Elevator guide rail installation monitoring system and method based on image processing
Through the elevator guide rail installation monitoring system based on image processing, using three-dimensional point cloud data, neural network, ant colony algorithm and other technologies, the problem of low detection accuracy of elevator shafts is solved, and high-precision and high-quality monitoring of elevator installation are achieved.
Patent Information
- Application Number
- CN202411141748.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-08-20
AI Technical Summary
The existing elevator shaft detection methods have low manual measurement accuracy, are prone to errors, time-consuming, and can only be sampled and inspected, and cannot be comprehensively measured, resulting in poor elevator installation accuracy control and installation process monitoring.
The elevator guide rail installation monitoring system is adopted based on image processing, and the three-dimensional point cloud data of the shaft is obtained through the suspended measurement device, a three-dimensional point cloud model is constructed, and the neural network is used to determine the opening position of the layer door, and a virtual reference line is generated based on the ant colony algorithm to transfer the elevator guide rail. The scaffold is fitted into the model, and the installation positioning information is optimized and analyzed, and the installation monitoring and correction is carried out through the monitoring image.
It has achieved comprehensive optimization of elevator shafts from measurement, modeling to installation positioning and process monitoring, improved measurement accuracy and installation accuracy, and ensured the high quality and reliability of elevator installation.
Smart Images

Figure CN118675113B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of engineering technology, and in particular to an elevator guide rail installation monitoring system and method based on image processing. Background Art
[0002] During elevator installation and construction, the elevator shaft requires extremely high precision and must be kept within a tight geometric tolerance range to ensure normal installation of the elevator. The shaft must be straight to prevent the elevator from swinging, the floor door openings must be aligned with each other, and the cross-section must also meet the tolerance to ensure assembly fit. However, the current common shaft inspection method is for the elevator company to use a plumb line to perform rough manual measurements in the shaft, measuring the distance between each floor wall and floor door relative to the plumb line. This method is not only prone to errors, but also extremely time-consuming. It can only be sampled for inspection and cannot achieve comprehensive measurement, resulting in poor installation results and cannot cope well with complex and changing installation requirements.
[0003] At present, there are technical problems in the relevant technologies, such as low manual measurement accuracy, easy errors, time-consuming and only sampling inspection in elevator shaft inspection leads to inability to conduct comprehensive measurement, which in turn affects the elevator installation precision control and installation process monitoring effect. Summary of the invention
[0004] The present application provides an elevator guide rail installation monitoring system and method based on image processing, which uses a suspended measuring device to obtain shaft three-dimensional point cloud data, constructs a three-dimensional point cloud model, determines the elevator door opening position through a neural network model, generates a virtual reference line based on this using an ant colony algorithm, fits the elevator guide rail and bracket into the model to determine the installation positioning information, and finally performs installation monitoring correction through monitoring images based on the installation positioning information and the virtual reference line, thereby achieving a comprehensive optimization of the elevator shaft from measurement, modeling to installation positioning and process monitoring, thereby achieving the technical effect of improving the measurement accuracy and installation accuracy of the elevator shaft and the monitoring effect of the installation process, and ensuring the high quality and reliability of the elevator installation.
[0005] The present application provides an elevator guide rail installation monitoring system based on image processing, including:
[0006] A three-dimensional point cloud data acquisition module, the three-dimensional point cloud data acquisition module is used to use a suspended measuring device equipped with a laser radar, an angle sensor, and a laser positioning device to uniformly lower the guide rail to scan the shaft and obtain three-dimensional point cloud data of the internal structure of the shaft; a three-dimensional point cloud model component module, the three-dimensional point cloud model component module is used to construct a three-dimensional point cloud model based on the three-dimensional point cloud data of the internal structure of the shaft through point cloud data processing and identification; a position determination module, the position determination module is used to analyze and identify the three-dimensional point cloud model through a trained neural network model to determine the opening position of the elevator floor door; a virtual reference line generation module, the virtual reference line generation module is used to generate a virtual reference line according to the opening position of the elevator floor door using an ant colony algorithm; an installation positioning information determination module, the installation positioning information determination module is used to fit the elevator guide rail and the guide rail fixing bracket into the three-dimensional point cloud model, optimize and analyze based on the virtual reference line, and determine the installation positioning information; a monitoring and correction module, the monitoring and correction module is used to perform installation monitoring and correction through monitoring and acquisition images based on the installation positioning information and the virtual reference line.
[0007] The present application provides an elevator guide rail installation monitoring method based on image processing, comprising:
[0008] A suspended measuring device is used, equipped with a laser radar, an angle sensor, and a laser positioning device, and is lowered at a uniform speed along the guide rail to scan the shaft, so as to obtain three-dimensional point cloud data of the internal structure of the shaft; a three-dimensional point cloud model is constructed through point cloud data processing and recognition based on the three-dimensional point cloud data of the internal structure of the shaft; the three-dimensional point cloud model is analyzed and recognized through a trained neural network model to determine the opening position of the elevator door; a virtual reference line is generated based on the opening position of the elevator door using an ant colony algorithm; the elevator guide rail and the guide rail fixing bracket are fitted into the three-dimensional point cloud model, and optimization analysis is performed based on the virtual reference line to determine the installation positioning information; based on the installation positioning information and the virtual reference line, installation monitoring and correction are performed through monitoring and acquisition images.
