CN116843704A - Intelligent twin body registration and reconstruction method, device, equipment and medium - Google Patents

Intelligent twin body registration and reconstruction method, device, equipment and medium Download PDF

Info

Publication number
CN116843704A
CN116843704A CN202310827456.4A CN202310827456A CN116843704A CN 116843704 A CN116843704 A CN 116843704A CN 202310827456 A CN202310827456 A CN 202310827456A CN 116843704 A CN116843704 A CN 116843704A
Authority
CN
China
Prior art keywords
point cloud
equipment
model
dimensional
power grid
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310827456.4A
Other languages
Chinese (zh)
Inventor
王鹤
王刚
王菲
范子恺
胡亚山
于海
周爱华
杨佩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Smart Grid Research Institute Co ltd
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
Economic and Technological Research Institute of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
State Grid Smart Grid Research Institute Co ltd
State Grid Corp of China SGCC
State Grid Jiangsu Electric Power Co Ltd
Economic and Technological Research Institute of State Grid Jiangsu Electric Power Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by State Grid Smart Grid Research Institute Co ltd, State Grid Corp of China SGCC, State Grid Jiangsu Electric Power Co Ltd, Economic and Technological Research Institute of State Grid Jiangsu Electric Power Co Ltd filed Critical State Grid Smart Grid Research Institute Co ltd
Priority to CN202310827456.4A priority Critical patent/CN116843704A/en
Publication of CN116843704A publication Critical patent/CN116843704A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/761—Proximity, similarity or dissimilarity measures
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
    • G06V10/765—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects using rules for classification or partitioning the feature space
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
    • G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10028—Range image; Depth image; 3D point clouds

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Databases & Information Systems (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Computer Graphics (AREA)
  • Geometry (AREA)
  • Processing Or Creating Images (AREA)

Abstract

The invention relates to the field of application of power system information technology, and discloses a twin body intelligent registration and reconstruction method, a device, equipment and a medium, wherein the method is realized by acquiring laser point cloud data of power transmission and transformation engineering; dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; performing association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result; and carrying out gesture alignment on the single equipment point cloud and the three-dimensional design drawing based on the mapping result, so that registration and reconstruction of two twin bodies of the single equipment point cloud and the three-dimensional design drawing are realized. By implementing the method, the problem that the design drawing of the transformer substation is lack of modification and adjustment based on the three-dimensional point cloud data of the transformer substation in the related technology is solved, and the accuracy of operation and maintenance management and control after the three-dimensional design drawing is transferred to an operation and maintenance link is improved.

Description

Intelligent twin body registration and reconstruction method, device, equipment and medium
Technical Field
The invention relates to the field of application of power system information technology, in particular to a twin intelligent registration and reconstruction method, a twin intelligent registration and reconstruction device, equipment and a twin intelligent registration and reconstruction medium.
Background
With the rapid development of the power industry, higher requirements are also put on the construction of a transformer substation, and digital reconstruction of the transformer substation is an important content for monitoring and diagnosis of the transformer substation. Based on the outline of the substation equipment, three-dimensional features of the substation equipment are extracted, the substation equipment is identified and classified according to the extracted features, and meanwhile, the spatial position relation of the substation equipment can be clearly known according to the position and posture information. According to the model number and the spatial position relation of the equipment, a digital reconstruction model is built for the existing transformer substation, the real position distribution and the equipment details of the transformer substation equipment can be reflected, the problems in design can be found conveniently and timely, the upgrading maintenance and the layout adjustment of the transformer substation are facilitated, and meanwhile more accurate data can be provided for reconstruction of the transformer substation according to requirements.
Classification identification of electrical equipment is a key technology for digital reconstruction of substations, and identification data of the electrical equipment is derived from three-dimensional laser point clouds. Three-dimensional point cloud data of the appearance of the transformer substation equipment is obtained through a three-dimensional laser scanner, so that the equipment can be classified and identified by utilizing the appearance characteristics of the transformer substation. Meanwhile, the automatic classification and identification of the transformer substation equipment is realized, and the method has a basic supporting effect on large-scale three-dimensional automatic digital reconstruction of the transformer substation.
