning and simulation modeling, which solves the problem that it is difficult to realize the outfitting path planning and dynamic simulation of ship system equipment in complex cabin environments in the prior art. By guiding the assembly through the combination of virtual and real, it significantly reduces the cognitive burden of workers on process information, thereby improving the assembly efficiency and quality of system equipment and reducing the error rate.

[0004] The solution adopted in this application is to provide a method for ship outfitting path planning and simulation modeling, including the steps of: Collecting the characteristic parameters of the on-site environment for ship system equipment outfitting, and constructing a virtual-real consistent three-dimensional virtual environment through surface fitting and data alignment; Based on multi-constraints of obstacles, narrow channels and dynamic environments, using the narrow channel optimization algorithm (NARRT*), dynamically planning a six-degree-of-freedom path based on the convex hull pose distance metric model, and generating the optimal outfitting path of ship system equipment through hierarchical bounding box collision detection and B-spline path smoothing. According to the optimal outfitting path, combining with a three-dimensional visualization engine to realize the dynamic simulation and collision warning of the outfitting process; Using industrial cameras, laser trackers and optical targets to collect outfitting process data, establishing a mapping relationship between feature deviations and the three-dimensional geometric model of the ship system equipment assembly body, predicting the deviation trend through a long short-term memory neural network and generating a correction instruction to correct the optimal outfitting path.

[0005] Further, the step of collecting the characteristic parameters of the on-site environment for ship system equipment outfitting and constructing a virtual-real consistent three-dimensional virtual environment through surface fitting and data alignment includes: Collect on-site feature data through a 3D laser scanner, an industrial camera, and a depth sensor. The feature data includes at least one of chimney segmentation, side passage, seawater main pipe unit, bulk carrier shaft and rudder system, tanker tube bundle unit system, auxiliary adjustment position, operation space layout, and equipment base. Extract point and surface feature information based on surface fitting and data alignment methods. Based on the collected feature data and the extracted point and surface feature information, construct an information model of the product assembly, an information model of the assembly process, and an information model of the assembly path simulation through 3D scene rendering. Form a virtual scene of the outfitting site with the information models of the product assembly, the assembly process, and the assembly path simulation. Define multiple coordinate systems, and correct the pose data of the ship system equipment in real time through marker recognition, correct the parameters of the virtual scene of the outfitting site, and reconstruct a 3D virtual environment consistent with the outfitting site.

[0006] Furthermore, the constructing of the information model of the product assembly, the information model of the assembly process, and the information model of the assembly path simulation through 3D scene rendering based on the collected feature data and the extracted point and surface feature information includes: Extract and integrate the geometric information, hierarchical structure information, basic attribute information, and pose information of the components from the collected feature data, integrate the information using an XML format file, create a component node file, and form an information model of the product assembly in the augmented reality scenario. Extract and integrate the task types, assembly processes and steps, operation objects, tools and resources, operation methods, assembly paths, and preconditions during the assembly process from the collected feature data, create an XML file for the steps, refine the process steps, and form an information model of the assembly process. Use a planning algorithm to plan and obtain the assembly object, the assembly base environment, the assembly movement path, the dynamic constraints, and the interference requirements as the component movement attitude information. Describe the movement details of the component assembly process according to the component movement attitude information, create an XML file for the component assembly path, and form an information model of the assembly path simulation.

[0007] Furthermore, the generating of the optimal outfitting path of the ship system equipment by using the narrow passage optimization algorithm (NARRT*), dynamically programming the six-degree-of-freedom path based on the convex hull pose distance metric model, and performing hierarchical bounding box collision detection and B-spline path smoothing processing includes: Map the ship outfitting workspace to a six-degree-of-freedom space , where the translational degrees of freedom are , and the rotational degrees of freedom are , which are the x-axis direction, the y-axis direction, and the z-axis direction respectively, They are the rotational degrees of freedom in the x-axis direction, the rotational degrees of freedom in the y-axis direction, and the rotational degrees of freedom in the z-axis direction respectively; Define the free space , the obstacle space and the narrow area ; Calculate the local obstacle density based on the hypersphere sampling method , and dynamically adjust the sampling strategy based on the local obstacle density ; when is greater than or equal to the first threshold, it is determined as a narrow channel, and directional bias sampling is performed along the main axis direction of the narrow channel; when is less than the first threshold, it is determined as an open area, and uniform random sampling is used within the open area; Calculate the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning; generate a smooth B-spline path using two-way tree synchronous expansion and hierarchical bounding box collision detection, and output a sequence of path key points including pose sequences and time parameters.

[0008] Furthermore, the calculation of the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning includes: By obtaining the three-dimensional geometric model of the ship system equipment assembly, calculate its convex hull vertex set ; where is the rigid body transformation of the nth pose, ; Define and two poses, and perform rigid body transformation on the convex hull vertex set , where the rigid body transformation matrix includes a translation component and a rotation component ; are the translation components in the x-axis direction, the y-axis direction, and the z-axis direction respectively, are the rotation components in the x-axis direction, the y-axis direction, and the z-axis direction respectively; Calculate the maximum displacement distance and the average displacement distance of the corresponding vertices , and use the maximum displacement distance as the metric distance between poses and .

[0009] Furthermore, the realization of dynamic simulation and collision warning of the outfitting process according to the optimal outfitting path in combination with a three-dimensional visualization engine includes: Based on a 3D visualization engine, load the 3D geometric model of the ship system equipment assembly and the planned path data, and render the motion trajectory and key pose points of the assembly in real time. When the detected minimum distance is less than the safety threshold, trigger a collision warning; Highlight the risk area through the 3D visualization engine, and automatically generate path correction suggestions to update the optimal outfitting path.