[0009] The elevator guide rail installation monitoring system and method based on image processing proposed in this application first uses a suspended measuring device to obtain the three-dimensional point cloud data of the shaft, constructs a three-dimensional point cloud model, and determines the opening position of the elevator door through a neural network model. Based on this, an ant colony algorithm is used to generate a virtual reference line, and the elevator guide rails and brackets are fitted into the model to determine the installation positioning information. Finally, based on the installation positioning information and the virtual reference line, installation monitoring correction is performed through monitoring images, thereby achieving a comprehensive optimization of the elevator shaft from measurement, modeling to installation positioning and process monitoring, thereby achieving the technical effect of improving the measurement accuracy, installation accuracy of the elevator shaft and the monitoring effect of the installation process, and ensuring the high quality and reliability of the elevator installation. BRIEF DESCRIPTION OF THE DRAWINGS
[0010] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the system according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0011] Figure 1 A schematic diagram of the structure of an elevator guide rail installation monitoring system based on image processing provided in an embodiment of the present application;
[0012] Figure 2 A flow chart of an elevator guide rail installation monitoring method based on image processing provided in an embodiment of the present application.
[0013] Explanation of the accompanying drawings: three-dimensional point cloud data acquisition module 10, three-dimensional point cloud model component module 20, position determination module 30, virtual reference line generation module 40, installation positioning information determination module 50, monitoring and correction module 60. DETAILED DESCRIPTION
[0014] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0015] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0016] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.
[0017] The present application embodiment provides an elevator guide rail installation monitoring system based on image processing, such as Figure 1 As shown, the method includes:
[0018] The three-dimensional point cloud data acquisition module 10 is used to use a suspended measuring device equipped with a laser radar, an angle sensor, and a laser positioning device to scan the well at a uniform speed along the guide rail to obtain the three-dimensional point cloud data of the internal structure of the well. Specifically, the three-dimensional point cloud data acquisition module 10 performs a comprehensive scanning operation on the well, identifies various structural features inside the well, and uses laser scanning and sensing technology to obtain key information such as the distance and angle inside the well. Based on the results of scanning and sensing, the spatial structure information of the well is determined. The spatial structure information includes information such as the shape, size, and internal obstacle distribution of the well, which is used for subsequent point cloud data processing. The data features of the obtained well spatial structure information are extracted and classified. The data features refer to key factors that affect the accuracy and integrity of point cloud data acquisition, such as scanning speed, sensor accuracy, environmental interference, etc. Through the classification algorithm, areas with similar data features are classified into one category, and corresponding acquisition parameters and strategies are configured according to the classification results and the characteristics of the well.
[0019] In a possible implementation, the three-dimensional point cloud data acquisition module 10 includes: a position correction unit, which is used to establish an absolute coordinate system through the laser positioning device, monitor the verticality of the device in real time, and perform position correction based on the verticality. Specifically, the laser positioning device is started to establish an absolute coordinate system, and the coordinate system provides a unified and accurate position reference frame for the entire measuring device. During the process of lowering the device along the guide rail, the laser positioning device monitors the verticality of the device in real time, and determines the verticality by measuring the inclination angle of the device relative to the absolute coordinate system. Once a deviation in verticality is detected, the position correction unit will calculate the required position correction amount according to the degree and direction of the deviation, and adjust the position of the device through the corresponding control mechanism to ensure that the collected data has high accuracy and reliability. For example, if the device tilts to the left at a certain angle, the position correction unit will calculate the correction amount to the right and achieve position correction by controlling the movement of the device.
[0020] The collaborative acquisition unit is used to collect three-dimensional point cloud data of the internal structure of the shaft, including the shaft wall, the floor door opening, and the steel bars, through the collaborative work of the laser radar and the angle sensor. Specifically, the laser radar and the angle sensor work together, the laser radar continuously emits a laser beam into the shaft, and receives the reflected light signal, thereby measuring the distance and angle information between the object inside the shaft and the sensor, and the angle sensor simultaneously measures the posture and angle change of the device itself. In the process of collaborative work, the three-dimensional point cloud data of the internal structure of the shaft can be fully collected, including the shape and position of the shaft wall, the size and position of the floor door opening, and the distribution of steel bars and other key information. Through collaborative collection, rich and comprehensive shaft structure data is obtained, which provides a basis for subsequent processing and analysis. For example, when the laser radar scans the shaft wall, the angle sensor will record the angle of the device at this time, thereby determining the exact position of each point on the shaft wall.
[0021] The preprocessing unit is used to perform noise reduction and abnormal point removal preprocessing on the collected three-dimensional point cloud data, and based on the classification of the shaft wall, floor door opening, and steel bar, the point cloud data is segmented into different categories using a segmentation algorithm to obtain the three-dimensional point cloud data of the internal structure of the shaft. Specifically, the original three-dimensional point cloud data obtained by the collaborative acquisition unit is subjected to noise reduction processing, which is achieved through various filtering algorithms, and invalid or erroneous data points generated by environmental interference, equipment noise and other factors are removed, and abnormal point removal operations are performed to identify and eliminate those data points that are obviously deviated from the normal range to improve the quality of the data. Different structural elements such as shaft walls, floor door openings, and steel bars are classified based on classification, and segmentation algorithms are used to divide the point cloud data into different categories according to its characteristics and distribution. For example, according to the distance and angle characteristics, the data points belonging to the shaft wall are classified into one category, the data points of the floor door opening are classified into another category, and the data points of the steel bars are classified into another category. Finally, the classified and processed three-dimensional point cloud data of the internal structure of the shaft are obtained, which is ready for subsequent modeling and analysis.
[0022] The three-dimensional point cloud model component module 20 is used to construct a three-dimensional point cloud model based on the three-dimensional point cloud data of the internal structure of the shaft through point cloud data processing and identification. Specifically, the three-dimensional point cloud model component module 20 performs in-depth analysis and processing on the three-dimensional point cloud data of the internal structure of the shaft, identifies key information such as the shape, size, and internal obstacle distribution of the shaft, extracts the various components of the shaft by using point cloud recognition and processing technology, and analyzes the spatial relationship and connection method between them. Based on the results of the analysis and processing, a three-dimensional point cloud model is constructed, and the three-dimensional point cloud model includes the shape and position information of elements such as the shaft wall, floor door opening, and steel bars for subsequent analysis and application. The constructed three-dimensional point cloud model is subjected to feature extraction and classification, and the feature refers to the key factors that affect the accuracy and usability of the model, such as point cloud density, data integrity, model accuracy, etc. Through the classification algorithm, model elements with similar features are classified into one category, and the corresponding model parameters and structures are optimized according to the classification results and the characteristics of the model.
