Elevator installation guidance method and system

Abstract

The present invention discloses an elevator installation guidance method and system, which relates to the field of elevator installation technology, including: scanning the internal structure of the elevator shaft through a laser rangefinder to obtain three-dimensional point cloud data, and synchronously collecting image data of the inner surface of the shaft through a structured light camera; fusing the laser rangefinder data with the image data to generate a three-dimensional environmental model including shaft dimensions, structural layout and environmental characteristics; applying an image recognition algorithm to analyze the generated three-dimensional environmental model, identifying and marking elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points; based on the marked points, using a path planning algorithm to generate the installation sequence and installation path of the elevator guide rails to guide the installation of elevator components. The elevator installation guidance method and system provided by the present invention can accurately obtain the actual spatial layout of the elevator shaft, effectively eliminate the errors existing in traditional manual measurement, and ensure the precise positioning of the elevator installation points.

Description

Technical Field

[0001] The present invention relates to the technical field of elevator installation, and in particular to an elevator installation guidance method and system. Background Art

[0002] The installation of elevators is a complex and high-precision process, which usually requires manual measurement and positioning one by one according to drawings. However, traditional elevator installation methods have many shortcomings, especially in terms of accuracy, efficiency and adaptability. First, the elevator shaft usually has a complex structure, and it is difficult to avoid small errors in actual installation. The traditional manual measurement method often cannot accurately obtain the actual position and size of each component inside the shaft, which makes the installation accuracy of the elevator components impossible to guarantee. Due to errors in manual measurement, elevator components may be installed in the wrong position, causing the elevator to malfunction during use, seriously affecting the safety of elevator operation. Secondly, the elevator installation process involves multiple complex installation points. The spatial relationship and installation sequence between these installation points must be accurately calculated and planned. Traditional manual experience cannot complete this task efficiently. Installers usually operate according to existing drawings, but lack the ability to analyze the installation sequence, path and installation environment changes in real time. The lack of an intelligent auxiliary system makes the installation process inefficient, manual adjustments are frequent, and it is difficult to detect and correct problems in a timely manner. Furthermore, the traditional method also lacks a real-time feedback mechanism. Once a deviation or delay occurs, it usually requires a lot of manpower and time to correct it, which seriously affects the overall installation progress of the elevator. In addition, with the diversification and complexity of elevator shaft design, traditional methods are unable to cope with these changes and lack adaptability and flexibility. In summary, the limitations of existing elevator installation technology, especially in terms of accuracy and flexibility, urgently need technological innovation to improve installation quality and efficiency. Summary of the invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by the present invention is that in the prior art, elevator installation relies on manual operation and measurement, which is difficult to avoid errors and has poor adaptability in complex installation environments. The traditional installation method cannot adjust the installation sequence in real time, which may lead to errors or delays during the installation process, affecting the installation quality and progress.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: an elevator installation guidance method, comprising:

[0006] The internal structure of the elevator shaft is scanned by a laser rangefinder to obtain three-dimensional point cloud data, and the image data of the inner surface of the shaft is simultaneously collected by a structured light camera. The laser rangefinder data and the image data are fused to generate a three-dimensional environmental model that includes the shaft size, structural layout and environmental characteristics. The image recognition algorithm is used to parse the generated three-dimensional environmental model to identify and mark the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points. Based on the marked points, the path planning algorithm is used to generate the installation sequence and installation path of the elevator guide rails to guide the installation of elevator components.

[0007] As a preferred solution of the elevator installation guidance method described in the present invention, the data fusion includes a first preprocessing step for the three-dimensional point cloud data, the first preprocessing step includes denoising and point cloud sparseness; a second preprocessing step is performed on the image data acquired by the structured light camera, the second preprocessing step includes color correction and image enhancement; the preprocessed three-dimensional point cloud data is spatially aligned with the image data using a point cloud registration algorithm; and the texture information in the image data is mapped to the three-dimensional point cloud data through color information fusion.

[0008] As a preferred solution of the elevator installation guidance method described in the present invention, the point cloud registration algorithm includes using feature matching technology to extract key feature points in three-dimensional point cloud data and image data; applying an iterative closest point algorithm to optimize the point cloud registration by minimizing the distance error between point clouds; calculating a rigid transformation matrix, rotating and translating the three-dimensional point cloud data according to the results of the iterative closest point algorithm, and aligning the image data.

[0009] As a preferred solution of the elevator installation guidance method described in the present invention, the analysis of the generated three-dimensional environment model includes dividing the generated three-dimensional environment model into multiple two-dimensional image slices, each slice corresponding to a specific area in the shaft; applying a pre-trained convolutional neural network model to each two-dimensional image slice to extract image feature vectors; matching the extracted image feature vectors with predefined installation point feature templates to calculate a similarity score; determining image slices whose similarity scores exceed a preset threshold, and marking the positions of elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points in the three-dimensional environment model.

[0010] As a preferred solution of the elevator installation guidance method described in the present invention, the path planning algorithm includes determining the priority and dependency of each installation step according to the spatial relationship between the elevator guide rail fixing point, the car positioning point, the counterweight installation point, the door machine installation point and the control system installation point; establishing a directed acyclic graph between the installation steps, in which the nodes represent the installation steps and the edges represent the dependency between the installation steps; applying a topological sorting algorithm to the established directed acyclic graph to generate the execution order of the installation steps.

[0011] As a preferred embodiment of the installation guidance method of the elevator described in the present invention, the guidance of the installation of elevator components includes installing the elevator guide rails to the marked guide rail fixing points using fixing tools according to the installation sequence generated by path planning; moving the car and installing it to the marked car positioning points using a positioning device; installing the counterweight assembly and adjusting the counterweight position to the marked counterweight installation point; installing the door machine to the marked door machine installation point using a symmetrical installation tool; and installing the control system to the marked control system installation point using a connection tool.

[0012] As a preferred solution of the elevator installation guidance method described in the present invention, the guidance of elevator component installation also includes real-time monitoring of the installation progress and comparing it with the preset installation progress. When the installation progress does not meet the preset progress, the node order and dependency relationship in the directed acyclic graph are adjusted according to the real-time data of the installation site, and the topological sorting algorithm is reapplied to update the execution order of the installation steps.

[0013] Another object of the present invention is to provide an elevator installation guidance system, which can solve the problem of unreasonable installation sequence in traditional methods and improve installation efficiency by constructing an elevator installation guidance system.

