A method for digital modeling and layout of elevator shaft space based on SLAM technology

Abstract

The present invention relates to a method for digital modeling and layout of an elevator shaft space based on SLAM technology, comprising: moving a depth camera in the vertical direction of the shaft, while rotating a pan / tilt head, and performing a full-scale scan of the elevator shaft around the depth camera; utilizing the correlation between map points to segment map points that do not have motion consistency, using static map points to perform pose estimation and static map construction, and obtaining a high-precision three-dimensional model of the elevator shaft space; placing and marking sample lines based on the high-precision three-dimensional digital model of the elevator shaft, and using a high-precision level as a reference, according to the elevator installation requirements, establishing the spatial position and posture of the layout reference relative to the elevator shaft model, projecting the elevator guide rail bracket mounting holes onto the three-dimensional model surface, obtaining accurate installation position information of the elevator guide rail bracket mounting hole surface, and realizing full digital layout of the shaft. The present invention can solve the installation error caused by the drift of the traditional layout reference.

Description

Technical Field

[0001] The present invention relates to the technical field of elevators, and in particular to a method for digital modeling and layout of an elevator shaft space based on SLAM technology. Background Art

[0002] Elevators are specialized equipment designed to transport people or cargo. Unlike ordinary industrial electromechanical products, they must be assembled on-site at the user's construction site and accepted before they can be put into use. Therefore, the performance and quality of an elevator depend not only on the factory's manufacturing quality but also on the quality of the on-site installation. Traditionally, installing elevators in a hoistway requires setting a template frame to establish an installation benchmark. Typically, the machine room lobby base line is used as a reference, and the template lines are placed and marked. A plumb bob is then hung at the corresponding position to guide subsequent construction. This installation process requires placing a plumb line to obtain a vertical reference and determine the verticality of the guide rails. However, the plumb line can swing slightly due to external interference. As the building height increases, it is prone to large horizontal displacement at the end, which inevitably affects the installation accuracy of the elevator. Summary of the Invention

[0003] The purpose of the present invention is to provide a method for digital modeling and layout of elevator shaft space based on SLAM technology, which can solve the problems of layout reference drift and large installation errors in traditional installation, and ensure high-quality installation of elevators under complex and harsh working conditions.

[0004] To achieve the above object, the present invention provides the following technical solution: a method for digital modeling and layout of an elevator shaft space based on SLAM technology, the method comprising the following steps:

[0005] Step S1: Fix the depth camera, the rotating pan-tilt head, and the high-precision level on the lifting platform. Driven by the lifting system, the depth camera moves vertically along the shaft. At the same time, the rotating pan-tilt head rotates, and the depth camera performs a full scan of the elevator shaft.

[0006] Step S2: To address the problem of scarce static textures in the shaft, the correlation between map points is used to segment map points that do not have motion consistency. Static map points are used to perform pose estimation and static map construction to obtain a high-precision 3D model of the elevator shaft space.

[0007] Step S3: Based on the high-precision three-dimensional digital model of the elevator shaft, sample lines are placed and marked. Using a high-precision level as a reference, the spatial position and posture of the layout reference relative to the elevator shaft model are determined according to the elevator installation requirements;

[0008] Step S4: Project the installation holes of the elevator guide rail bracket onto the surface of the three-dimensional model to obtain accurate installation position information of the installation holes of the elevator guide rail bracket, thereby realizing full digital layout of the hoistway.

[0009] Furthermore, the "depth camera performs a full-scale scan of the elevator shaft" in step S1 is further: the depth camera achieves 360-degree continuous rotation through the pan-tilt platform, and the depth camera is lowered to perform a full-scale scan of the elevator shaft.

[0010] Furthermore, the descent speed V of the depth camera is less than the vertical height H of the depth camera field of view*the angular velocity of the gimbal w / 360.

[0011] Furthermore, the map points in step S2 are point clouds measured by a depth camera.

[0012] Furthermore, the step S2 of "using the correlation between map points to segment map points that do not have motion consistency" is further as follows: in the point cloud continuously scanned by the depth camera, the same area will be scanned multiple times, that is, the point cloud data at different times will overlap, and the duplicate point cloud data is removed from the point cloud data collected at different times.