[0023] Position determination module 30, the position determination module 30 is used to analyze and identify the three-dimensional point cloud model through the trained neural network model to determine the elevator floor door opening position. Specifically, the three-dimensional point cloud model data containing the elevator floor door opening position is collected as a training sample. The sample should cover elevator shafts of various types and structures. The training data is preprocessed, including data cleaning, normalization and other operations to improve the data quality and the efficiency of model training. The structure of the neural network is designed, including determining the number of layers, the number of neurons, the activation function, etc. The preprocessed training data is input into the neural network, and the output result is calculated through forward propagation. According to the difference between the output result and the actual elevator floor door opening position label, the gradient is calculated using the back propagation algorithm, and the weights and biases of the network are updated to continuously optimize the parameters of the model. After multiple iterations of training, when the performance of the model on the validation set reaches When a certain accuracy requirement is reached or convergence is achieved, the training is completed, the constructed 3D point cloud model is obtained, and the 3D point cloud model is input into the trained neural network model. The neural network model extracts and analyzes the input 3D point cloud model layer by layer. The model will comprehensively consider the distribution, shape, density and other characteristics of the point cloud, and gradually narrow the range of possible elevator door opening positions through calculation, and finally output the determined elevator door opening position. For example, in a specific application scenario, the model may first notice that the density and distribution pattern of the point cloud in a certain area are similar to the characteristics of the elevator door opening position in the training data, and then further analyze the situation of the surrounding point clouds to finally accurately determine the position of the elevator door opening.
[0024] The virtual reference line generation module 40 is used to generate a virtual reference line according to the elevator floor door opening position by using an ant colony algorithm. Specifically, the virtual reference line generation module 40 analyzes the elevator floor door opening position, identifies key information such as the spatial coordinates of the position, the surrounding environment characteristics, and generates a virtual reference line using an ant colony algorithm. In the initial stage of the ant colony algorithm, some initial path exploration points are randomly generated near the elevator floor door opening position. Each path exploration point selects the next path point according to a certain probability. The probability is affected by the pheromone concentration on the path and the heuristic information. The pheromone concentration reflects the selection frequency of the previous path exploration point on the path. The higher the concentration, the greater the probability of being selected. The heuristic information guides the path exploration point in a direction that is more likely to generate a valid reference line based on factors such as distance and direction. As the path exploration point continuously explores the path, the pheromone on the path will be updated according to certain rules. The pheromone concentration of the path selected by more path exploration points will increase; while the pheromone concentration of the path that is less selected will gradually decrease. After multiple iterations, the path exploration points will gradually concentrate on some better paths, forming an obvious path trend. By integrating and optimizing these better paths, a continuous, smooth and satisfactory virtual reference line is generated. For example, assuming that the elevator floor door opening position is in a more complex shaft environment. Initially, the path exploration points may randomly explore in various directions. As the algorithm progresses, more path exploration points will be concentrated on paths with shorter distances and more reasonable directions. Ultimately, the generated virtual reference line can avoid obstacles and effectively connect the elevator door opening position with other key positions, providing an accurate reference for subsequent installation work.
[0025] In a possible implementation, the virtual reference line generation module 40 includes: an initial point determination unit, the initial point determination unit is used to take the elevator floor door opening position as the starting point. Specifically, the accurate coordinates of the elevator floor door opening position are obtained, and the position is clearly identified as the starting point of the subsequent path search, which provides a clear starting point for the path search of the entire ant colony algorithm, ensuring that the generation of the path is based on the association and relative relationship with the elevator floor door opening position.
[0026] The pheromone matrix initialization unit is used to initialize the pheromone matrix. The pheromone concentrations on all paths are the same. Starting from the starting point, the path search is performed through the ant colony algorithm to obtain an updated pheromone matrix. Specifically, a pheromone matrix is established to represent the pheromone concentrations on each possible path. Initially, the pheromone concentrations on all paths are set to the same value, so that the path exploration points have no prior preference when selecting a path. It is completely based on random exploration and heuristic information. Starting from a determined starting point, the path exploration points are allowed to perform path search through the ant colony algorithm. In each search, the path exploration point selects the next node according to a certain probability, and the probability is jointly affected by the pheromone concentration and heuristic information on the current path. As the path exploration points continue to search, the pheromone matrix is continuously updated, reflecting the frequency and effect of different paths being explored.
[0027] An iterative convergence unit is used to iteratively search until the pheromone converges based on the updated pheromone matrix, connect the paths with the largest pheromone concentration, and perform smoothing to obtain the virtual reference line. Specifically, based on the continuously updated pheromone matrix, iterative search is continuously performed. In each iteration, the path exploration point selects a path according to the latest pheromone concentration and heuristic information. As the iteration proceeds, the pheromone concentration gradually changes. The pheromone concentration on some paths increases, while the concentration on other paths decreases. When the pheromone distribution reaches a relatively stable state, that is, convergence, it is considered that the search has reached a relatively optimal result. The path with the largest pheromone concentration is found and connected. In order to make the generated path smoother and meet actual needs, the connected paths are smoothed to finally obtain the optimal path that can be used as a virtual reference line. For example, assuming that the elevator door opening position is in a complex building structure, the initial point determination unit clarifies the starting coordinates. The pheromone matrix initialization unit provides a fair starting point for path search, and the path exploration point begins random exploration. After multiple iterations, the iterative convergence unit finds the optimal path and, after smoothing, generates a virtual reference line that can accurately guide subsequent work.