[0010] Furthermore, the utilization of industrial cameras, laser trackers, and optical targets to collect outfitting process data, establish the mapping relationship between feature deviations and the 3D geometric model of the ship system equipment assembly, and predict the deviation trend through a long short-term memory neural network and generate correction instructions to correct the optimal outfitting path includes: Arrange industrial cameras, laser trackers, and optical targets to form a multi-sensor network, set up multi-view observation stations, and plan the spatial distribution of the target point array; Calculate the 3D coordinates of feature points through bundle adjustment, and establish the conversion relationship between the global coordinate system and the local coordinate system; Extract the measured pose data of key control points, calculate the deviation value from the theoretical model, and establish the mapping relationship between the deviation field and the 3D geometric model of the ship system equipment assembly; Use radial basis function interpolation to display the deviation distribution heat map in real time, and predict the deviation development trend based on a long short-term memory neural network and generate correction instructions, and correct the optimal outfitting path according to the correction instructions.

[0011] This application also provides a ship outfitting path planning and simulation modeling system, including: A data acquisition and virtual modeling module, used to collect the characteristic parameters of the ship system equipment outfitting site environment, and construct a virtual 3D environment that is consistent between the virtual and the real through surface fitting and data alignment; A path planning and simulation module, used to dynamically plan a six-degree-of-freedom path based on multiple constraints of obstacles, narrow channels, and dynamic environments, using the narrow channel optimization algorithm (NARRT*), based on the convex hull pose distance metric model, and generate the optimal outfitting path of the ship system equipment through hierarchical bounding box collision detection and B-spline path smoothing processing, and realize the dynamic simulation and collision warning of the outfitting process according to the optimal outfitting path combined with the 3D visualization engine; A deviation detection and correction module, used to utilize industrial cameras, laser trackers, and optical targets to collect outfitting process data, establish the mapping relationship between feature deviations and the 3D geometric model of the ship system equipment assembly, and predict the deviation trend through a long short-term memory neural network and generate correction instructions to correct the optimal outfitting path.

[0012] Furthermore, the path planning and simulation module includes: A narrow channel determination unit, used to calculate the local obstacle density based on the hypersphere sampling method , dynamically adjust the sampling strategy based on the local obstacle density ; A convex hull pose distance calculation unit, which is used to calculate the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set, and use it as the distance metric for path planning; A path smoothing unit, which is used to generate a smooth B-spline path by means of bidirectional tree synchronous expansion and hierarchical bounding box collision detection, and output a sequence of path key points including pose sequences and time parameters.

[0013] Furthermore, the deviation detection and correction module includes: A multi-sensor calibration unit, which is used to arrange industrial cameras, laser trackers and optical targets to form a multi-sensor network, set multi-view observation stations, and plan the spatial distribution of the target point array; A deviation heat map generation unit, which is used to display the deviation distribution heat map in real time through radial basis function interpolation; A warning level classification unit, which is used to predict the deviation development trend based on long short-term memory and output a warning signal including normal, attention, and danger levels.

[0014] This application also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method described above are implemented.

[0015] This application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described above are implemented.

[0016] This application solves the problems of low path planning efficiency, insufficient virtual-real fusion accuracy, and lack of dynamic collision warning in complex cabin environments, significantly reduces the cognitive load of workers, improves the qualified rate of assembly and the accuracy of collision warning, and is applicable to digital outfitting operations of complex system equipment such as ship pipelines and shaft rudders, providing key technical support for digital shipbuilding. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is the logic diagram for constructing the virtual scene of the outfitting site in this application; Figure 2 is the flowchart of the NARRT* algorithm in this application; Figure 3 is the multi-sensor network and deviation tracking diagram in this application; Figure 4 is the schematic diagram of the ship outfitting path planning and simulation modeling method in this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following clearly and completely describes the technical solution of the present application in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0019] In this embodiment, taking the outfitting operation of the chimney section in a large container ship construction project as an example, the specific implementation manner of the present application is elaborated in detail.

[0020] Embodiment 1 This embodiment provides a ship outfitting path planning and simulation modeling method based on multi-constraint virtual-real fusion, which is implemented according to the following steps: Step 1: Collect the characteristic parameters of the outfitting site environment of the ship system equipment, extract the point-surface feature information through surface fitting and data alignment methods, correct the virtual environment parameter model of the outfitting site, and construct a three-dimensional virtual outfitting environment based on key features consistent with the outfitting site.

[0021] Step 2: Combine multi-constraint conditions such as obstacles, narrow channels, and dynamic environments, and based on the narrow channel optimization algorithm (NARRT*), generate the optimal outfitting sequence and path of the ship system equipment. Verify the outfitting process through three-dimensional visual simulation, and conduct visual analysis and early warning processing on potential collision problems that may occur during the outfitting process.

[0022] Among them, based on the multi-constraint conditions of obstacles, narrow channels, and dynamic environments, the narrow channel optimization algorithm (NARRT*) is used to generate the optimal outfitting path sequence of the ship system equipment. Through hierarchical bounding box collision detection and B-spline path smoothing processing, and combined with a three-dimensional visualization engine, dynamic simulation and collision warning of the outfitting process are realized.

[0023] Step 3: Use industrial cameras, laser trackers, optical targets, etc. to collect outfitting process data, form a set of target characteristic parameters for the outfitting of ship system equipment, reconstruct the three-dimensional model of the ship system equipment, establish a mapping relationship between the feature measurement deviation and the three-dimensional system model, and predict the deviation trend through a long short-term memory neural network to realize the visual tracking and intelligent prediction and correction of the outfitting deviation information of the system equipment.