[0014] To solve the above technical problems, the present invention provides the following technical solutions: an elevator installation guidance system, comprising: a data acquisition module, used to scan the internal structure of the elevator shaft through a laser rangefinder to obtain three-dimensional point cloud data, and synchronously collect image data of the inner surface of the shaft through a structured light camera; a data fusion module, used to fuse the laser ranging data with the image data to generate a three-dimensional environmental model including the shaft size, structural layout and environmental characteristics; an image recognition module, used to apply an image recognition algorithm to analyze the generated three-dimensional environmental model, identify and mark the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points; and; an installation guidance module, used to generate the installation sequence and installation path of the elevator guide rail based on the marked points and use a path planning algorithm to guide the installation of elevator components.

[0015] Beneficial effects of the present invention: The elevator installation guidance method provided by the present invention generates a three-dimensional environment model by fusing laser ranging and image data. The present invention can accurately obtain the actual spatial layout of the elevator shaft, effectively eliminate the errors existing in traditional manual measurement, ensure the precise positioning of the elevator installation point, thereby improving the installation accuracy, avoiding installation deviations caused by measurement errors, and ensuring the correct installation of elevator components;

[0016] The present invention generates the installation sequence and path through the path planning algorithm, which can efficiently plan the installation steps of elevator components and optimize the installation sequence error problem that may be caused by insufficient manual experience in the traditional method. Using the topological sorting algorithm and directed acyclic graph, the dependency relationship between the installation steps is accurately processed, further improving the efficiency and controllability of the installation process;

[0017] In addition, the real-time monitoring and adjustment mechanism of the present invention enables the installation process to dynamically respond to changes in the on-site environment, monitor the installation progress in real time, and adjust the installation steps according to real-time data, thereby avoiding delays in the installation progress or errors. Through this intelligent installation method, the present invention not only improves the accuracy and safety of elevator installation, but also significantly reduces the installation cost and time cost, and has higher adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0019] Figure 1 The present invention provides an overall flow chart of an elevator installation guidance method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0022] Example 1, reference Figure 1 , as an embodiment of the present invention, provides an elevator installation guidance method, comprising:

[0023] Step S1: Scan the internal structure of the elevator shaft through a laser rangefinder to obtain three-dimensional point cloud data, and synchronously collect image data of the inner surface of the shaft through a structured light camera.

[0024] Step S2: Fusing the laser ranging data with the image data to generate a three-dimensional environment model including the shaft dimensions, structural layout and environmental features.

[0025] In step S2, the data fusion of the method of this embodiment includes performing a first preprocessing step on the three-dimensional point cloud data, the first preprocessing step includes denoising and point cloud thinning; performing a second preprocessing step on the image data acquired by the structured light camera, the second preprocessing step includes color correction and image enhancement; using a point cloud registration algorithm, spatially aligning the preprocessed three-dimensional point cloud data with the image data; and mapping the texture information in the image data to the three-dimensional point cloud data through color information fusion.

[0026] Specifically, the random noise points are removed by applying the Gaussian filtering algorithm to the acquired three-dimensional point cloud data. The Gaussian filtering uses the Gaussian kernel function to smooth the point cloud through convolution operation, retaining the main structural features and removing abnormal points.

[0027] The point cloud is sparsely processed using the voxel grid filtering method. By setting the voxel size parameter (such as 1 cubic centimeter), the point cloud is divided into a fixed-size voxel grid and a representative point is selected in each voxel to reduce the amount of point cloud data while maintaining key structural features.

[0028] The image data obtained by the structured light camera is color corrected, and an algorithm based on white balance is used to adjust the color temperature and hue of the image, eliminate the influence of ambient light on the image color, and ensure the accuracy of the image color.

[0029] Apply histogram equalization technology to enhance the contrast of the image, improve the visibility of image details, and enhance the accuracy of subsequent image recognition. Histogram equalization makes the contrast of the image more uniform by redistributing the grayscale of the image.

[0030] It should be noted that denoising and sparse processing not only improves the quality and processing efficiency of point cloud data, but also reduces the computational burden of subsequent algorithms. Gaussian filtering effectively removes noise from point clouds, while voxel grid filtering retains the key features of the shaft structure through uniform sampling, making the three-dimensional environment model more concise and accurate. Color correction and image enhancement processing not only improve the quality of image data, but also enhance the visibility of key features in the image, making the subsequent image recognition and annotation process more accurate and efficient.

[0031] Furthermore, in this embodiment, by adjusting the standard deviation of the Gaussian filter and the size parameters of the voxel grid, the smoothness and sparsity of the point cloud data can be flexibly controlled to adapt to the complexity and accuracy requirements of different wellbore environments. In this embodiment, adaptive histogram equalization technology is used to perform local contrast enhancement according to the lighting conditions of different areas, further improving the image details and feature recognition capabilities.

[0032] The point cloud registration algorithm includes using feature matching technology to extract key feature points in 3D point cloud data and image data; applying an iterative closest point algorithm to optimize the point cloud registration by minimizing the distance error between point clouds; calculating the rigid transformation matrix, rotating and translating the 3D point cloud data according to the results of the iterative closest point algorithm, and aligning the image data.

[0033] Feature matching uses ORB feature matching technology to extract key feature points from pre-processed 3D point cloud data and image data and establish feature point pairs. ORB technology is highly efficient and anti-rotation, and is suitable for real-time applications.

[0034] The color-based ICP algorithm is applied to optimize the point cloud registration accuracy by minimizing the color-weighted distance error between point clouds. Based on the traditional ICP algorithm, the weight of color information is increased, so that the registration process not only considers the geometric position but also the color consistency, thus improving the accuracy of registration.

[0035] Based on the results of the ICP algorithm, the rigid transformation matrix, including the rotation matrix and translation vector, is calculated to rotate and translate the 3D point cloud data so that it is accurately aligned with the image data.

[0036] It should be noted that the combination of feature matching and ICP algorithm ensures high-precision alignment of point cloud data and image data. ORB feature matching technology extracts stable key feature points to establish a preliminary correspondence between point cloud and image, and ICP algorithm further optimizes the registration accuracy to ensure the consistency of geometry and color information of the 3D environment model.

[0037] Furthermore, in this embodiment, through multiple rounds of iterations and dynamic adjustment of the parameters of the ICP algorithm (including the number of iterations and convergence conditions), it is possible to adapt to the complexity of different shaft environments and the diversity of point cloud data, thereby improving the robustness and accuracy of the registration algorithm.