[0028] In a possible implementation, the pheromone matrix initialization unit includes: an objective function setting subunit, the objective function setting subunit is used to set the objective function with the goal of minimizing the path length of the virtual reference line and with the goal of maximizing the degree of decoupling between the guide rail and the bracket being independent of each other. Specifically, the objective function setting subunit defines two key goals. First, the path length of the virtual reference line is minimized to reduce resource consumption and improve efficiency. Second, the maximum degree of decoupling between the guide rail and the bracket being independent of each other is achieved to ensure the stability and reliability of the system. Based on the goal, the objective function is set, and the objective function is obtained by adding the product of the path length and the path length configuration weight to the product of the decoupling degree and the decoupling degree configuration weight.
[0029] The target weight configuration subunit is used to configure the target weights of minimizing the path length and maximizing the degree of decoupling according to the requirements of the elevator guide rail installation scenario. Specifically, the specific requirements of the elevator guide rail installation scenario are analyzed in depth. If the installation space is relatively narrow, more attention may be paid to minimizing the path length, and a higher path length configuration weight may be configured for it. If the installation scenario has extremely high requirements for the independence of the guide rail and the bracket, such as in a complex mechanical environment, a larger decoupling degree configuration weight will be configured for the degree of decoupling. Through careful study and weighing of the scenario, the target weight that can accurately reflect the actual needs is determined.
[0030] The weight fitting subunit is used to fit the target weight to the objective function. Specifically, the path length configuration weight and the decoupling degree configuration weight determined by the target weight configuration subunit are substituted into the objective function, so that the objective function can accurately reflect the relative importance of the path length and the decoupling degree in the current scene, and provide an accurate mathematical expression for subsequent optimization calculations.
[0031] The heuristic function component subunit is used to combine the path length and the decoupling degree information to generate comprehensive heuristic information, and construct a heuristic function based on the comprehensive heuristic information and the target weight. Specifically, the relevant information of the path length and the decoupling degree is integrated to generate comprehensive heuristic information that can reflect the quality of the path, and the heuristic function is constructed based on the heuristic information and the target weight configured previously. The heuristic function can help the ant colony algorithm to be more directional and efficient when searching for paths, and find a better path faster.
[0032] A path search subunit, the path search subunit is used to perform path search based on the objective function and the heuristic function through the ant colony algorithm. Specifically, based on the constructed objective function and heuristic function, the ant colony algorithm is used to perform path search. During the search process, each path search point selects the next node according to the information provided by the heuristic function. At the same time, the objective function is used to evaluate the quality of the generated path. Through continuous iteration and group collaboration, the ant colony algorithm can gradually find the optimal path that satisfies the goals of minimizing the path length and maximizing the degree of decoupling. For example, assuming that in a specific elevator guide rail installation scenario, the space is limited and the mechanical environment is complex, the objective function setting subunit sets the objective function, the objective weight configuration subunit configures the weight of 0.6 for the path length and the weight of 0.4 for the degree of decoupling, and the weight fitting subunit substitutes these weights into the objective function. The heuristic function component subunit calculates the path length, constructs the decoupling degree expression, and synthesizes the heuristic function. The path search subunit uses the ant colony algorithm, and after multiple iterations, it finally finds a path that is relatively short and can ensure a high degree of decoupling between the guide rail and the bracket.
[0033] In a possible implementation, the target weight configuration subunit includes: an installation index determination microunit, which is used to determine the installation progress and track the installation effect index by monitoring and collecting real-time images and data during installation. Specifically, during the installation of the elevator guide rail, the monitoring equipment at the installation site continuously collects real-time images and related data, and analyzes and processes the images and data to accurately determine the current installation progress, such as the proportion of the completed installation workload to the total workload, and track various indicators of the installation effect, such as the installation flatness of the guide rail, the firmness of the bracket, etc.
[0034] The installation cycle node setting micro-unit is used to set the installation cycle nodes, and establish the mapping association between the installation progress and the tracking installation effect indicators and the installation cycle. Specifically, according to the overall plan and requirements for the installation of elevator guide rails, clear installation cycle nodes are set. The nodes can be stage marks divided by time, or they can be key milestones set according to the degree of completion of the installation workload. At the same time, the mapping association between the installation progress and the tracking installation effect indicators and the installation cycle is established to clarify the progress and effect standards that should be achieved at each installation cycle node.
[0035] The installation evaluation micro-unit is used to perform installation evaluation according to the installation progress and tracking installation effect indicators, calculate the matching of the installation cycle node targets, and determine the cycle evaluation value, which includes the installation efficiency evaluation value and the installation accuracy evaluation value. Specifically, the installation progress and installation effect indicator data obtained by the micro-unit are determined according to the installation indicators, and the current installation work is comprehensively evaluated. The actual installation situation is matched with the target set by the installation cycle node to determine the cycle evaluation value. The cycle evaluation value includes the installation efficiency evaluation value and the installation accuracy evaluation value. The installation efficiency evaluation value reflects the speed and progress of the installation work, and the installation accuracy evaluation value reflects the quality and accuracy of the installation.
[0036] The feedback mechanism setting micro-unit is used to fit the adjustment relationship between the path length, the decoupling degree weight and the periodic evaluation value according to the historical case data, and set the adaptive feedback mechanism. Specifically, the data in the historical cases are collected and analyzed, including the path length, the decoupling degree weight and the corresponding periodic evaluation value in different installation projects, and the internal relationship and law between these variables are found through the data fitting method. Based on the analysis, an adaptive feedback mechanism is set, so that when the periodic evaluation value changes, the path length and the decoupling degree weight can be adjusted accordingly.