[0024] Such as Figure 1As shown in the figure, step 1 is as follows: first, the feature data of the outfitting site is collected by using a 3D laser scanner, an industrial camera and a depth sensor, and the point and surface feature information is extracted by combining surface fitting and data alignment methods; secondly, based on the collected data, a virtual scene of the outfitting site is constructed by scene 3D rendering, including a product assembly information model, an assembly process information model and an assembly path simulation information model; finally, a hybrid tracking technology based on natural features and multi-sensor fusion is used to define multiple coordinate systems and describe the position and posture of ship system equipment, and the posture data is corrected in real time through marker recognition to improve the model accuracy, correct the virtual environment parameters of the ship outfitting site, and reconstruct a 3D virtual environment consistent with the outfitting site.

[0025] Furthermore, the on-site characteristic data include container ship chimney sections, side channels, seawater main units, bulk carrier shaft and rudder systems, oil tanker tube bundle unit systems, etc., as well as key features such as auxiliary adjustment positions, work space layout, and equipment bases.

[0026] Furthermore, the product assembly information model is developed through secondary development of CAD software, extracting and integrating information such as component geometry and hierarchical structure, integrating information using XML format files, creating component node files, and forming a product assembly information model in an augmented reality scenario; the model information includes component geometry information, hierarchical structure information, basic attribute information, and posture information.

[0027] Optionally, the component geometry information includes the three-dimensional geometry, surface model and wireframe model of each component of the ship system equipment; Optionally, the hierarchical structure information includes the parent-child hierarchical relationship between the components of the ship system equipment and the assembly constraints between the components; Optionally, the basic attribute information includes the name, number, material properties (density, elastic modulus, thermal expansion coefficient, etc.), mass and center of gravity, outfitting process requirements, etc. of each component of the ship system equipment; Optionally, the pose information includes the initial pose, target pose, relative pose, absolute pose and bounding box center pose of each component of the ship system equipment.

[0028] Furthermore, the assembly process information model is developed through secondary development of CAPP software, extracts and integrates key information such as task types and processes in the assembly process, creates XML files for the work steps, refines the specific process, and forms an AR-guided assembly process information model; the model information includes assembly processes and steps, operation objects, tools and resources, operation methods, assembly paths and prerequisites.

[0029] Optionally, the processes and steps include decomposition of assembly tasks such as mounting bases for ship system equipment and positioning bolt holes; Optionally, the objects of operation include part numbers, accessories, etc.; Optionally, the tools and resources include wrench models, crane load capacity, man - power allocation, etc.; Optionally, the operation methods include descriptions of assembly actions and precautions such as collision - prevention sensors; Optionally, the assembly path and pre - conditions include motion trajectory planning files and previous working steps of various components.

[0030] Furthermore, the assembly path simulation information model is obtained by secondary development of CAD software. Using planning algorithms, it automatically plans to obtain the motion posture information of components, describes the motion details during the component assembly process, creates an XML file for the component assembly path, and forms an assembly path simulation information model; the model information includes assembly objects, assembly matrix environment, assembly motion path, dynamic constraints, and interference requirements.

[0031] Optionally, the assembly matrix environment includes a list of assembled components, obstacles, workspace constraints, etc.; Optionally, the assembly motion path includes initial pose, target pose, sequence of key points on the motion path, motion duration, motion speed, etc.; Optionally, the dynamic constraints include degree - of - freedom limitations, equipment capabilities, etc.; Optionally, the interference requirements include collision - detection rules, safety warning thresholds, etc.

[0032] As Figure 2 shown, the NARRT* algorithm in step 2 includes: Mapping the ship outfitting workspace to a six - degree - of - freedom configuration space , defining the free space , obstacle space and narrow areas ; Calculating the local obstacle density based on the hypersphere sampling method , when it is determined as a narrow channel, and directional offset sampling is performed along the main axis direction of the channel; Calculating the maximum displacement distance between poses based on the rigid - body transformation of the convex hull as the distance metric for path planning; Generating a smooth B - spline path using bidirectional tree synchronous expansion and hierarchical bounding - box collision detection, and outputting an XML file containing the pose sequence and time parameters.

[0033] ​Furthermore, Step 2 is specifically as follows: First, construct a three-dimensional configuration space model for ship outfitting, and define the narrow passage area in its workspace; Second, based on the pose distance metric model of the convex hull, calculate the pose distance of the assembly in the configuration space; Then, use the NARRT* algorithm to search for paths in the configuration space, dynamically adjust the sampling strategy according to the regional environment type, and verify the path feasibility through the hierarchical bounding box collision algorithm, so as to generate an optimal outfitting path sequence and perform path smoothing; Finally, based on the three-dimensional visualization engine, realize the dynamic simulation of the outfitting process, monitor potential collision risks in real time and generate warning information.

[0034] Furthermore, the configuration space model maps the ship outfitting workspace into a six-degree-of-freedom space, including translational degrees of freedom and rotational degrees of freedom , and at the same time define the free space as , the obstacle space and the narrow area ; The narrow passage area is judged by calculating the local area obstacle density through the hypersphere sampling method .

[0035] Furthermore, the pose distance metric model based on the convex hull includes calculating the convex hull vertex set by obtaining the three-dimensional geometric model of the assembly of the ship system equipment to be measured ; Define and two poses, and perform a rigid body transformation on the convex hull vertex set , where the rigid body transformation matrix includes a translation component and a rotation component ; Calculate the maximum displacement distance and average displacement distance of the corresponding vertices , and use the maximum displacement distance as the metric distance between the poses and .