[0013] Furthermore, the step S2 of "performing pose estimation and static map construction to obtain a high-precision three-dimensional model of the elevator shaft space" is further as follows: through the ICP registration algorithm of the point cloud, the point cloud data at different times are transformed into the same reference coordinate system, and after the registered point cloud data is simplified, triangular patches are used for fitting to obtain a three-dimensional surface, thereby obtaining a high-precision three-dimensional model of the elevator shaft space.

[0014] Furthermore, the step S3 of "establishing the spatial position and posture of the layout reference relative to the elevator shaft model according to the elevator installation requirements" is further as follows: the projector and the depth camera are fixed to the same reference, the depth camera is used to scan the shaft surface, the scanned point cloud data is aligned with the three-dimensional model of the shaft, and the position and posture information of the installation reference of the projector and the depth camera is obtained, and the position of the projection pattern is adjusted accordingly to ensure that the projection pattern is consistent with the construction position.

[0015] Furthermore, the step S3 of "placement and position marking of sample line" is further as follows: using a projector to project the placement and position marking of sample line onto the well surface.

[0016] Beneficial effects of the present invention: The accuracy of the elevator shaft layout benchmark and the accuracy of the position of the installation holes of the elevator guide rail brackets and floor door sills and door frames are the key to achieving high-quality installation of elevators. The present invention combines a depth camera to obtain high-precision point cloud data of the shaft, and on this basis constructs a three-dimensional digital model of the elevator shaft space. Based on the three-dimensional digital model, a vertical layout benchmark is further established, and its spatial position and posture relative to the elevator shaft model are obtained, which serves as the benchmark for elevator installation. In addition, by scanning the local shaft point cloud data with a depth camera and combining the fusion matching algorithm of the local point cloud data with the shaft digital model, accurate position information is provided for the construction of the installation holes of the elevator guide rail brackets and floor door sills and door frames. It fundamentally solves the problems of drift of the traditional layout benchmark and large installation errors, and ensures high-quality installation of elevators under complex and harsh working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a schematic diagram of the depth camera scan;

[0018] Figure 2 This is the registration diagram of the three-dimensional model of the elevator shaft of the present invention;

[0019] Figure 3 This is a schematic diagram of the projector projection.

[0020] Among them: 1. Rotating gimbal and high-precision level, 2. Depth camera, 3. Elevator shaft, 4. Rotational motion direction, 5. Vertical motion direction, 6. Scanning surface, 7. Projection surface, 8. Structured light, 9. Rotating gimbal coordinate system, 10. Elevator shaft coordinate system, 11. Camera coordinate system. DETAILED DESCRIPTION

[0021] See also Figures 1 to 3 As shown, the present invention provides a method for digital modeling and layout of an elevator shaft space based on SLAM technology, the method comprising the following steps:

[0022] Step S1: Fix the depth camera, the rotating pan-tilt head, and the high-precision level on the lifting platform. Driven by the lifting system, the depth camera moves vertically along the shaft. At the same time, the rotating pan-tilt head rotates, and the depth camera performs a full scan of the elevator shaft.

[0023] Step S2: To address the problem of scarce static textures in the shaft, the correlation between map points is used to segment map points that do not have motion consistency. Static map points are used to perform pose estimation and static map construction to obtain a high-precision 3D model of the elevator shaft space.

[0024] Step S3: Based on the high-precision three-dimensional digital model of the elevator shaft, sample lines are placed and marked. Using a high-precision level as a reference, the spatial position and posture of the layout reference relative to the elevator shaft model are determined according to the elevator installation requirements;

[0025] Step S4: Project the installation holes of the elevator guide rail bracket onto the surface of the three-dimensional model to obtain accurate installation position information of the installation holes of the elevator guide rail bracket, thereby realizing full digital layout of the hoistway.

[0026] The present invention will be further described below with reference to a specific embodiment:

[0027] A digital modeling and layout method for elevator shaft space based on SLAM technology.

[0028] Step S1: Fix the depth camera, the rotating pan-tilt head, and the high-precision level on the lifting platform. Driven by the lifting system, the depth camera moves vertically along the shaft. At the same time, the rotating pan-tilt head rotates, and the depth camera performs a full scan of the elevator shaft.