[0037] The adaptive adjustment micro-unit is used to match the installation efficiency evaluation value and the installation accuracy evaluation value with the adaptive feedback mechanism, determine the target weight adjustment strategy, and perform weight adaptive adjustment on the path length and decoupling degree, wherein the path length adjustment weight corresponds to the installation efficiency evaluation value, and the decoupling degree corresponds to the installation accuracy evaluation value. Specifically, the installation efficiency evaluation value and the installation accuracy evaluation value obtained by the installation evaluation micro-unit are matched with the adaptive feedback mechanism set by the feedback mechanism setting micro-unit. If the installation progress is fast but the accuracy is low, it means that more attention needs to be paid to the installation accuracy. At this time, the weight of the decoupling degree is increased to encourage the installation process to pay more attention to improving the accuracy. On the contrary, if the installation progress is slow but the accuracy is high, the weight of the path length is increased to speed up the installation speed and improve the installation efficiency. Through the adaptive adjustment of the weight, the installation strategy can be dynamically optimized according to the actual situation to adapt to different installation stages and needs. For example, in a certain installation stage, the micro-unit determined by the installation index finds that the installation progress is 70% but the accuracy evaluation value is low. After the installation evaluation micro-unit calculates the cycle evaluation value, the adaptive adjustment micro-unit increases the weight of the decoupling degree according to the feedback mechanism, so that more attention is paid to improving the accuracy in the subsequent installation. In another stage, if the installation progress is only 40% but the accuracy is high, the adaptive adjustment micro-unit increases the weight of the path length to speed up the installation.
[0038] In a possible implementation, the heuristic function component subunit includes: a path length calculation microunit, the path length calculation microunit is used to calculate the path length of the guide rail and the bracket in the three-dimensional space, ,in, , Specifically, the coordinates of two points on the three-dimensional space path of the guide rail and the bracket are obtained, which are recorded as and , substitute the coordinate values into the path length calculation formula By calculating the square root of the sum of the squares of the distance differences between two points in the x, y, and z directions, we can get the straight-line distance between the two points, that is, the path length. The calculation process accurately quantifies the path length of the guide rail and the bracket in three-dimensional space.
[0039] The decoupling degree calculation expression component micro-unit is used to construct a calculation expression of the decoupling degree: , where d is the geometric distance between the guide rail and the bracket, F is the mechanical interference value between the guide rail and the bracket, e is a minimum constant used to avoid the denominator being zero, and α and β are the weight coefficients of the geometric distance and the mechanical interference value, respectively. Specifically, in order to evaluate the degree of decoupling between the guide rail and the bracket, the geometric distance d and the mechanical interference value F between them are first measured, and the minimum constant e is introduced to avoid the denominator being zero. According to the given weight coefficients α and β, the calculation expression of the degree of decoupling is constructed ,Through the expression, the influence of geometric distance and mechanical interference on the decoupling degree is comprehensively considered and quantified into a numerical value.
[0040] The heuristic function component micro-unit is used to integrate the path length and the degree of decoupling to construct the heuristic function as follows: , where γ is the strength coefficient of the heuristic function, is the weight coefficient of the path length, is the weight coefficient of the decoupling degree. Specifically, obtain the path length calculated above and the degree of decoupling The value of determines the strength coefficient γ of the heuristic function and the weight coefficient of the path length and the weight coefficient of the degree of decoupling , substitute the parameters into the heuristic function In the , the path length and decoupling degree are comprehensively weighted and calculated. The heuristic function H can comprehensively reflect the importance of path length and decoupling degree, and provide a basis for subsequent decision-making and optimization. For example, in a specific elevator guide rail installation plan, the length of a certain path is calculated to be 10 meters by the path length calculation micro-unit, and the decoupling degree calculation expression component micro-unit obtains a decoupling degree of 0.8. Assuming that the strength coefficient γ is 2, the path length weight coefficient is 0.6, the decoupling degree weight coefficient is 0.4, and the heuristic function component micro-unit constructs the heuristic function H=2×(0.6×10+0.4×0.8)=12.32 to guide subsequent installation planning and decision-making.
[0041] The installation positioning information determination module 50 is used to fit the elevator guide rails and the guide rail fixing brackets into the three-dimensional point cloud model, perform optimization analysis based on the virtual reference lines, and determine the installation positioning information. Specifically, the installation positioning information determination module 50 receives the previously constructed three-dimensional point cloud model and the generated virtual reference line, and accurately fits the relevant data and model of the elevator guide rail and the guide rail fixing bracket into the three-dimensional point cloud model, taking into account various parameters such as the shape, size, and connection method of the guide rail and the bracket to ensure that the fitting in the model has a high degree of accuracy and authenticity, and optimizes and analyzes the fitted situation based on the virtual reference line. The virtual reference line provides a reference and direction for installation positioning. In the analysis process, multiple factors will be comprehensively considered, such as the degree of matching between the guide rail and the bracket and the hoistway structure, the spatial relationship with other equipment or structures, the stability and safety of the installation, etc. Through calculation, the position and posture of the guide rail and the bracket in the three-dimensional point cloud model are continuously adjusted and optimized to determine accurate installation positioning information, including detailed parameters such as the specific installation position, angle, and height of the guide rail and the bracket. The information will provide accurate and reliable guidance for the actual elevator guide rail installation work. For example, in an actual installation scenario, the installation positioning information determination module 50 first fits the elevator guide rail and bracket model of a specific model into the three-dimensional point cloud model of the hoistway. Then, based on the virtual reference line, it was found that the current fitting position might cause a conflict with other structures, so the position and angle of the guide rail and bracket were adjusted. After multiple optimization analyses, the precise installation positioning information that can ensure smooth, stable installation and compliance with safety standards was finally determined.