[0036] Furthermore, the NARRT* algorithm improves the classical RRT algorithm and adopts a sensitive adaptive sampling strategy in the ship outfitting workspace, where uniform random sampling is used in the open area , and directional bias sampling is performed along the main axis of the channel in the narrow area . Based on the convex hull model, accurately measure the pose distance and combine it with the dynamic environment classification mechanism. Through the two-way synchronous expansion of the initial tree and the target tree, use the hierarchical bounding box for real-time collision monitoring to complete dynamic path optimization, generate a smooth B-spline path curve, and output an XML file containing the pose sequence and time parameters.

[0037] Furthermore, for the dynamic simulation of the outfitting process, based on a three-dimensional visualization engine, the three-dimensional geometric model of the ship system equipment assembly and the planned path data are loaded, and the movement trajectory and key pose points of the assembly are rendered in real time. When it is detected that the minimum distance is less than the safety threshold, a collision warning is triggered and the risk area is highlighted, and path correction suggestions are automatically generated and the path planning scheme is updated.

[0038] That is, the dynamic simulation includes: Rendering the movement trajectory and key pose points of the assembly in real time, and triggering a collision warning when it is detected that the minimum distance is less than the safety threshold; Highlighting the risk area through a three-dimensional visualization engine and automatically generating path correction suggestions.

[0039] As Figure 3 shown, step 3 is specifically as follows: Arrange industrial cameras, laser trackers, and optical targets to form a multi-sensor network, design multi-view observation stations, and plan the spatial distribution of the target point array; Calculate the three-dimensional coordinates of feature points through bundle adjustment, and establish the conversion relationship between the global coordinate system and the local coordinate system; Extract the measured pose data of key control points and calculate the deviation value from the theoretical model; Correct the integrated information model based on the measured data and establish the mapping relationship between the deviation field and the three-dimensional model; Real-time display the deviation heat map through a visualization method and predict the deviation development trend based on a machine learning algorithm.

[0040] Furthermore, for the industrial cameras, optical targets, and laser trackers, the layout spacing of industrial cameras does not exceed 5 meters, the optical targets are made of high-reflectivity materials, and the measurement accuracy of the laser tracker reaches 0.01 mm / m.

[0041] Furthermore, for the mapping between the deviation field and the three-dimensional model, the deviation field mapping uses radial basis function interpolation, and the matching error between the reconstructed model and the theoretical model is less than 0.5 mm; that is, radial basis function interpolation is used to construct the deviation field mapping, and the matching error is less than 0.5 mm.

[0042] Furthermore, for the deviation prediction, a long short-term memory neural network is used for deviation prediction, and the prediction time span can be set from 1 to 24 hours, that is, the deviation trend is predicted based on the long short-term memory neural network for 1 to 24 hours, and a warning signal including normal, attention, and danger levels is output.

[0043] Advantages of the present application: The present application realizes efficient path planning in a narrow-channel environment through the NARRT* algorithm, improves the planning accuracy by using the convex hull pose distance metric, constructs a deviation prediction model by combining multi-sensor data, and significantly improves the assembly efficiency and quality. This method is compatible with mainstream CAD / CAPP systems, realizes the dynamic simulation of the outfitting process through a 3D visualization engine, supports six-degree-of-freedom motion simulation, AR / VR interaction, realizes "what you see is what you get" assembly guidance, establishes a complete "planning - simulation - detection - prediction" technical route, creates a full-process digital closed-loop from virtual planning to actual assembly, greatly improves the success rate of path planning, the accuracy of collision detection and warning, and the deviation detection accuracy, reduces the assembly error rate and the cognitive burden of workers, and is applicable to the outfitting operations of complex system equipment such as pipelines and chimneys in shipbuilding, providing key technical support for digital shipbuilding.

[0044] Embodiment 2 Embodiment 2 includes all the technical features of Embodiment 1. In Embodiment 2, a three-dimensional virtual environment for ship outfitting is constructed, and the specific implementation includes the following steps: (1) Data collection: Use a three-dimensional laser scanner (scanning accuracy ±2 mm), an industrial camera (resolution 2592×1944 pixels), and a depth sensor to collect comprehensive data on the outfitting site of the chimney section of a container ship. The collection content includes the structure of the chimney section itself, the surrounding side channels, installed equipment such as the seawater main pipe unit, as well as key feature data such as the auxiliary position adjustment positions, the layout of the working space, and the equipment bases.

[0045] (2) Feature extraction and model construction: Import the collected data into MATLAB software, process the complex surface through a surface fitting algorithm, and combine the data alignment method based on ICP (Iterative Closest Point) to extract point and surface feature information. Based on the extracted features, perform secondary development on AutoCAD software to extract information such as the geometric shape (3D geometric shape, patch model, wireframe model) and hierarchical structure of the chimney section components, integrate them in an XML format file, create a component node file, and form a product assembly information model; perform secondary development on CAPP software to extract and integrate key information such as the task types and processes (such as the installation of the base and the decomposition of the positioning bolt hole assembly task) during the assembly process, create an XML file for the work steps, refine the process, and construct an assembly process information model; use AutoCAD secondary development again, apply the Dijkstra planning algorithm to automatically plan and obtain the motion pose information of the components, describe the motion details during the assembly process, create an XML file for the component assembly path, and form an assembly path simulation information model.