[0029] The "depth camera performs a full-scale scan of the elevator shaft" in step S1 is further: the depth camera is rotated 360 degrees continuously through the pan-tilt platform, and the depth camera is lowered to perform a full-scale scan of the elevator shaft.

[0030] The descent speed V of the depth camera is less than the vertical height H of the depth camera field of view*the angular velocity of the gimbal w / 360.

[0031] Step S2: To address the problem of scarce static textures in the shaft, we use the correlation between map points to segment map points that do not have motion consistency, and use static map points to perform pose estimation and static map construction to obtain a high-precision three-dimensional model of the elevator shaft space; introduce vision-based simultaneous localization and mapping technology (SLAM), use a high-precision depth camera to perform a global scan of the elevator shaft space, and construct a three-dimensional digital model of the elevator shaft. The settings of the depth camera, rotating pan / tilt head, and high-precision level are as follows: Figure 1 shown.

[0032] The map points in step S2 are point clouds measured by the depth camera.

[0033] The step S2 of "using the correlation between map points to segment map points that do not have motion consistency" is further as follows: in the point cloud continuously scanned by the depth camera, the same area will be scanned multiple times, that is, the point cloud data at different times will overlap, and the duplicate point cloud data is removed from the point cloud data collected at different times.

[0034] The step S2 of "performing pose estimation and static map construction to obtain a high-precision three-dimensional model of the elevator shaft space" is further as follows: through the ICP registration algorithm of the point cloud, the point cloud data at different times are transformed into the same reference coordinate system, and after the registered point cloud data is simplified, triangular patches are used for fitting to obtain a three-dimensional surface, thereby obtaining a high-precision three-dimensional model of the elevator shaft space.

[0035] Step S3: Based on the high-precision three-dimensional digital model of the elevator shaft, sample lines are placed and marked. Using a high-precision level as a reference, and according to the elevator installation requirements, the spatial position and posture of the layout reference relative to the elevator shaft model are established; the sample line is a marking line used as a reference line for drilling or installation.

[0036] The step S3 of "establishing the spatial position and posture of the layout reference relative to the elevator shaft model according to the elevator installation requirements" is further as follows: the projector and the depth camera are fixed to the same reference, the shaft surface is scanned by the depth camera, the scanned point cloud data is aligned with the shaft 3D model, the position and posture information of the projector and depth camera installation reference is obtained, and the projection pattern position is adjusted accordingly to ensure that the projection pattern is consistent with the construction position. The alignment of the scanned point cloud data with the shaft 3D model is achieved by using the sampling ICP (Iterative Closest Point) algorithm. Before layout, a complete shaft 3D point cloud model will be obtained by scanning. Figure 2 As shown, when laying out, the depth camera will scan the 3D data of the local well to be laid out, obtain the data of the scanned surface, and align it with the 3D data of the whole well. After the alignment, the transformation matrix of the two point cloud data can be obtained, thereby obtaining the position and posture information of the projector and the depth camera installation reference. According to the position and posture information of the projector and the depth camera installation reference, the coordinate position of the layout line pattern is changed, and a picture of the layout line is generated. Finally, the layout line is projected onto the well surface by the projector to form a Figure 2 The projection surface shown, the position information such as the camera coordinate system on the depth camera, the elevator shaft coordinate system on the elevator shaft and the rotating pan-tilt coordinate system as shown Figure 2 shown.

[0037] The step S3 of "placement and position marking of sample line" is further as follows: using a projector to project the placement and position marking of sample line onto the well surface, such as Figure 3 shown.

[0038] Step S4: Project the elevator guide rail bracket mounting holes onto the 3D model surface to obtain precise installation position information on the surface of the elevator guide rail bracket mounting holes, thereby achieving fully digitalized hoistway layout. After the ideal model of the construction drawing is discretized into a point cloud, it is aligned to the measured 3D hoistway model using the ICP algorithm. The elevator guide rail bracket mounting holes are then projected onto the 3D model surface to obtain the position information of the mounting holes. Based on the high-precision fully digital 3D model of the elevator hoistway, the sample lines are placed and marked. Using an electronic level as a reference, the spatial position and posture of the layout reference relative to the elevator hoistway model are established according to the elevator installation requirements. The coordinate position of the layout line pattern is then changed to generate a layout line image. The layout line is then projected onto the hoistway surface using a projector to achieve fully digitalized layout. This invention fundamentally solves the installation error caused by the drift of the traditional layout reference, ensuring high-quality installation of the elevator.