[0042] In a possible implementation, the installation location information determination module 50 includes: a cloud model component unit, which is used to construct a three-dimensional point cloud model of the steel bar structure distribution based on the three-dimensional point cloud data of the steel bar. Specifically, the three-dimensional point cloud data of the steel bar is obtained, and the data contains detailed information such as the position, shape and distribution of the steel bar in the shaft. The point cloud data is processed and integrated using specific modeling algorithms and technologies. By identifying and connecting adjacent points, a three-dimensional structural framework of the steel bar is constructed, and this framework is refined and improved to form a three-dimensional point cloud model that can accurately reflect the distribution of the steel bar structure.
[0043] The interfering steel bar distribution determination unit is used to fit the three-dimensional point cloud model of the steel bar structure distribution to the three-dimensional point cloud data of the internal structure of the shaft, and determine the interfering steel bar distribution based on the virtual reference line. Specifically, the constructed three-dimensional point cloud model of the steel bar structure distribution is accurately fitted to the three-dimensional point cloud data of the internal structure of the shaft. During the fitting process, the previously generated virtual reference line is referenced to determine the distribution of steel bars that may conflict with the installation operation. By comparing the spatial relationship between the steel bar model, the virtual reference line and the shaft model, the position and range of the steel bars that may hinder the installation of the elevator guide rails and brackets are identified, thereby clarifying the distribution of the interfering steel bars.
[0044] The constraint optimization unit is used to use the conflicting steel bar distribution as a constraint condition to constrain the optimization analysis and determine the installation positioning information. Specifically, the conflicting steel bar distribution is used as a constraint condition, and the optimization algorithm is used for optimization analysis. The optimization algorithm will try different installation positions and parameter combinations under the premise of satisfying these constraints. For example, the algorithm may start from multiple initial installation position assumptions and calculate the installation effect evaluation index under each assumption. In the calculation process, factors such as the distance from the conflicting steel bar, the stability of the installation, and the compatibility with other structures will be considered. If a hypothesis violates the constraint condition of the conflicting steel bar distribution, it will be excluded or given a lower evaluation value. Through continuous iteration and comparison, the algorithm gradually converges to one or a group of optimal installation positions and parameter combinations, and finally determines the installation positioning information that meets the requirements. For example, in an actual elevator guide rail installation plan, the cloud model component unit successfully constructs a three-dimensional model of the steel bar. The conflicting steel bar distribution determination unit finds that some steel bars will hinder the originally planned installation position. The constraint optimization unit uses the optimization algorithm, and after many attempts and adjustments, it finally determines an optimal installation position that avoids conflicting steel bars and meets other installation requirements.
[0045] The monitoring and correction module 60 is used to perform installation monitoring and correction based on the installation positioning information and the virtual reference line by monitoring and collecting images. Specifically, the monitoring and correction module 60 receives the relevant data of the installation positioning information and the virtual reference line. During the installation process, the monitoring equipment arranged at the installation site continuously collects real-time images. The images contain the actual conditions of the elevator guide rails and related components being installed. The collected images are compared and analyzed with the installation positioning information and the virtual reference line. The installation positioning information provides detailed parameters such as the installation position and posture that should be achieved in theory, and the virtual reference line provides the overall installation path and direction guidance. During the comparison process, carefully check whether the position, angle, spacing, etc. of the installation components are consistent with the installation positioning information, and judge whether the actual installation path is along the direction of the virtual reference line and whether there is any deviation. If it is found that the actual installation situation deviates from the expectation For example, if the position of the installation component is offset, the angle is wrong, or the installation path deviates from the virtual reference line, the amount of correction required will be calculated. According to the calculated correction amount, the corresponding instructions will be issued to guide the installer to make adjustments. The installer can reposition the installation component, adjust the angle, or change the installation path according to the instructions to ensure the accuracy of the installation work and meet the design requirements. The entire monitoring and correction process is a continuous cycle, constantly collecting images, comparing and analyzing, calculating the correction amount, and making adjustments until the installation work fully meets the requirements of the installation positioning information and the virtual reference line. For example, at a certain stage of installing the elevator guide rail, the monitoring image shows that the installation position of the guide rail is 5 cm lower than the position specified by the installation positioning information and deviates from the virtual reference line by 3 degrees. The monitoring and correction module calculates the correction amount that requires the guide rail to be lifted up by 5 cm and rotated 3 degrees in the direction of the virtual reference line, and then notifies the installer to perform the corresponding operations to get the installation work back on track.
[0046] In the above, refer to Figure 1 The elevator guide rail installation monitoring system based on image processing according to an embodiment of the present invention is described in detail. Figure 2 The following describes an elevator guide rail installation monitoring method based on image processing according to an embodiment of the present invention.
[0047] Elevator guide rail installation monitoring method based on image processing, such as Figure 2 As shown, the method includes:
[0048] A suspended measuring device is used, equipped with a laser radar, an angle sensor, and a laser positioning device, and is lowered at a uniform speed along the guide rail to scan the shaft, so as to obtain three-dimensional point cloud data of the internal structure of the shaft; a three-dimensional point cloud model is constructed through point cloud data processing and recognition based on the three-dimensional point cloud data of the internal structure of the shaft; the three-dimensional point cloud model is analyzed and recognized through a trained neural network model to determine the opening position of the elevator door; a virtual reference line is generated based on the opening position of the elevator door using an ant colony algorithm; the elevator guide rail and the guide rail fixing bracket are fitted into the three-dimensional point cloud model, and optimization analysis is performed based on the virtual reference line to determine the installation positioning information; based on the installation positioning information and the virtual reference line, installation monitoring and correction are performed through monitoring and acquisition images.