[0046] (3)Model correction: Adopt a hybrid tracking technology based on natural features and multi-sensor fusion, and define multiple coordinate systems such as the world coordinate system, device coordinate system, and sensor coordinate system. AR markers are arranged in the segmented chimney and surrounding areas, and the markers are recognized by industrial cameras to correct the pose data in real time. The constructed virtual scene is corrected using the collected real-time data to ensure that the virtual environment parameters are consistent with the actual outfitting site, and the three-dimensional virtual environment construction is completed.

[0047] Among them, generating the optimal outfitting sequence and path specifically includes the following steps when implemented: (1)Configuration of spatial model construction: Map the outfitting workspace of the segmented chimney into a six-degree-of-freedom space, where the translational degrees of freedom are , and the rotational degrees of freedom are , and the free space , obstacle space and narrow area are defined. The hypersphere sampling method is adopted, with the segmented chimney as the center, sampling within a radius of 5 meters, and calculating the obstacle density in the local area . After calculation, some areas around the segmented chimney are determined as narrow channel areas.

[0048] (2)Pose distance calculation: Obtain the three-dimensional geometric model of the segmented chimney assembly, and calculate the convex hull vertex set using the convex hull generation function of SolidWorks software . Define and two poses, and perform a rigid body transformation on the convex hull vertex set, where the rigid body transformation matrix includes a translation component and a rotation component . Calculate the maximum displacement distance and the average displacement distance of the corresponding vertices, and use as the metric distance between poses and .

[0049] (3)Path search and optimization: Use the NARRT* algorithm for path search. In the open area adopt a uniform random sampling strategy, and in the narrow area perform directional offset sampling along the main axis of the channel. Expand the initial tree and the target tree synchronously in both directions, and use a hierarchical bounding box (OBB bounding box) for real-time collision monitoring. After multiple iterations of optimization, a smooth B-spline path curve is generated, and an XML file containing the pose sequence and time parameters is output.

[0050] (4) Dynamic simulation and warning: Based on the Unity 3D visualization engine, load the 3D model of the chimney segmented assembly and the planned path data, and render the movement trajectory and key pose points of the assembly in real time. Set the safety threshold to 200 mm. When the detected minimum distance is less than the safety threshold, trigger a collision warning, highlight the risk area in red, automatically generate path correction suggestions and update the path planning scheme.

[0051] Among them, for deviation tracking and prediction correction, the specific implementation includes the following steps: (1) Sensor network layout: Around the chimney segmented outfitting area, install an industrial camera every no more than 5 meters, with a total of 8 cameras installed; install a laser tracker (measurement accuracy 0.01 mm / m); use optical targets made of high-reflectivity materials and paste 20 optical targets at key parts of the chimney segments to form a multi-sensor network. Design 4 multi-view observation stations to ensure that all parts of the chimney segments can be comprehensively observed.

[0052] (2) Data processing and deviation calculation: Calculate the three-dimensional coordinates of feature points through bundle adjustment, and establish the conversion relationship between the global coordinate system and the local coordinate system. Use the laser tracker and industrial cameras to collect the measured pose data of key control points, compare with the theoretical model, and calculate the deviation value. For example, at a certain key control point, the measured coordinate is 12.345 m, and the theoretical coordinate is 12.350 m, then the direction deviation value is -0.005 m.

[0053] (3) Model correction and prediction: Correct the integrated information model based on the measured data, establish the mapping relationship between the deviation field and the 3D model using radial basis function interpolation, and ensure that the matching error between the reconstructed model and the theoretical model is less than 0.5 mm. Use Python to build a long short-term memory neural network model, input the deviation data of the past 6 hours, set the prediction time span to 12 hours, and output the deviation warning level (normal, attention, danger). When it is predicted that the deviation of a certain part will exceed the allowable range, promptly issue a warning to the construction personnel to guide them to make adjustments to ensure the high-precision completion of the chimney segmented outfitting operation.

[0054] Embodiment 3 Embodiment 3 includes all the technical features of Embodiment 2. As Figure 4 shown, in Embodiment 3, a ship outfitting path planning and simulation modeling method is provided, including the steps: Collect the outfitting site environment characteristic parameters of the ship system equipment, and construct a virtual-real consistent 3D virtual environment through surface fitting and data alignment; Based on multiple constraints such as obstacles, narrow channels, and dynamic environments, the narrow-channel optimization algorithm (NARRT*) is adopted to dynamically plan a six-degree-of-freedom path based on the convex hull pose distance metric model. Through hierarchical bounding box collision detection and B-spline path smoothing, the optimal outfitting path of the ship system equipment is generated. According to the optimal outfitting path, combined with a three-dimensional visualization engine, dynamic simulation and collision warning of the outfitting process are realized; Industrial cameras, laser trackers, and optical targets are used to collect outfitting process data, establish the mapping relationship between feature deviations and the three-dimensional geometric model of the ship system equipment assembly, and predict the deviation trend through a long short-term memory neural network and generate correction instructions to correct the optimal outfitting path.

[0055] Furthermore, the characteristic parameters of the outfitting site environment of the ship system equipment are collected, and a virtual-reality-consistent three-dimensional virtual environment is constructed through surface fitting and data alignment, including: Field characteristic data are collected through a three-dimensional laser scanner, industrial cameras, and depth sensors. The characteristic data include at least one of the chimney section, side passage, seawater main pipe unit, bulk carrier shaft and rudder system, tanker pipe bundle unit system, auxiliary adjustment position, operation space layout, and equipment base; Point and surface feature information is extracted based on surface fitting and data alignment methods; Based on the collected characteristic data and the extracted point and surface feature information, an information model of the product assembly, an information model of the assembly process, and an information model of the assembly path simulation are constructed through scene three-dimensional rendering. The information model of the product assembly, the information model of the assembly process, and the information model of the assembly path simulation form a virtual scene of the outfitting site; Multiple coordinate systems are defined, and the pose data of the ship system equipment are corrected in real time through marker recognition, the parameters of the virtual outfitting site scene are corrected, and a three-dimensional virtual environment consistent with the outfitting site is reconstructed.