In addition, the design drawing in the engineering construction of the transformer substation is usually transferred to the operation and maintenance link of the equipment, however, due to the difference between the actual construction process and the design drawing, the modification and adjustment of the design drawing are usually carried out manually in the operation and maintenance link of the equipment. The manual adjustment has the problems of large workload and low efficiency. Therefore, how to modify and adjust the design drawing of the transformer substation based on the three-dimensional point cloud data of the transformer substation is a problem to be solved urgently at present.
Disclosure of Invention
In view of the above, the invention provides a twin intelligent registration and reconstruction method, device, equipment and medium, which are used for solving the problem that the existing three-dimensional point cloud data based on the transformer substation is lack to modify and adjust the design drawing of the transformer substation.
In a first aspect, the present invention provides a twin intelligent registration and reconstruction method, the method comprising: acquiring laser point cloud data of power transmission and transformation engineering; dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; performing association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result; and carrying out gesture alignment on the single equipment point cloud and the three-dimensional design drawing based on the mapping result, so that registration and reconstruction of two twin bodies of the single equipment point cloud and the three-dimensional design drawing are realized.
According to the intelligent twin body registration and reconstruction method provided by the invention, laser point cloud data of power transmission and transformation engineering are obtained; dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; performing association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result; and aligning the single equipment point cloud with the three-dimensional design drawing based on the mapping result. The method realizes the gesture alignment of the equipment obtained by adopting the point cloud scanning and the equipment in the three-dimensional design drawing, solves the problem that the three-dimensional point cloud data based on the transformer substation is lack in the related technology to modify and adjust the design drawing of the transformer substation, and improves the accuracy of operation and maintenance management and control after the three-dimensional design drawing is transferred to an operation and maintenance link.
In an alternative embodiment, the power grid engineering point cloud processing model library is constructed in the following manner: acquiring a power grid equipment point cloud model obtained by discretizing a power grid equipment point cloud scanning model and a power grid equipment three-dimensional model; matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data; carrying out depth feature extraction on point cloud data in a point cloud scanning model of the power grid equipment and three-dimensional model difference space data by adopting a three-dimensional deep learning network model; clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed.
In an alternative embodiment, the device cluster model includes: the three-dimensional model based on the power grid equipment business knowledge comprises plane, sphere and cone model topology construction information in the defined point cloud according to power grid equipment construction, and the equipment clustering model based on the area growth comprises preset conditions for stopping growth between different targets in the defined point cloud according to the expansion condition of the power grid equipment.
In an optional implementation manner, the power transmission and transformation engineering laser point cloud data is segmented by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain a single equipment point cloud, and the method comprises the following steps: comparing the three-dimensional design drawing with laser point cloud data of the power transmission and transformation project to obtain a reference comparison point; determining partitions of laser point cloud data of the power transmission and transformation project based on a three-dimensional model point cloud association model of the power grid equipment and a reference comparison point; performing discrete point removal, point cloud ground removal, downsampling and smoothing on laser point cloud data of power transmission and transformation engineering based on a point cloud model data preprocessing range, a related environment association range, a related equipment association range and a related equipment topology range; and carrying out matching segmentation on the partitioned power transmission and transformation project laser point cloud data based on the equipment clustering model to obtain single equipment point cloud.
In an optional implementation manner, the method for obtaining the single device point cloud by matching and dividing the partitioned power transmission and transformation project laser point cloud data based on the device cluster model comprises the following steps: determining a point cloud similarity candidate set in the partitioned power transmission and transformation project laser point cloud data according to a three-dimensional model based on power grid equipment business knowledge; removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value; and generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting laser point cloud data of the power transmission and transformation project to obtain single equipment point clouds.
In an optional implementation manner, before removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the device cluster model based on the region growth to obtain a point cloud matching scheme with similarity greater than a threshold value, the method further comprises: removing point clouds which do not meet the conditions in the point cloud similarity candidate set based on the point cloud model data preprocessing range; and generating a discretization point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretization point cloud model.