[0049] In a possible implementation, the elevator guide rail installation monitoring method based on image processing also includes: establishing an absolute coordinate system through the laser positioning device, and monitoring the verticality of the device in real time, and performing position correction based on the verticality; collecting three-dimensional point cloud data of the internal structure of the shaft, including shaft walls, floor door openings, and steel bars, through the collaborative work of laser radar and angle sensor; performing noise reduction and abnormal point removal preprocessing on the collected three-dimensional point cloud data, and using a segmentation algorithm to segment the point cloud data into different categories based on the classification of the shaft walls, floor door openings, and steel bars to obtain the three-dimensional point cloud data of the internal structure of the shaft.
[0050] In a possible implementation, the elevator guide rail installation monitoring method based on image processing also includes: constructing a three-dimensional point cloud model of the steel bar structure distribution based on the three-dimensional point cloud data of the steel bars; fitting the three-dimensional point cloud model of the steel bar structure distribution to the three-dimensional point cloud data of the internal structure of the shaft, and determining the distribution of the interfering steel bars based on the virtual reference line; using the distribution of the interfering steel bars as a constraint condition to constrain the optimization analysis and determine the installation positioning information.
[0051] In a possible implementation, the elevator guide rail installation monitoring method based on image processing also includes: taking the elevator floor door opening position as the starting point; initializing the pheromone matrix, the pheromone concentration on all paths is the same, and performing path search through an ant colony algorithm from the starting point to obtain an updated pheromone matrix; based on the updated pheromone matrix, iteratively searching until the pheromone converges, connecting the paths with the largest pheromone concentration, and performing smoothing to obtain the virtual reference line.
[0052] In a possible implementation, the elevator guide rail installation monitoring method based on image processing also includes: setting an objective function with the goal of minimizing the path length of the virtual reference line and with the goal of maximizing the degree of decoupling between the guide rail and the bracket being independent of each other; configuring the target weights of minimizing the path length and maximizing the degree of decoupling according to the requirements of the elevator guide rail installation scenario; fitting the target weights to the objective function; combining the path length and decoupling degree information to generate comprehensive heuristic information, and constructing a heuristic function based on the comprehensive heuristic information and the target weights; and performing path search through the ant colony algorithm based on the objective function and the heuristic function.
[0053] In a possible implementation, the elevator guide rail installation monitoring method based on image processing also includes: determining the installation progress and tracking the installation effect index by monitoring and collecting real-time images and data during installation; setting the installation cycle node, and establishing the mapping association between the installation progress and tracking the installation effect index and the installation cycle; performing installation evaluation according to the installation progress and tracking the installation effect index, and calculating the matching of the target with the installation cycle node to determine the cycle evaluation value, wherein the cycle evaluation value includes the installation efficiency evaluation value and the installation accuracy evaluation value; setting an adaptive feedback mechanism according to the adjustment relationship between the path length, the decoupling degree weight and the cycle evaluation value fitting the historical case data; matching the installation efficiency evaluation value and the installation accuracy evaluation value with the adaptive feedback mechanism to determine the target weight adjustment strategy, and adaptively adjusting the weights of the path length and the decoupling degree, wherein the path length adjustment weight corresponds to the installation efficiency evaluation value and the decoupling degree and the installation accuracy evaluation value, respectively.
[0054] In a possible implementation, the elevator guide rail installation monitoring method based on image processing further includes: calculating the path length of the guide rail and the bracket in three-dimensional space, ,in, , are the coordinates of two points on the path respectively; construct the calculation expression of the decoupling degree: , where d is the geometric distance between the guide rail and the bracket, F is the mechanical interference value between the guide rail and the bracket, e is a minimum constant used to avoid the denominator being zero, and α and β are weight coefficients of the geometric distance and the mechanical interference value, respectively; the path length and the degree of decoupling are combined to construct the heuristic function: , where γ is the strength coefficient of the heuristic function, is the weight coefficient of the path length, is the weight coefficient of the decoupling degree.
[0055] The elevator guide rail installation monitoring system based on image processing provided in the embodiment of the present invention can execute the elevator guide rail installation monitoring method based on image processing provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0056] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0057] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
Claims
1. The elevator guide rail installation monitoring system based on image processing is characterized by: The elevator guide rail installation monitoring system based on image processing includes: A three-dimensional point cloud data acquisition module is used to use a suspended measuring device equipped with a laser radar, an angle sensor, and a laser positioning device, and to lower the device at a uniform speed along the guide rail to scan the shaft and acquire three-dimensional point cloud data of the internal structure of the shaft; A three-dimensional point cloud model component module, wherein the three-dimensional point cloud model component module is used to construct a three-dimensional point cloud model based on the three-dimensional point cloud data of the internal structure of the shaft through point cloud data processing and recognition; A position determination module, the position determination module is used to analyze and identify the three-dimensional point cloud model through a trained neural network model to determine the elevator door opening position; A virtual reference line generation module, the virtual reference line generation module is used to generate a virtual reference line using an ant colony algorithm according to the elevator door opening position; An installation positioning information determination module, the installation positioning information determination module is used to fit the elevator guide rail and the guide rail fixing bracket into the three-dimensional point cloud model, perform optimization analysis based on the virtual reference line, and determine the installation positioning information; A monitoring and correction module, the monitoring and correction module is used to perform installation monitoring and correction through monitoring and acquisition images based on the installation positioning information and the virtual reference line; Wherein, the virtual reference line generation module includes: An initial point determination unit, the initial point determination unit is used to take the elevator door opening position as a starting point; A pheromone matrix initialization unit, wherein the pheromone matrix initialization unit is used to initialize the pheromone matrix, the pheromone concentrations on all paths are the same, and a path search is performed from the starting point through an ant colony algorithm to obtain an updated pheromone matrix; An iterative convergence unit, the iterative convergence unit being used to iteratively search based on the updated pheromone matrix until the pheromone converges, connect the paths with the maximum pheromone concentration, and perform smoothing to obtain the virtual reference line; The pheromone matrix initialization unit comprises: An objective function setting subunit, wherein the objective function setting subunit is used to set an objective function with the goal of minimizing the path length of the virtual reference line and with the goal of maximizing the degree of decoupling between the guide rail and the bracket being independent of each other and not affected by each other; A target weight configuration subunit, the target weight configuration subunit is used to configure target weights of minimizing the path length and maximizing the decoupling degree according to the elevator guide rail installation scenario requirements; A weight fitting subunit, the weight fitting subunit is used to fit the target weight to the target function; A heuristic function component subunit, the heuristic function component subunit is used to combine the path length and the decoupling degree information to generate comprehensive heuristic information, and construct a heuristic function based on the comprehensive heuristic information and the target weight; A path search subunit is used to perform path search through the ant colony algorithm based on the objective function and the heuristic function.