[0056] Furthermore, the construction of the information model of the product assembly, the information model of the assembly process, and the information model of the assembly path simulation through scene three-dimensional rendering based on the collected characteristic data and the extracted point and surface feature information includes: Extract and integrate the geometric information, hierarchical structure information, basic attribute information, and pose information of the components from the collected characteristic data, integrate the information using XML format files, create component node files, and form an information model of the product assembly under the augmented reality scenario; Extract and integrate the task types, assembly processes and steps, operation objects, tools and resources, operation methods, assembly paths, and preconditions during the assembly process from the collected characteristic data, create XML files for the steps, refine the process steps, and form an information model of the assembly process; Use a planning algorithm to plan and obtain the assembly object, the assembly matrix environment, the assembly motion path, the dynamic constraints, and the interference requirements as the motion attitude information of the components. Describe the motion details of the component assembly process according to the motion attitude information of the components, create an XML file for the component assembly path, and form an assembly path simulation information model.

[0057] Further, the narrow passage optimization algorithm (NARRT*) is adopted to dynamically plan a six-degree-of-freedom path based on the convex hull pose distance metric model. The optimal outfitting path of the ship system equipment generated through hierarchical bounding box collision detection and B-spline path smoothing processing includes: Map the ship outfitting workspace to a six-degree-of-freedom space , where the translational degrees of freedom are , and the rotational degrees of freedom are , respectively the x-axis direction, the y-axis direction, and the z-axis direction, respectively the rotational degree of freedom in the x-axis direction, the rotational degree of freedom in the y-axis direction, and the rotational degree of freedom in the z-axis direction; Define the free space , the obstacle space and the narrow area ; Calculate the local obstacle density based on the hypersphere sampling method , and dynamically adjust the sampling strategy based on the local obstacle density ; when is greater than or equal to the first threshold, it is determined as a narrow passage, and directional offset sampling is performed along the main axis direction of the narrow passage; when is less than the first threshold, it is determined as an open area, and uniform random sampling is adopted within the open area; Calculate the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning; generate a smooth B-spline path using two-way tree synchronous expansion and hierarchical bounding box collision detection, and output a sequence of path key points including pose sequences and time parameters.

[0058] Further, the calculation of the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning includes: By obtaining the three-dimensional geometric model of the ship system equipment assembly, calculate its convex hull vertex set ; where is the rigid body transformation of the nth pose, ; Define and two poses, and perform rigid body transformation on the convex hull vertex set , where the rigid body transformation matrix includes the translation component and the rotational component ; are respectively the translational component in the x-axis direction, the translational component in the y-axis direction, and the translational component in the z-axis direction, and are respectively the rotational component in the x-axis direction, the rotational component in the y-axis direction, and the rotational component in the z-axis direction; Calculate the maximum displacement distance and the average displacement distance of the corresponding vertices , and use the maximum displacement distance as the metric distance between the pose

[0059] Furthermore, the realizing the dynamic simulation and collision warning of the outfitting process according to the optimal outfitting path in combination with a 3D visualization engine includes: Loading the 3D geometric model of the ship system equipment assembly and the planned path data based on the 3D visualization engine, rendering the motion trajectory and key pose points of the assembly in real time, and triggering a collision warning when it is detected that the minimum distance is less than the safety threshold; Highlighting the risk area through the 3D visualization engine, automatically generating path correction suggestions, and updating the optimal outfitting path.

[0060] Furthermore, the using industrial cameras, laser trackers, and optical targets to collect outfitting process data, establishing a mapping relationship between the feature deviation and the 3D geometric model of the ship system equipment assembly, predicting the deviation trend through a long short-term memory neural network, and generating a correction instruction to correct the optimal outfitting path includes: Arranging industrial cameras, laser trackers, and optical targets to form a multi-sensor network, setting multi-view observation stations, and planning the spatial distribution of the target point array; Calculating the 3D coordinates of the feature points through bundle adjustment, and establishing the conversion relationship between the global coordinate system and the local coordinate system; Extracting the measured pose data of the key control points, calculating the deviation value from the theoretical model, and establishing the mapping relationship between the deviation field and the 3D geometric model of the ship system equipment assembly; Using radial basis function interpolation to display the deviation distribution heat map in real time, predicting the deviation development trend based on a long short-term memory (LSTM) neural network, and generating a correction instruction, and correcting the optimal outfitting path according to the correction instruction.

[0061] Embodiment 4 The present application provides a ship outfitting path planning and simulation modeling system in Embodiment 4 of the present application, including: A data acquisition and virtual modeling module, which is used to collect the feature parameter of the ship system equipment outfitting site environment, and construct a virtual 3D environment that is consistent between virtual and real through surface fitting and data alignment; Path planning and simulation module, which is used to dynamically plan a six-degree-of-freedom path based on a multi-constraint condition of obstacles, narrow channels and dynamic environment, adopt a narrow-channel optimization algorithm (NARRT*), and based on a convex hull pose distance metric model. An optimal outfitting path for the ship system equipment is generated through hierarchical bounding box collision detection and B-spline path smoothing processing. According to the optimal outfitting path, combined with a three-dimensional visualization engine, dynamic simulation and collision warning of the outfitting process are realized; Deviation detection and correction module, which is used to collect outfitting process data by using industrial cameras, laser trackers and optical targets, establish a mapping relationship between feature deviations and the three-dimensional geometric model of the ship system equipment assembly, predict the deviation trend through a long short-term memory neural network, and generate correction instructions to correct the optimal outfitting path.