[0038] Through color information fusion technology, the texture information in the image data is accurately mapped to the 3D point cloud data to generate a 3D environment model with rich color information. Color information fusion combines the color attributes in the point cloud (such as RGB values) with the color information in the image to enhance the visualization of the 3D model and facilitate subsequent image recognition and annotation.

[0039] It should be noted that color information fusion not only improves the visual effect of the 3D environment model, but also provides richer feature information for subsequent image recognition and annotation. By combining color and geometric information, the position of each installation point can be more accurately identified and located, reducing the recognition error caused by the singleness of the data.

[0040] Furthermore, in this embodiment, a mapping method based on RGB color space is adopted to optimize the fusion effect of color information and point cloud data by adjusting the color mapping ratio and weight parameters, thereby ensuring that the three-dimensional environment model has good visualization and recognition performance under different lighting and environmental conditions.

[0041] It should be noted that the data fusion process not only improves the accuracy and detail of the 3D environment model, but also solves the shortcomings that may exist in a single data source. For example, the laser rangefinder provides high-precision geometric data, but lacks color and texture information; the structured light camera provides rich color and texture information, but there may be deviations in geometric accuracy. Through data fusion, the advantages of both are combined to generate a high-precision 3D environment model with rich texture information.

[0042] Furthermore, the data fusion technology in the method of this embodiment can effectively solve the positioning accuracy problem during the installation of elevators. Through high-precision point cloud data and detailed image texture information, the location of each installation point can be accurately identified and marked, reducing human errors and improving installation quality and efficiency. At the same time, the application of data fusion technology enables the installation guidance system to operate stably in complex hoistway environments and adapt to the installation requirements of elevators of different types and specifications.

[0043] Step S3: Apply an image recognition algorithm to analyze the generated three-dimensional environment model, identify and mark the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points.

[0044] In step S3, the method of this embodiment parses the generated three-dimensional environment model, including dividing the generated three-dimensional environment model into multiple two-dimensional image slices, each slice corresponding to a specific area in the shaft; applying a pre-trained convolutional neural network model to each two-dimensional image slice to extract an image feature vector; matching the extracted image feature vector with a predefined installation point feature template to calculate a similarity score; determining the image slices whose similarity scores exceed a preset threshold, and marking the positions of the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points in the three-dimensional environment model.

[0045] Specifically, the image recognition process in this embodiment is as follows: the generated three-dimensional environment model is divided into multiple two-dimensional image slices using a fixed-interval plane slicing method. Each slice corresponds to a specific area in the shaft, and the slice interval is set to 10 cm to ensure that each slice covers sufficient spatial details.

[0046] In the 3D environment model, a plane is cut horizontally every 10 cm in the height direction to generate corresponding 2D image slices. Each slice is saved as a high-resolution grayscale image for subsequent feature extraction and recognition.

[0047] It should be noted that the fixed-interval planar slicing method can evenly distribute the slices, ensuring that the feature information of each area is fully captured and analyzed, avoiding blind spots in recognition due to uneven slicing.

[0048] Furthermore, in this embodiment, the slice interval can be dynamically adjusted according to the specific structural characteristics of the well to adapt to the complexity and detail requirements of different areas. For example, in areas with complex structures or rich details, the slice interval can be appropriately reduced to improve the accuracy of recognition.

[0049] Using convolutional neural networks, a pre-trained ResNet-50 convolutional neural network model is applied to each 2D image slice to extract image feature vectors. The ResNet-50 model can effectively train deep neural networks and extract high-level image features through the residual network structure.

[0050] The image feature vector is processed through a fully connected layer to reduce the dimensionality of the image, converting the high-dimensional feature vector into a low-dimensional vector of a fixed length (such as 256 dimensions) to facilitate subsequent matching calculations.

[0051] It should be noted that the ResNet-50 model is pre-trained on the ImageNet dataset and has powerful image feature extraction capabilities. It can effectively capture the key features of elevator installation points, such as the geometry of guide rail fixing points and identification symbols of car positioning points.

[0052] Furthermore, in this embodiment, the ResNet-50 model is fine-tuned on sample data of specific elevator installation points through transfer learning technology, further improving the model's recognition accuracy of target installation point features.

[0053] The predefined installation point feature templates are generated by engineers based on historical installation data and expert knowledge, covering the typical features of elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points.

[0054] The extracted image feature vector is matched with the predefined installation point feature template, and the similarity score is calculated using cosine similarity. Cosine similarity can measure the similarity of two vectors in direction and is suitable for comparing high-dimensional feature vectors.

[0055] A similarity score threshold is set (such as 0.8). When the similarity score of an image slice exceeds the threshold, it is considered that the target installation point exists in the area corresponding to the slice.

[0056] It should be noted that the feature matching process can effectively distinguish the target installation point from the non-target area by calculating the cosine similarity between the image feature vector and the installation point template, thereby ensuring the accuracy and reliability of the annotation.

[0057] Furthermore, in this embodiment, the accuracy of the matching result is optimized by dynamically adjusting the similarity threshold according to the complexity of the actual installation environment and the characteristic differences of the installation points. For example, in a complex environment, the threshold is appropriately lowered to improve the recognition rate, or the threshold is increased in a specific area to reduce mislabeling.

[0058] After determining the image slices whose similarity scores exceed the preset threshold, the corresponding installation point positions are marked in the 3D environment model. The markings include the positions of the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points, and control system installation points.

[0059] By combining the geometric information of the point cloud data, the position of the marked points is further corrected to ensure the accurate positioning of the installation points.

[0060] It should be noted that the installation point annotation process not only relies on the image feature matching results, but also combines the geometric accuracy of the point cloud data to ensure the accuracy of the position of the annotation point in three-dimensional space. For example, by mapping the two-dimensional image coordinates of the annotation point back to the three-dimensional point cloud coordinate system and using the depth information of the point cloud, the spatial accuracy of the annotation point can be further improved.

[0061] Furthermore, the marking system in this embodiment also supports manual review and correction functions, allowing engineers to verify and adjust the automatic marking results to further ensure the accuracy and reliability of the installation points.

[0062] It should be noted that the image recognition and annotation process can effectively identify and locate the installation points of various elevator components, reduce human errors, and improve installation accuracy and efficiency. The pre-trained ResNet-50 model has powerful image feature extraction capabilities through training on large-scale data sets, and can accurately identify target installation points in complex hoistway environments.