According to the intelligent twin body registration and reconstruction method provided by the invention, discrete point removal, ground removal, downsampling, smoothing and other processing steps of the laser point cloud data of the power transmission and transformation project are realized through the pre-constructed network engineering point cloud processing model library, so that the point cloud is easier to segment and extract the monomers. Meanwhile, a point cloud similarity candidate set of the laser point cloud data of the power transmission and transformation project and point cloud removal which does not meet the conditions are selected based on the equipment clustering model, so that the single equipment is segmented, and the segmentation accuracy is ensured.
In an optional implementation manner, the mapping method for performing association mapping on the single device point cloud based on the three-dimensional device model point cloud library of the power grid to obtain a mapping result comprises the following steps: converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library; generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment and the equipment point cloud in the equipment point cloud library; based on the equipment point cloud library, the noise data and the three-dimensional point cloud training siamese deep learning model of the power grid equipment, obtaining compact learning local feature descriptors and generating a classifier; the class of the point cloud of the single device is determined based on the classifier.
According to the intelligent twin body registration and reconstruction method provided by the invention, a deep neural network is introduced, a point cloud key point extraction method and a feature descriptor suitable for complex power equipment are researched, and the accuracy of three-dimensional matching and retrieval of a power grid based on appearance is improved. Noise data is introduced at the same time, so that the recognition efficiency and accuracy are further improved.
In an alternative embodiment, aligning the single device point cloud with the three-dimensional design drawing based on the mapping result includes: matching the single equipment point cloud with a three-dimensional model in a three-dimensional design drawing based on the mapping result; and adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the point cloud of the single equipment and the posture of the three-dimensional model.
In a second aspect, the present invention provides a twin intelligent registration and reconstruction device, the device comprising: the data acquisition module is used for acquiring laser point cloud data of power transmission and transformation engineering; the segmentation module is used for segmenting the laser point cloud data of the power transmission and transformation project by adopting a pre-constructed power grid project point cloud processing model library and a three-dimensional design drawing to obtain a single equipment point cloud; the mapping module is used for carrying out association mapping on the point cloud of the single equipment based on the point cloud library of the three-dimensional equipment model of the power grid to obtain a mapping result; and the registration module is used for aligning the single equipment point cloud with the three-dimensional design drawing based on the mapping result, so that the registration and reconstruction of the two twin bodies of the single equipment point cloud and the three-dimensional design drawing are realized.
In an alternative embodiment, the power grid engineering point cloud processing model library is constructed in the following manner: acquiring a power grid equipment point cloud model obtained by discretizing a power grid equipment point cloud scanning model and a power grid equipment three-dimensional model; matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data; carrying out depth feature extraction on point cloud data in a point cloud scanning model of the power grid equipment and three-dimensional model difference space data by adopting a three-dimensional deep learning network model; clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed.
In an alternative embodiment, the device cluster model includes: the three-dimensional model based on the power grid equipment business knowledge comprises plane, sphere and cone model topology construction information in the defined point cloud according to power grid equipment construction, and the equipment clustering model based on the area growth comprises preset conditions for stopping growth between different targets in the defined point cloud according to the expansion condition of the power grid equipment.
In an alternative embodiment, the segmentation module includes: the comparison module is used for comparing the three-dimensional design drawing with the laser point cloud data of the power transmission and transformation project to obtain a reference comparison point; the partitioning module is used for determining the partitioning of the power transmission and transformation project laser point cloud data based on the three-dimensional model point cloud association model of the power grid equipment and the reference comparison points; the preprocessing module is used for performing discrete point removal, point cloud ground removal, downsampling and smoothing on the laser point cloud data of the power transmission and transformation project based on the point cloud model data preprocessing range, the related environment association range, the related equipment association range and the related equipment topology range; and the segmentation sub-module is used for carrying out matching segmentation on the partitioned power transmission and transformation project laser point cloud data based on the equipment clustering model to obtain single equipment point cloud.
In an alternative embodiment, the segmentation submodule is specifically configured to: determining a point cloud similarity candidate set in the partitioned power transmission and transformation project laser point cloud data according to a three-dimensional model based on power grid equipment business knowledge; removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value; and generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting laser point cloud data of the power transmission and transformation project to obtain single equipment point clouds.