[0062] Among them, the product assembly information model includes: Geometric information of parts (three-dimensional geometric shape, patch model, wireframe model), hierarchical structure information (parent-child relationship, assembly constraints); Basic attribute information (name, material attributes, mass, process requirements) and pose information (initial pose, target pose, bounding box center pose).

[0063] Furthermore, the path planning and simulation module includes: Narrow-channel determination unit, which is used to calculate the local obstacle density based on the hypersphere sampling method , and dynamically adjust the sampling strategy based on the local obstacle density ; Convex hull pose distance calculation unit, which is used to calculate the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set, as the distance metric for path planning; Path smoothing unit, which is used to generate a smooth B-spline path by using two-way tree synchronous expansion and hierarchical bounding box collision detection, and output a sequence of path key points including pose sequences and time parameters.

[0064] Furthermore, the deviation detection and correction module includes: Multi-sensor calibration unit, which is used to arrange industrial cameras, laser trackers and optical targets to form a multi-sensor network, set multi-view observation stations, and plan the spatial distribution of the target point array; Deviation heat map generation unit, which is used to display the deviation distribution heat map in real time through radial basis function interpolation; Early warning level classification unit, which is used to predict the deviation development trend based on long short-term memory, and output an early warning signal including normal, attention, and danger levels.

[0065] Among them, the system supports AR / VR interaction, realizes the visual guidance of the assembly process through a virtual-real fusion engine, and is compatible with mainstream CAD / CAPP systems to form a full-process digital closed loop.

[0066] This application also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the steps of the method described above are implemented.

[0067] This application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described above are implemented.

[0068] This application solves the problems of low path planning efficiency, insufficient virtual-real fusion accuracy, and lack of dynamic collision warning in a complex cabin environment, significantly reduces the cognitive load of workers, improves the assembly qualification rate and the accuracy of collision warning, and is applicable to the digital outfitting operations of complex system equipment such as ship pipelines and shaft and rudder systems, providing key technical support for digital shipbuilding.

[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for ship outfitting path planning and simulation modeling, characterized in that Includes steps: Collect the environmental characteristic parameters of the ship system equipment outfitting site, and build a three-dimensional virtual environment that is consistent with reality through surface fitting and data alignment; Based on the multi-constraint conditions of obstacles, narrow channels and dynamic environments, a narrow channel optimization algorithm is adopted to dynamically plan a six-degree-of-freedom path based on the convex hull pose distance measurement model. The optimal outfitting path of the ship system equipment is generated through hierarchical bounding box collision detection and B-spline path smoothing. According to the optimal outfitting path, the dynamic simulation and collision warning of the outfitting process are realized in combination with a three-dimensional visualization engine. Industrial cameras, laser trackers and optical targets are used to collect outfitting process data, and a mapping relationship between feature deviation and three-dimensional geometric model of ship system equipment assembly is established. The deviation trend is predicted through a long short-term memory neural network and correction instructions are generated to correct the optimal outfitting path.

2. The method for ship outfitting path planning and simulation modeling according to claim 1, wherein, The collecting of on-site environmental characteristic parameters of ship system equipment outfitting and the construction of a three-dimensional virtual environment consistent with reality by means of surface fitting and data alignment include: Collecting on-site feature data by means of a 3D laser scanner, an industrial camera and a depth sensor, the feature data including at least one of chimney sections, side channels, seawater main pipe units, bulk carrier shaft rudder systems, oil tanker tube bundle unit systems, auxiliary positioning positions, work space layout and equipment bases; Extract point and surface feature information based on surface fitting and data alignment; Based on the collected feature data and the extracted point and surface feature information, a product assembly information model, an assembly process information model and an assembly path simulation information model are constructed through scene 3D rendering, and the product assembly information model, the assembly process information model and the assembly path simulation information model are formed into a virtual scene of the outfitting site; A multi-coordinate system is defined, and the position and posture data of the ship system equipment are corrected in real time through marker recognition, the parameters of the outfitting site virtual scene are corrected, and a three-dimensional virtual environment consistent with the outfitting site is reconstructed.

3. The method for ship outfitting path planning and simulation modeling according to claim 2, characterized in that, The method of constructing a product assembly information model, an assembly process information model and an assembly path simulation information model through scene 3D rendering based on the collected feature data and the extracted point and surface feature information includes: Extract and integrate component geometry information, hierarchical structure information, basic attribute information and posture information from the collected feature data, integrate the information using XML format files, create component node files, and form a product assembly information model in an augmented reality scenario; Extract and integrate the types of tasks, assembly processes and steps, operation objects, tools and resources, operation methods, assembly paths and prerequisites in the assembly process from the collected feature data, create XML files for the steps, refine the process steps, and form an assembly process information model; The planning algorithm is used to plan and acquire the assembly object, and the assembly base environment, assembly motion path, dynamic constraints and interference requirements are used as the motion posture information of the parts. The motion details of the parts assembly process are described according to the motion posture information of the parts, and the parts assembly path XML file is created to form an assembly path simulation information model.