[0063] Furthermore, the method of this embodiment achieves high-precision installation point annotation by matching image features with predefined templates and combining the geometric information of the three-dimensional point cloud. This method not only improves the accuracy and reliability of annotation, but also adapts to the installation requirements of elevators of different types and specifications, and has strong versatility and scalability.

[0064] Step S4: Based on the marked points, a path planning algorithm is used to generate the installation sequence and installation path of the elevator guide rails to guide the installation of elevator components.

[0065] In step S4, the path planning algorithm used in the method of this embodiment includes determining the priority and dependency of each installation step according to the spatial relationship between the elevator guide rail fixing point, the car positioning point, the counterweight installation point, the door machine installation point and the control system installation point; establishing a directed acyclic graph between the installation steps, in which the nodes represent each installation step and the edges represent the dependency between the installation steps; applying a topological sorting algorithm to the established directed acyclic graph to generate the execution order of the installation steps.

[0066] Specifically, the path planning algorithm in this embodiment first determines which installation steps need to be completed before other steps based on spatial relationship analysis. For example, the fixing of the guide rail must be completed before the positioning of the car to ensure the stability of the guide rail and the accurate positioning of the car.

[0067] It should be noted that in the construction of the directed acyclic graph (DAG), each installation step is treated as a node, and the dependencies are represented by directed edges. For example, if step A (installing the guide rail) must be completed before step B (positioning the car), a directed edge is drawn from node A to node B in the graph, indicating that A is a prerequisite for B. In this way, the dependencies between the installation steps are clearly represented in the graph, ensuring that the installation process is carried out in a reasonable order.

[0068] Furthermore, in this embodiment, the Kahn algorithm is used for topological sorting, and a linear installation step execution order is generated by traversing the directed acyclic graph to ensure that all dependencies are satisfied. The topological sorting algorithm can effectively handle complex dependencies, generate the optimal installation order, and avoid conflicts or errors during the installation process. For example, if step A must precede step B and step C, and there is no dependency between step B and step C, the topological sorting algorithm may generate an execution order of A→B→C or A→C→B, which ensures the rationality of the installation steps.

[0069] In addition, the implementation of the path planning algorithm makes the installation process highly flexible and adaptable, and can generate the optimal installation path and sequence according to different installation requirements and site conditions, thereby improving installation efficiency and quality. By analyzing the spatial relationship and dependency between each installation step, the path planning algorithm can effectively avoid spatial conflicts and ensure the smooth progress of each installation step.

[0070] It should be noted that the path planning algorithm not only considers the direct dependencies between the installation steps, but also comprehensively considers the coordination of the installation steps in the spatial layout. For example, if the guide rail installation position is close to the car positioning point, the path planning algorithm will give priority to the guide rail installation to reduce the moving distance of the installation tools and equipment and improve the overall installation efficiency.

[0071] Furthermore, the path planning algorithm can dynamically adjust the order of installation steps and path planning according to the actual installation situation on site and real-time monitoring data to ensure the continuity and efficiency of the installation process. For example, if an installation step needs to be delayed due to equipment failure, the path planning algorithm will automatically adjust the order of subsequent installation steps to ensure that the installation process is not affected.

[0072] Guiding the installation of elevator components includes installing the elevator guide rails to the marked guide rail fixing points using fixing tools according to the installation sequence generated by the path planning; using the positioning device to move the car and install it to the marked car positioning points; installing the counterweight assembly and adjusting the counterweight position to the marked counterweight installation point; using the symmetrical installation tool to install the door machine to the marked door machine installation point; using the connection tool to install the control system to the marked control system installation point.

[0073] In guiding the installation of elevator components, the guide rails are installed according to the installation sequence generated by the path planning, and the elevator guide rails are firmly installed to the marked guide rail fixing points using bolt fixing tools. The bolt fixing tool is equipped with an automatic torque control function, which can ensure that the bolts are tightened according to the predetermined torque value, avoiding over-tightening or over-loosening, and ensuring the installation quality of the guide rails.

[0074] Specifically, the bolt fixing tool used in this embodiment monitors the tightening degree of the bolts in real time through automated control, ensuring that each bolt is evenly stressed, thereby improving the stability and safety of the guide rail installation.

[0075] It should be noted that the automatic torque control function of the bolt fixing tool can automatically adjust the torque parameters according to the material and specifications of the bolts to ensure that the bolts are tightened to the optimal state and prevent guide rail installation problems caused by loose or over-tightened bolts.

[0076] Furthermore, the bolt fixing tool in this embodiment is also equipped with an intelligent feedback system, which can record the tightening status of each bolt in real time and transmit the data to the central control system for subsequent quality inspection and installation report generation, ensuring the transparency and traceability of the guide rail installation process.

[0077] The car positioning uses an automatic positioning device to accurately move and install the car to the marked car positioning point. The automatic positioning device uses laser ranging and real-time feedback control technology to monitor the position and posture of the car in real time, ensuring that it is accurately aligned to the positioning point and reducing installation errors.

[0078] Specifically, the automatic positioning device in this embodiment adjusts the position and angle of the car through real-time comparison with the three-dimensional environment model to ensure the smooth operation of the car on the guide rails and avoid operation failures caused by inaccurate positioning.

[0079] It should be noted that the automatic positioning device has multi-axis rotation and movement capabilities, and can be precisely adjusted in different directions and angles to ensure the installation accuracy and smooth operation of the car.

[0080] Furthermore, the automatic positioning device is also equipped with a visual recognition system that can identify identification points and positioning marks on the guide rails, further improving the accuracy and reliability of car positioning. Through the combination of the visual system and laser ranging, the automatic positioning device can achieve multi-dimensional data fusion to ensure high accuracy and high reliability of car positioning.

[0081] When installing the counterweight assembly, use the balance adjustment device to adjust the counterweight position to the marked counterweight installation point. The balance adjustment device has an automatic adjustment function, which can automatically adjust the position of the counterweight according to real-time load changes, maintain the balance of the elevator, and reduce energy consumption and mechanical wear.

[0082] Specifically, the balance adjustment device in this embodiment monitors the load condition of the elevator through sensors, automatically calculates and adjusts the position of the counterweight, and ensures that the elevator can maintain balance under different load conditions, thereby improving operating efficiency and safety.

[0083] It should be noted that the balancing adjustment device can not only automatically adjust the position of the counterweight, but also has remote monitoring and control functions, allowing engineers to monitor the status and adjustment process of the counterweight in real time through the control center to ensure the accuracy and timeliness of the counterweight installation.