In an alternative embodiment, the segmentation sub-module is further configured to: removing point clouds which do not meet the conditions in the point cloud similarity candidate set based on the point cloud model data preprocessing range; and generating a discretization point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretization point cloud model.
In an alternative embodiment, the mapping module is specifically configured to: converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library; generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment and the equipment point cloud in the equipment point cloud library; based on the equipment point cloud library, the noise data and the three-dimensional point cloud training siamese deep learning model of the power grid equipment, obtaining compact learning local feature descriptors and generating a classifier; the class of the point cloud of the single device is determined based on the classifier.
In an alternative embodiment, the registration module is specifically configured to: matching the single equipment point cloud with a three-dimensional model in a three-dimensional design drawing based on the mapping result; and adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the point cloud of the single equipment and the posture of the three-dimensional model.
In a third aspect, the present invention provides a computer device comprising: the intelligent twin registering and reconstructing method comprises a memory and a processor, wherein the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions so as to execute the intelligent twin registering and reconstructing method according to the first aspect or any corresponding implementation mode.
In a fourth aspect, the present invention provides a computer readable storage medium having stored thereon computer instructions for causing a computer to perform the twin body intelligent registration and reconstruction method of the first aspect or any of its corresponding embodiments.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow diagram of a twin intelligent registration and reconstruction method according to an embodiment of the present invention;
FIG. 2 is a flow diagram of another twin intelligent registration and reconstruction method according to an embodiment of the present invention;
FIG. 3 is a flow diagram of a further twin intelligent registration and reconstruction method according to an embodiment of the present invention;
FIG. 4 is a flow diagram of yet another twin intelligent registration and reconstruction method according to an embodiment of the present invention;
FIG. 5 is a block diagram of a twin intelligent registration and reconstruction device according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a hardware structure of a computer device according to an embodiment of the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In accordance with an embodiment of the present invention, a twin intelligent registration and reconstruction method embodiment is provided, it being noted that the steps illustrated in the flow diagrams of the figures may be performed in a computer system, such as a set of computer executable instructions, and that although a logical order is illustrated in the flow diagrams, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.
In this embodiment, a twin body intelligent registration and reconstruction method is provided, which can be used for electronic devices, such as computers, mobile phones, tablet computers, etc., fig. 1 is a flowchart of a twin body intelligent registration and reconstruction method according to an embodiment of the present invention, and as shown in fig. 1, the flowchart includes the following steps:
step S101, acquiring laser point cloud data of power transmission and transformation engineering; the power transmission and transformation project can be a transformer substation in a power grid, and laser point cloud data of the transformer substation are obtained by scanning the transformer substation by adopting point cloud scanning equipment. In addition, the power transmission and transformation project may be a converter station, etc., and the embodiment of the present invention is not particularly limited.
Step S102, dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; the three-dimensional design drawing is obtained by carrying out model design through current three-dimensional software in the early stage of power transmission and transformation project construction, such as a power grid project construction link, and is needed to be transferred to a power grid equipment operation and maintenance link in the follow-up process. The power grid engineering point cloud processing model library comprises a plurality of power grid equipment construction rules obtained through calculation. The power transmission and transformation project laser point cloud data obtained through point cloud scanning comprises a plurality of power grid devices such as transformers, isolating switches, capacitors and the like, and the extraction of the plurality of power grid device point cloud data is realized by dividing a pre-constructed power grid project point cloud processing model library and a three-dimensional design drawing.
Step S103, carrying out association mapping on the point cloud of the single equipment based on the point cloud library of the three-dimensional equipment model of the power grid to obtain a mapping result; specifically, the point cloud library of the three-dimensional equipment model of the power grid contains common or most of point cloud data of the power grid equipment, and the single equipment and the point cloud library are mapped, so that the extracted single equipment can be identified.
And step S104, aligning the single device point cloud and the three-dimensional design drawing based on the mapping result, and realizing the registration and reconstruction of the two twin bodies of the single device point cloud and the three-dimensional design drawing. Specifically, because a certain difference exists between the power transmission and transformation project obtained through actual construction and the three-dimensional design drawing, the single equipment obtained through point cloud scanning and segmentation is adopted to adjust the equipment three-dimensional model in the three-dimensional design drawing, so that the gesture alignment of the power transmission and transformation project and the three-dimensional design drawing is realized. Therefore, the three-dimensional design drawing after the gesture alignment is handed over to the operation and maintenance link of the power grid equipment, so that the operation and maintenance link can be more accurately realized.