[0039] In summary, during the elevator installation process, a depth camera scans the local shaft point cloud data around the installation work platform, providing precise location information for the construction of the elevator guide rail brackets and the installation holes for the door thresholds and door frames. To address the possibility of wobbling on the installation work platform, the information sources of the depth camera and the level are integrated, and classic SLAM methods such as point cloud registration and loopback testing are integrated. By aligning the point cloud data with the 3D shaft model, the projector and depth camera are fixed to the same reference. The depth camera is used to scan the shaft surface, and the scanned point cloud data is aligned with the 3D shaft model. The position and posture information of the projector and depth camera installation references are obtained. The projection pattern position is adjusted accordingly to ensure that the projection pattern is consistent with the construction position, thus eliminating the impact of wobbling.

[0040] The above description is only a preferred embodiment of the present invention and should not be understood as limiting the present application. All equivalent changes and modifications made within the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A method for digital modeling and layout of elevator shaft space based on SLAM technology, characterized in that: The method comprises the following steps: Step S1: Fix the depth camera, the rotating pan-tilt head, and the high-precision level on the lifting platform. Driven by the lifting system, the depth camera moves vertically along the shaft. At the same time, the rotating pan-tilt head rotates, and the depth camera performs a full scan of the elevator shaft. Step S2: To address the problem of scarce static textures in the shaft, the correlation between map points is used to segment map points that do not have motion consistency. Static map points are used to perform pose estimation and static map construction to obtain a high-precision 3D model of the elevator shaft space. Step S3: Based on the high-precision 3D model of the elevator shaft space, sample lines are placed and marked. Using a high-precision level as a reference, and according to elevator installation requirements, the spatial position and posture of the layout reference relative to the high-precision 3D model of the elevator shaft space are determined; Step S4: Project the installation holes of the elevator guide rail bracket onto the surface of the three-dimensional model to obtain accurate installation position information of the installation holes on the surface of the elevator guide rail bracket, thereby realizing full digital layout of the hoistway; The step S2 of "performing pose estimation and static map construction to obtain a high-precision three-dimensional model of the elevator shaft space" further includes: transforming the point cloud data at different times into a common reference coordinate system using an ICP registration algorithm for the point cloud; simplifying the registered point cloud data; and fitting the three-dimensional surface using triangular patches to obtain a high-precision three-dimensional model of the elevator shaft space. The step S3 of "establishing the spatial position and posture of the lofting reference relative to the high-precision three-dimensional model of the elevator shaft space according to the elevator installation requirements" further includes: fixing the projector and the depth camera to the same reference, scanning the shaft surface using the depth camera, aligning the scanned point cloud data with the high-precision three-dimensional model of the elevator shaft space, obtaining the position and posture information of the projector and depth camera installation references, and adjusting the projection pattern position accordingly to ensure that the projection pattern is consistent with the construction position; The “placement and position marking of the sample line” in step S3 further includes: using a projector to project the placement and position marking of the sample line onto the well surface.

2. The method for digital modeling and layout of an elevator shaft space based on SLAM technology according to claim 1, characterized in that: The "depth camera performs a full-scale scan of the elevator shaft" in step S1 is further: the depth camera is rotated 360 degrees continuously through the pan-tilt head, and the depth camera is lowered to perform a full-scale scan of the elevator shaft.

3. The method for digital modeling and layout of an elevator shaft space based on SLAM technology according to claim 2, characterized in that: The value V of the descent speed of the depth camera is less than the value of the vertical height H* (pan-tilt angular velocity w / 360) of the depth camera's field of view.

4. The method for digital modeling and layout of an elevator shaft space based on SLAM technology according to claim 1, characterized in that: The map points in step S2 are point clouds measured by the depth camera.

5. The method for digital modeling and layout of an elevator shaft space based on SLAM technology according to claim 1, characterized in that: The step S2 of "using the correlation between map points to segment map points that do not have motion consistency" is further as follows: in the point cloud continuously scanned by the depth camera, the same area will be scanned multiple times, that is, the point cloud data at different times will overlap, and the duplicate point cloud data is removed from the point cloud data collected at different times.

Images