[0084] Furthermore, the balance adjustment device also uses an intelligent algorithm to optimize the counterweight adjustment strategy according to the elevator's operating data and load changes, further improving the elevator's energy efficiency and operating stability. The intelligent algorithm can predict future load change trends and adjust the counterweight position in advance to ensure that the elevator is always in the best balance state.

[0085] The symmetrical installation tool is used to install the door machine to the marked door machine installation point. The symmetrical installation tool is equipped with a precise positioning system to ensure the symmetrical installation of the door machine in the horizontal and vertical directions, avoiding deviations and failures during the operation of the door machine.

[0086] Specifically, the symmetrical installation tool achieves high-precision positioning and installation of the door machine through laser ranging and robotic arm control, ensuring the stability and switching flexibility of the door machine.

[0087] It should be noted that the symmetrical installation tool has automatic alignment and locking functions, which can automatically correct the position and angle of the door machine during the installation process to ensure its symmetry and operating accuracy.

[0088] Furthermore, the symmetrical installation tool is also equipped with a real-time monitoring system that can detect the installation status and symmetry of the gantry crane in real time. Through the feedback control mechanism, it automatically adjusts the installation parameters to ensure the perfect symmetry and efficient operation of the gantry crane.

[0089] Control system installation Use electrical connection tools to install the control system to the marked control system installation points. The electrical connection tools adopt modular design, support quick plug-in and efficient connection, reduce installation time and labor costs, and ensure the stability of the control system and the reliability of data transmission.

[0090] Specifically, the electrical connection tool used in this embodiment is equipped with an intelligent recognition system that can automatically identify the connector type and connection method of the control system, ensure the correctness and safety of the connection, and avoid system failures caused by connection errors.

[0091] It should be noted that the modular design of the electrical connection tool makes the installation process of the control system simpler and more efficient, supports various types and specifications of control system interfaces, and adapts to the installation requirements of control systems of different models and brands.

[0092] Furthermore, the electrical connection tool also has fault detection and alarm functions, which can monitor the status of the electrical connection in real time during the connection process, promptly detect and eliminate connection faults, and ensure the stable operation of the control system. Through the intelligent detection system, the electrical connection tool can automatically identify and correct connection errors, improving the safety and reliability of the installation process.

[0093] Furthermore, the method of this embodiment can also be combined with an automated installation robot to further improve the automation and installation accuracy of the installation process, reduce human errors, and improve overall installation efficiency. The automated installation robot is equipped with a multi-axis robotic arm and a precise motion control system, which can automatically complete the installation operation of each component according to a predetermined installation path and sequence.

[0094] It should be noted that the automated installation robot can efficiently and accurately complete the installation of elevator components through pre-programmed installation paths and action sequences, reduce the tediousness and errors of manual operations, and improve overall installation efficiency and quality.

[0095] Furthermore, combined with intelligent sensors and real-time monitoring systems, the automated installation robot can provide real-time feedback on the installation status and location, ensuring the controllability and traceability of the installation process and improving the intelligence and automation level of the installation process.

[0096] Guiding the installation of elevator components also includes real-time monitoring of the installation progress and comparing it with the preset installation progress. When the installation progress does not meet the preset progress, the node order and dependencies in the directed acyclic graph are adjusted according to the real-time data of the installation site, and the topological sorting algorithm is reapplied to update the execution order of the installation steps.

[0097] In this embodiment, the real-time monitoring of the installation progress uses Internet of Things (IoT) sensors and an installation progress tracking system to collect and analyze key data during the installation process in real time.

[0098] Specifically, installation progress sensors are installed at key installation steps to record the completion status and required time of each step in real time.

[0099] The installation progress sensor uses RFID technology and position sensors to accurately detect the start and completion time of the installation steps and record the actual time taken for each installation step. The sensor data is transmitted to the central monitoring system in real time via a wireless network to ensure the real-time and accuracy of the data.

[0100] Equipment status sensors monitor the operating status of equipment used during the installation process (such as fixing tools, positioning devices, etc.) to detect whether there are any faults or abnormalities. Equipment status sensors include temperature sensors, vibration sensors, and current sensors, which can monitor the operating status of equipment in real time, detect whether the equipment is overheating, abnormal vibration, or abnormal current, and prevent equipment failure. Sensor data is centrally managed through the Internet of Things platform and fed back to the monitoring system in real time, making it easy to promptly discover and handle equipment failures.

[0101] Environmental monitoring sensors monitor the environmental conditions at the installation site, such as temperature, humidity, and space restrictions, to ensure that the environmental conditions meet the installation requirements. Environmental monitoring sensors include temperature and humidity sensors and space measurement sensors, which can collect the temperature, humidity, and space dimensions of the installation site in real time to ensure that the environmental conditions are within a safe and controllable range. Sensor data is transmitted to the central monitoring system in real time through a wireless transmission module, facilitating real-time monitoring and adjustment of the installation environment.

[0102] It should be noted that the installation progress tracking system compares the installation progress with the preset installation progress of the project plan to evaluate in real time whether the installation progress meets expectations. When the actual installation time exceeds 10% of the preset time, equipment failure occurs, or the installation environment changes significantly, the path planning adjustment step will be automatically triggered.

[0103] Furthermore, in this embodiment, the path planning adjustment step includes, when the condition is met, the adjustment includes:

[0104] Progress delay detection: When it is detected that the actual installation progress delay exceeds the preset time threshold (such as 10%), the specific installation steps that are delayed will be identified and the sequence of steps that need to be adjusted will be marked;

[0105] Equipment fault detection: When the equipment status sensor detects that the installation equipment has a fault or anomaly, it will pause the relevant installation steps and rearrange the installation sequence to ensure the continuity and safety of the installation process;

[0106] Environmental change detection: When environmental monitoring sensors detect significant changes in the installation environment (such as a sharp rise in temperature, humidity changes beyond a safe range, and space limitations that hinder installation steps), the installation path and step sequence will be adjusted to ensure that the installation process is carried out in a safe and controllable environment.

[0107] Based on real-time data, the priorities and dependencies of each installation step will be re-evaluated and a new order of installation steps will be determined. For example, if the guide rail installation is delayed, the order of the car positioning step will be adjusted to accommodate the new installation schedule.

[0108] Based on the re-evaluated priorities and dependencies, the node order and edge dependencies in the directed acyclic graph will be adjusted to ensure the correctness and acyclicity of the graph structure.