2. The elevator guide rail installation monitoring system based on image processing according to claim 1, characterized in that: The three-dimensional point cloud data acquisition module includes: A position correction unit, the position correction unit is used to establish an absolute coordinate system through the laser positioning device, monitor the verticality of the device in real time, and perform position correction based on the verticality; A collaborative acquisition unit, which is used to collect three-dimensional point cloud data of the internal structure of the shaft, including the shaft wall, floor door opening, and steel bars, by working in collaboration with a laser radar and an angle sensor; A preprocessing unit is used to perform noise reduction and abnormal point removal preprocessing on the collected three-dimensional point cloud data, and based on the classification of the shaft wall, floor door opening, and steel bars, a segmentation algorithm is used to segment the point cloud data into different categories to obtain the three-dimensional point cloud data of the internal structure of the shaft.
3. The elevator guide rail installation monitoring system based on image processing according to claim 2, characterized in that: The installation location information determination module includes: A cloud model component unit, wherein the cloud model component unit is used to construct a three-dimensional point cloud model of the steel bar structure distribution based on the three-dimensional point cloud data of the steel bar; A conflicting steel bar distribution determination unit, the conflicting steel bar distribution determination unit is used to fit the three-dimensional point cloud model of the steel bar structure distribution to the three-dimensional point cloud data of the internal structure of the shaft, and determine the conflicting steel bar distribution based on the virtual reference line; A constraint optimization unit is used to use the conflicting steel bar distribution as a constraint condition to constrain the optimization analysis and determine the installation positioning information.
4. The elevator guide rail installation monitoring system based on image processing according to claim 1, characterized in that: The target weight configuration subunit also includes: An installation indicator determination micro-unit, which is used to determine the installation progress and track the installation effect indicators by monitoring and collecting real-time images and data during installation; An installation cycle node setting micro-unit, wherein the installation cycle node setting micro-unit is used to set the installation cycle nodes and establish a mapping association between the installation progress and the installation effect tracking indicator and the installation cycle; An installation evaluation micro-unit, the installation evaluation micro-unit is used to perform installation evaluation according to the installation progress and tracking installation effect indicators, calculate the matching of the installation cycle node target, and determine the cycle evaluation value, the cycle evaluation value includes the installation efficiency evaluation value and the installation accuracy evaluation value; A feedback mechanism setting micro-unit, wherein the feedback mechanism setting micro-unit is used to fit the adjustment relationship between the path length, the decoupling degree weight and the period evaluation value according to the historical case data, and set the adaptive feedback mechanism; An adaptive adjustment micro-unit is used to match the installation efficiency evaluation value and the installation accuracy evaluation value with the adaptive feedback mechanism, determine the target weight adjustment strategy, and perform weight adaptive adjustment on the path length and the degree of decoupling, wherein the path length adjustment weight corresponds to the installation efficiency evaluation value, and the degree of decoupling corresponds to the installation accuracy evaluation value, respectively.
5. The elevator guide rail installation monitoring system based on image processing according to claim 1, characterized in that: The heuristic function component subunit comprises: A path length calculation micro unit, wherein the path length calculation micro unit is used to calculate the path length of the guide rail and the bracket in three-dimensional space, ,in, , are the coordinates of two points on the path respectively; The decoupling degree calculation expression component micro-unit is used to construct a calculation expression of the decoupling degree: , where d is the geometric distance between the guide rail and the bracket, F is the mechanical interference value between the guide rail and the bracket, e is a minimum constant used to avoid the denominator being zero, and α and β are weight coefficients of the geometric distance and the mechanical interference value respectively; The heuristic function component micro-unit is used to integrate the path length and the degree of decoupling to construct the heuristic function as follows: , where γ is the strength coefficient of the heuristic function, is the weight coefficient of the path length, is the weight coefficient of the decoupling degree.
6. The elevator guide rail installation monitoring method based on image processing is characterized in that: The method is applied to the elevator guide rail installation monitoring system based on image processing according to any one of claims 1 to 5, and the method comprises: Use a suspended measuring device equipped with a laser radar, angle sensor, and laser positioning equipment to scan the shaft at a constant speed along the guide rail to obtain three-dimensional point cloud data of the shaft's internal structure; Based on the three-dimensional point cloud data of the internal structure of the shaft, a three-dimensional point cloud model is constructed through point cloud data processing and recognition; The three-dimensional point cloud model is analyzed and identified by using a trained neural network model to determine the opening position of the elevator door; According to the elevator door opening position, a virtual reference line is generated by using an ant colony algorithm; Fitting the elevator guide rail and the guide rail fixing bracket into the three-dimensional point cloud model, performing optimization analysis based on the virtual reference line, and determining installation positioning information; Based on the installation positioning information and the virtual reference line, installation monitoring and correction are performed by monitoring and collecting images.