4. The method for ship outfitting path planning and simulation modeling according to claim 1, characterized in that The adoption of the narrow-channel optimization algorithm to dynamically plan a six-degree-of-freedom path based on the convex hull pose distance metric model, and generate the optimal outfitting path of the ship system equipment through hierarchical bounding box collision detection and B-spline path smoothing processing includes: Map the ship fitting-out workspace to a six-degree-of-freedom space , where the translational degrees of freedom are , and the rotational degrees of freedom are , in the x-axis direction, y-axis direction, and z-axis direction respectively, and are the rotational degrees of freedom about the x-axis, y-axis, and z-axis respectively; Define free space and obstacle space as well as narrow areas ; Calculating local obstacle density based on hypersphere sampling method , dynamically adjusting the sampling strategy based on the local obstacle density ; when is greater than or equal to the first threshold, it is determined as a narrow channel, and directional bias sampling is performed along the main axis direction of the narrow channel; when is less than the first threshold, it is determined as an open area, and uniform random sampling is used within the open area; Calculating the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning; adopting bidirectional tree synchronous expansion and hierarchical bounding box collision detection to generate a smooth B-spline path, and outputting a sequence of path key points including pose sequences and time parameters.

5. The ship fitting path planning and simulation modeling method according to claim 4, characterized in that The calculation of the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set as the distance metric for path planning includes: By obtaining the three-dimensional geometric model of the ship system equipment assembly, calculate the set of convex hull vertices ; where is a rigid body transformation for the nth pose, ; Definition and Two poses are used to perform a rigid body transformation on the convex hull vertex set , where the rigid body transformation matrix includes a translation component and a rotation component ; are the translation components in the x-axis direction, y-axis direction, and z-axis direction respectively, are the rotation components in the x-axis direction, y-axis direction, and z-axis direction respectively; Calculate the maximum displacement distance and average displacement distance of the corresponding vertices , and use the maximum displacement distance as the metric distance between poses.

6. The method for ship outfitting path planning and simulation modeling according to claim 1, wherein The realization of dynamic simulation and collision warning of the outfitting process according to the optimal outfitting path in combination with a three-dimensional visualization engine includes: Loading the three-dimensional geometric model of the ship system equipment assembly and the planned path data based on the three-dimensional visualization engine, rendering the motion trajectory and key pose points of the assembly in real time, and triggering a collision warning when the detected minimum distance is less than the safety threshold; Highlighting the risk area through the three-dimensional visualization engine, automatically generating path correction suggestions, and updating the optimal outfitting path.

7. The method for ship outfitting path planning and simulation modeling according to claim 1, wherein The use of industrial cameras, laser trackers, and optical targets to collect outfitting process data, establish the mapping relationship between feature deviations and the three-dimensional geometric model of the ship system equipment assembly, and predict the deviation trend through a long short-term memory neural network and generate a correction instruction to correct the optimal outfitting path includes: Arranging industrial cameras, laser trackers, and optical targets to form a multi-sensor network, setting multi-view observation stations, and planning the spatial distribution of the target point array; Calculating the three-dimensional coordinates of feature points through bundle adjustment and establishing the conversion relationship between the global coordinate system and the local coordinate system; Extracting the measured pose data of key control points, calculating the deviation value from the theoretical model, and establishing the mapping relationship between the deviation field and the three-dimensional geometric model of the ship system equipment assembly; Using radial basis function interpolation to display the deviation distribution heat map in real time, predicting the deviation development trend based on a long short-term memory neural network and generating a correction instruction, and correcting the optimal outfitting path according to the correction instruction.

8. A ship outfitting path planning and simulation modeling system, characterized in that, Including: A data acquisition and virtual modeling module for collecting the characteristic parameters of the outfitting site environment of the ship system equipment and constructing a virtual-real consistent three-dimensional virtual environment through surface fitting and data alignment; A path planning and simulation module for dynamically planning a six-degree-of-freedom path based on the convex hull pose distance metric model using the narrow-channel optimization algorithm under multi-constraint conditions of obstacles, narrow channels, and dynamic environments, generating the optimal outfitting path of the ship system equipment through hierarchical bounding box collision detection and B-spline path smoothing processing, and realizing dynamic simulation and collision warning of the outfitting process according to the optimal outfitting path in combination with a three-dimensional visualization engine; A deviation detection and correction module for using industrial cameras, laser trackers, and optical targets to collect outfitting process data, establishing the mapping relationship between feature deviations and the three-dimensional geometric model of the ship system equipment assembly, and predicting the deviation trend through a long short-term memory neural network and generating a correction instruction to correct the optimal outfitting path.

9. The shipfitting path planning and simulation modeling system according to claim 8, wherein, The path planning and simulation module includes: A narrow passage determination unit for calculating the local obstacle density based on the hypersphere sampling method , and dynamically adjusting the sampling strategy based on the local obstacle density ; The convex hull pose distance calculation unit is used to calculate the maximum displacement distance between poses based on the rigid body transformation of the convex hull vertex set, and use it as the distance metric for path planning; The path smoothing unit is used to generate a smooth B-spline path by means of bidirectional tree synchronous expansion and hierarchical bounding box collision detection, and output a sequence of path key points including pose sequences and time parameters.

10. The ship outfitting path planning and simulation modeling system according to claim 8, characterized in that, The deviation detection and correction module includes: The multi-sensor calibration unit is used to arrange industrial cameras, laser trackers and optical targets to form a multi-sensor network, set multi-view observation stations, and plan the spatial distribution of the target point array; The deviation heat map generation unit is used to display the deviation distribution heat map in real time through radial basis function interpolation; The warning level classification unit is used to predict the deviation development trend based on long short-term memory, and output warning signals including normal, attention, and danger levels.

11. A computer device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 7.

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

Images