[0109] Apply the Kahn algorithm to the adjusted directed acyclic graph to generate a new order of execution of the installation steps. The Kahn algorithm ensures the linear order of topological sorting by gradually deleting nodes without incoming edges, which is suitable for processing complex dependencies.

[0110] The generated new installation step execution sequence will be fed back to the installation guidance system in real time to guide the installers to perform the installation steps in the new sequence, ensuring the smooth progress of the installation process and improving overall efficiency.

[0111] It should be noted that the path planning adjustment step can ensure that the installation process remains flexible and adaptable under dynamically changing site conditions, respond to delays in the installation progress, equipment failures and environmental changes in a timely manner, and avoid interruptions and errors in the installation process.

[0112] Furthermore, the path planning adjustment mechanism in this embodiment realizes intelligent management of the installation process through automated data analysis and decision support. It can automatically identify problems based on real-time data, quickly adjust the installation path and step sequence, ensure the continuity and efficiency of the installation process, and significantly improve the installation quality and efficiency.

[0113] In an optional embodiment of the present invention, the original three-dimensional point cloud data Apply Gaussian filter To denoise, Represents three-dimensional Euclidean space and generates denoised point cloud data :

[0114] ;

[0115] in, represents the original 3D point cloud dataset, Represents a single point in the point cloud, located in the three-dimensional space coordinates, represents a single point in the denoised point cloud, Represents a Gaussian filter function, applying Gaussian noise At each point To remove noise, Represents the denoised point cloud dataset.

[0116] Filtering using a voxel grid For the denoised point cloud Perform sparse processing and set the voxel size to (e.g. 1 cubic centimeter), generating a sparse point cloud :

[0117] ;

[0118] in, represents the voxel grid filter function, which is used to sparse point cloud data. represents the voxel size parameter, which controls the degree of sparsification. Indicates The set of points within a voxel, Representing voxels The representative point in is obtained by calculating the average value of all points in the voxel.

[0119] The acquired image data Apply color correction, generating a color-corrected image :

[0120] ;

[0121] in, represents the original image data, represents the pixel coordinates of the image, Represents the pixel value of the image in the red, green, and blue channels, Represents the white balance correction factor, which is used for the red, green, and blue channels respectively. represents the color correction function, Represents image data after color correction.

[0122] Apply histogram equalization to the color corrected image Perform contrast enhancement to generate an enhanced image :

[0123] ;

[0124] in, represents the image enhancement function, HE represents the histogram equalization algorithm, which improves the contrast by adjusting the grayscale distribution of image pixels. Represents the enhanced image data.

[0125] From the preprocessed point cloud and images Extract ORB features from and :

[0126] ;

[0127] Among them, ORB represents the ORB feature extraction algorithm, Represents the set of ORB feature points in the point cloud data. Represents a set of ORB feature points in image data.

[0128] Create a feature point pair set :

[0129] ;

[0130] in, Represents a set of feature point pairs, including the correspondence between point cloud feature points and image feature points. Indicates the point cloud data ORB feature points, Represents the feature points in the image data and the point cloud The corresponding ORB feature points.

[0131] Define optimization goal to minimize geometric error and color error , solve the rigid transformation matrix :

[0132] ;

[0133] in, Represents a rigid transformation matrix, including a rotation matrix and translation vectors , Represents the rotation matrix, which is used to adjust the rotation angle of the point cloud. Represents the translation vector, which is used to adjust the position of the point cloud. Represents the first feature points, Represents the image data corresponding feature points, Represents the color error weight parameter, which controls the relative importance of geometric error and color error. represents the Euclidean distance, which is used to measure the geometric error. Represents a distance metric based on color information (such as Euclidean distance in RGB space), which is used to measure color error. Indicates that all To sum, is the index of the feature point.

[0134] The rigid transformation matrix Application to point cloud data , generate the aligned point cloud . Through the rotation matrix For the original sparse point cloud Rotate, plus translate vector Realize spatial alignment of point clouds to generate aligned point clouds :

[0135] ;

[0136] in, represents the aligned point cloud dataset, Represents the rotation matrix, which is used to adjust the rotation angle of the point cloud. Represents the translation vector, which is used to adjust the position of the point cloud. It represents the multiplication operation of matrix and vector, and the rotation operation; + represents the vector addition operation, and the translation operation.

[0137] Through multiple rounds of iterations, the point cloud registration process is optimized to gradually improve the alignment accuracy. Set the maximum number of iterations And the convergence condition , iteratively update the rigid transformation matrix :

[0138] ;

[0139] in, Indicates The rigid transformation matrix of the round iteration, Indicates The aligned point cloud dataset after rounds of iterations, represents an iterative closest point algorithm based on color information, which is used to optimize the rigid transformation matrix. Indicates The rotation matrix in the iteration is used to rotate the point cloud data from the current estimated coordinate system to the target coordinate system. Indicates the The translation vector in iteration , used to translate the point cloud data from the current estimated coordinate system to the target coordinate system.

[0140] The iteration stop condition is expressed as follows:

[0141] ;

[0142] in, Indicates the magnitude of change in the rigid transformation matrix, which is used to judge convergence. Indicates the current iteration number, represents the maximum number of iterations, Indicates the convergence condition value.

[0143] According to the progress of point cloud registration, the number of iterations and convergence conditions are dynamically adjusted to improve the adaptability and robustness of the algorithm:

[0144] ;

[0145] in, Indicates the current iteration number, represents the minimum number of iterations, represents the maximum number of iterations, represents the maximum convergence threshold, represents the high threshold, used for initial iterations, Represents the low threshold, used for subsequent iterations, Indicates the initial iteration number threshold.

[0146] The texture mapping process includes: Through the camera intrinsic matrix Project to the two-dimensional image plane and obtain the corresponding pixel coordinates From the image Extract the color information of the corresponding pixel , assigned to .

[0147] The color information fusion mechanism includes calculating the fusion weight and , dynamically adjust the fusion ratio of geometric information and color information. Color attributes with texture information Fusion, Spawn Point The final color attribute .

[0148] The point cloud generation with color information includes the fusion of color attributes Assign the corresponding points , generate a point cloud dataset with color information .

[0149] Specifically, first, define the color information fusion function , used to combine the geometric color attributes of the point cloud With texture color attributes extracted from the image , generating point The final color attribute . Weight parameter and Control the fusion ratio of geometric information and color information to ensure a balanced fusion effect.

[0150] ;

[0151] in, Indicate point The final color attribute, expressed as an RGB value , Represents the color information fusion function, which is used to combine geometric information and texture information to generate the final color of the point. Representing a point cloud dataset The points, with coordinates in three-dimensional space , Represents the enhanced image data, including the color and texture details of the image, Represents the fusion weight of geometric information, based on point The correspondence with the image pixels is dynamically adjusted. Represents the fusion weight of color information and satisfies , Represents the color attribute based on geometric information, which comes from the original color or preset color of the point cloud. Represents texture attributes based on color information by mapping texture information in the image to point cloud points On to get.

[0152] Define and implement the texture mapping function MapTexture, which is used to map the image Texture information in According to the point cloud Spatial location Mapping to color attributes This function maps texture information by projecting three-dimensional points onto a two-dimensional image plane and extracting the color values ​​of corresponding pixels.

[0153] ;

[0154] Among them, MapTexture represents the texture mapping function, which is used to convert the texture information in the image According to the point cloud Spatial location Mapping to color attributes superior, Represents the camera intrinsic parameter matrix, which is used to project three-dimensional points onto the two-dimensional image plane. Indicate point In the image The corresponding pixel coordinates in .

[0155] The three-dimensional point Through the camera intrinsic matrix Project to the two-dimensional image plane to get the normalized pixel coordinates Then, by rounding Convert to actual integer pixel coordinates , and perform boundary processing to ensure that the pixel coordinates are within the image range;

[0156] ;

[0157] ;

[0158] in, Indicate point The coordinates in three-dimensional space, represents the camera intrinsic parameter matrix, represents the focal length of the camera in pixels, represents the principal point coordinates of the image (pixels), Represents the normalized pixel coordinates, without rounding. Round represents the rounding operation, converting the floating-point coordinates to the nearest integer pixel coordinates. Indicates the width and height of the image (in pixels), Indicate point In the image The final pixel coordinates in , after boundary processing.

[0159] Point projection representation using the camera intrinsic matrix The three-dimensional point Project to the two-dimensional image plane to get the normalized pixel coordinates ; Coordinate conversion is done by rounding Convert to actual pixel coordinates ; Boundary processing ensures is within the valid pixel range of the image, that is and .

[0160] According to the determined pixel coordinates , from the enhanced image Extract the color value of the corresponding pixel , as a point Color information ;

[0161] ;

[0162] in, Indicate point The color information is extracted from the image through the texture mapping function MapTexture. Represents the enhanced image In pixel coordinates The color value at , expressed as RGB value .

[0163] According to the final pixel coordinates From the image Extract the corresponding color value , assigned to .

[0164] Calculate fusion weight parameters and , controls the fusion ratio of geometric information and color information. Weight Based on point With image pixels The spatial distance measure between and weight adjustment parameters Dynamically adjust to balance the contribution of geometric and color information;

[0165] ;

[0166] in, Represents the fusion weight of geometric information, based on point The correspondence with the image pixels is dynamically adjusted to ensure the balance between geometric information and color information. Represents the fusion weight of color information and satisfies , Represents the weight adjustment parameter, which is used to control the weight and The rate of change of is usually a positive real number, Represents spatial distance measurement, reflecting point cloud points With image pixels The smaller the correspondence accuracy between Indicates a better correspondence, the larger Indicates poor correspondence, measured by spatial distance and weight adjustment parameters , dynamically calculate weights and , in order to balance the fusion ratio of geometric information and color information.

[0167] Combined with the calculated weight and , the geometric information color attribute Color attributes with texture information Fusion, calculate the point The final color attribute Then, all the fused points Composing a new point cloud dataset , to achieve comprehensive integration of color information;

[0168] ;

[0169] in, Indicate point The final color attribute combines geometric information and texture information. Represents the fused point cloud dataset with color information, Representing a point cloud dataset The points, with coordinates in three-dimensional space , Indicate point The final color attribute, expressed as an RGB value , Represents the fusion weight parameter, which controls the fusion ratio of geometric information and color information and satisfies 1, Represents the color attribute based on geometric information, which comes from the original color or preset color of the point cloud. Represents the enhanced image The texture color attributes extracted from .

[0170] Using the calculated weights and , the geometric information color attribute Color attributes with texture information Fusion, get the point The final color attribute . The merged color attributes Assign the corresponding points , generate a point cloud dataset with color information , to achieve comprehensive integration of color information.

[0171] Embodiment 2 is an embodiment of the present invention, which provides an elevator installation guidance system, including:

[0172] A data acquisition module is used to scan the internal structure of the elevator shaft through a laser rangefinder to obtain three-dimensional point cloud data, and to synchronously collect image data of the inner surface of the shaft through a structured light camera;

[0173] Data fusion module, used to fuse laser ranging data with image data to generate a three-dimensional environmental model including shaft dimensions, structural layout and environmental characteristics;

[0174] Image recognition module, used to apply image recognition algorithm to analyze the generated 3D environment model, identify and mark the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points;

[0175] The installation guidance module is used to generate the installation sequence and installation path of the elevator guide rails based on the marked points and use the path planning algorithm to guide the installation of elevator components.

[0176] Embodiment 3, an embodiment of the present invention, is different from the first two embodiments in that:

[0177] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0178] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0179] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0180] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0181] Example 4 is an embodiment of the present invention, which provides an elevator installation guidance method. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0182] This experiment simulated 5 different types of elevator shaft environments in the laboratory. The experiment included straight shafts, standard curved shafts, and shafts with complex geometries to ensure that the test conditions covered applications in different actual installation environments.

[0183] The shaft width is 2.5 meters for both straight and curved shafts, and the turning radius of the curved shaft is 5 meters; the complex geometric shaft has multiple corners, obstacles and narrow sections. The environment of mixed natural light and artificial lighting is simulated, and the influence of changing lighting conditions (for example, darker light or uneven light in the depth of the shaft) is ensured. Fixing tools, positioning devices, and symmetrical installation tools are used to simulate the real tools used in the installation process.

[0184] The traditional method uses manual measurement and manual calculation, combined with existing simple path planning tools to guide the installation of the shaft. First, the installer manually measures the basic dimensions of the shaft and records the length, width and turning position of the shaft. Then, based on the measurement results, the installation path of the elevator guide rail is manually drawn. Then, the installation sequence is manually set and each installation point is marked. Next, the installer begins to gradually install the various components of the elevator according to the manually calculated installation path.

[0185] The method of the present invention adopts a combination of automated measurement and intelligent path planning. The elevator shaft is scanned by a laser rangefinder and a structured light camera to obtain three-dimensional point cloud data and fuse it with image data. First, the laser rangefinder scans the inner wall of the shaft, and the structured light camera synchronously captures the image data of the shaft surface. Then, through data fusion, a three-dimensional model containing the shaft size, structural layout and environmental characteristics is generated. Then, the image recognition algorithm is used to parse the three-dimensional model, and key installation points such as elevator guide rail fixing points, car positioning points, and counterweight installation points are automatically identified and marked. Next, based on the marked points, the path planning algorithm is used to generate the installation order and path of the elevator guide rails, and the installation order is adjusted in real time according to the actual situation on site. The experimental results are shown in Table 1.

[0186] Table 1 Comparison of experimental results

[0187] Test scenario Straight shaft Standard curve Complex geometry Curve + Obstacle Narrow paths and obstacles Traditional installation time (hours) 10 12 18 15 14 Installation time of the method of the present invention (hours) 6 7 10 9 8 Accuracy difference (mm) 8 10 12 11 9 Installation Error Rate 5% 8% 12% 10% 9% Installation order optimization 60% 70% 80% 75% 85% On-site adjustment response time (minutes) 30 35 40 38 32

[0188] The method of the present invention obtains three-dimensional point cloud and image data by combining a laser rangefinder with a structured light camera, automatically generates an accurate three-dimensional environmental model, and greatly improves the accuracy of measurement. Especially when there are complex geometric shapes or obstacles in the shaft, it can achieve more accurate installation path planning than manual methods. Traditional methods mainly rely on manual judgment in path planning, and it is difficult to ensure the rationality of the optimization and installation sequence of each step. Even if a simple path planning tool is used, its optimization effect is relatively limited, and it is often impossible to adjust the installation sequence in real time to cope with changes in field conditions. In contrast, the present invention uses a path planning algorithm based on the spatial relationship between each installation point in the shaft to automatically generate the optimal installation path, and adjust the path in real time to adapt to changes in the installation process. This solution not only effectively improves the installation efficiency, but also significantly reduces the errors and time waste caused by manual judgment.

[0189] Especially in complex elevator shaft scenarios, traditional methods often have difficulty dealing with shaft geometry changes and space limitations, resulting in improper path planning, unreasonable installation sequence, and misplaced installation components during the installation process, which in turn leads to low installation efficiency and a high error rate. Through the automated path planning and real-time adjustment mechanism of the present invention, possible risks can be predicted in the early stages of installation, and the installation sequence and path can be dynamically adjusted according to the actual situation on site, thereby avoiding common problems of blocked paths and space obstructions, and improving the overall efficiency and accuracy of elevator installation.

[0190] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. An elevator installation guidance method, characterized in that: include: The internal structure of the elevator shaft is scanned by a laser rangefinder to obtain 3D point cloud data, and the image data of the inner surface of the shaft is collected synchronously by a structured light camera; The laser ranging data and image data are fused to generate a three-dimensional environmental model including shaft dimensions, structural layout and environmental characteristics; Apply image recognition algorithms to analyze the generated 3D environment model, identify and mark the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points; Based on the marked points, the path planning algorithm is used to generate the installation sequence and installation path of the elevator guide rails to guide the installation of elevator components; The data fusion comprises: Performing a first preprocessing step on the three-dimensional point cloud data, wherein the first preprocessing step includes denoising and point cloud thinning; Performing a second preprocessing step on the image data acquired by the structured light camera, wherein the second preprocessing step includes color correction and image enhancement; Use the point cloud registration algorithm to spatially align the preprocessed 3D point cloud data with the image data; By fusing color information, the texture information in the image data is mapped to the 3D point cloud data; The point cloud registration algorithm includes: Use feature matching technology to extract key feature points from 3D point cloud data and image data; Applying the iterative closest point algorithm to optimize the registration of point clouds by minimizing the distance error between point clouds; Calculate the rigid transformation matrix, rotate and translate the 3D point cloud data according to the results of the iterative closest point algorithm, and align the image data; The analyzing of the generated three-dimensional environment model comprises: Segment the generated three-dimensional environment model into a plurality of two-dimensional image slices, each slice corresponding to a specific area in the well; Apply a pre-trained convolutional neural network model to each 2D image slice to extract image feature vectors; Match the extracted image feature vector with the predefined installation point feature template and calculate the similarity score; The image slices whose similarity scores exceed the preset threshold are determined, and the positions of the elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points are marked in the 3D environment model.

2. The elevator installation guidance method according to claim 1, characterized in that: The application of the path planning algorithm includes: Determine the priority and dependency of each installation step based on the spatial relationship between the elevator guide rail fixing point, car positioning point, counterweight installation point, door machine installation point and control system installation point; Create a directed acyclic graph between the installation steps, where nodes represent each installation step and edges represent dependencies between the installation steps; A topological sorting algorithm is applied to the established directed acyclic graph to generate the execution order of the installation steps.

3. The elevator installation guidance method according to claim 2, characterized in that: The guidance of elevator component installation includes: According to the installation sequence generated by the path planning, use the fixing tools to install the elevator guide rails to the marked guide rail fixing points; Use the positioning device to move the car and install it to the marked car positioning point; Install the counterweight assembly and adjust the counterweight position to the marked counterweight installation point; Use the symmetrical installation tool to install the door operator to the marked door operator installation point; Use the connection tool to install the control system to the marked control system installation points.

4. The elevator installation guidance method according to claim 3, characterized in that: The guiding the installation of elevator components also includes: real-time monitoring of the installation progress, comparing it with the preset installation progress, and when the installation progress does not meet the preset progress, adjusting the node order and dependency relationship in the directed acyclic graph according to the real-time data of the installation site, and reapplying the topological sorting algorithm to update the execution order of the installation steps.

5. A system using the elevator installation guidance method according to any one of claims 1 to 4, characterized in that: include: A data acquisition module is used to scan the internal structure of the elevator shaft through a laser rangefinder to obtain three-dimensional point cloud data, and to synchronously collect image data of the inner surface of the shaft through a structured light camera; Data fusion module, used to fuse laser ranging data with image data to generate a three-dimensional environmental model including shaft dimensions, structural layout and environmental characteristics; An image recognition module, for applying an image recognition algorithm to analyze the generated three-dimensional environment model, identifying and marking elevator guide rail fixing points, car positioning points, counterweight installation points, door machine installation points and control system installation points; and; The installation guidance module is used to generate the installation sequence and installation path of the elevator guide rails based on the marked points and use the path planning algorithm to guide the installation of elevator components.

Images