The method comprises the steps of scanning a power transmission and transformation project by using a point cloud scanning device to obtain power transmission and transformation laser point cloud data, and performing model design by using three-dimensional software to obtain a three-dimensional design drawing, so that the combination mapping of entity equipment to a digital space is realized, namely the power transmission and transformation laser point cloud data and the three-dimensional design drawing are twins of the entity equipment such as the power transmission and transformation project; through the association mapping and posture alignment process, comparison of equipment in the point cloud data and equipment in the three-dimensional design drawing and adjustment of equipment posture are realized, namely, registration and reconstruction of two twin bodies of single equipment point cloud and three-dimensional design drawing are realized.
According to the twin body intelligent registration and reconstruction method provided by the embodiment of the invention, the laser point cloud data of the power transmission and transformation project is obtained; dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; performing association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result; and aligning the single equipment point cloud with the three-dimensional design drawing based on the mapping result. The method realizes the gesture alignment of the equipment obtained by adopting the point cloud scanning and the equipment in the three-dimensional design drawing, solves the problem that the three-dimensional point cloud data based on the transformer substation is lack in the related technology to modify and adjust the design drawing of the transformer substation, and improves the accuracy of operation and maintenance management and control after the three-dimensional design drawing is transferred to an operation and maintenance link.
In this embodiment, a twin intelligent registration and reconstruction method is provided, fig. 2 is a flowchart of the twin intelligent registration and reconstruction method according to an embodiment of the present invention, as shown in fig. 2, and the flowchart includes the following steps:
step S201, acquiring laser point cloud data of power transmission and transformation engineering; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.
Step S202, dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud;
specifically, the power grid engineering point cloud processing model library is constructed in the following manner:
s21, acquiring a grid equipment point cloud model obtained by discretizing a grid equipment point cloud scanning model and a grid equipment three-dimensional model; specifically, the power grid equipment point cloud scanning model is an actual equipment point cloud model obtained by scanning an actual power grid equipment by adopting point cloud scanning equipment, wherein the actual power grid equipment comprises all known power grid equipment or typical equipment such as a transformer, a mutual inductor, a disconnecting switch, a capacitor and the like of common power grid equipment. The three-dimensional model of the power grid equipment is obtained by modeling and designing the power grid equipment by adopting three-dimensional design software. The point cloud model of the power grid equipment is obtained by discretizing the three-dimensional model of the power grid equipment by adopting a two-stage grid sampling method based on Poisson-Disk distribution.
Step S22, matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data; specifically, the two are matched, the same point cloud data which is successfully matched is removed, and different point cloud data which is reserved is three-dimensional model difference space data.
S23, carrying out depth feature extraction on point cloud data in a point cloud scanning model of power grid equipment and three-dimensional model difference space data by adopting a three-dimensional deep learning network model; specifically, point cloud data in a point cloud scanning model of the power grid equipment and three-dimensional model difference space data are taken as samples, and a three-dimensional deep learning network model is adopted to extract features so as to obtain depth features. When the feature extraction is performed, the point cloud data in the sample can be converted into voxel grids, and the feature extraction can be performed by using a model constructed by a convolutional neural network, a cyclic neural network or a graph neural network and the like.
And step S24, clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed. Wherein, a K-means clustering algorithm is adopted in the clustering process, and the working flow is as follows: k points are randomly selected as an initial centroid (i.e., the center of all points in a cluster), and each point in the dataset is then assigned to a cluster, specifically, each point is found its nearest centroid and assigned to the cluster to which it corresponds. After this step is completed, the centroid of each cluster is updated to be the average of all points of the cluster. The above steps are repeated until the centroid does not change. In this step, a specific k value is determined from the number of three-dimensional models of the grid device, for example, the k value is set to 200. And carrying out K-means algorithm clustering on the extracted depth features to obtain a plurality of corresponding grid equipment point cloud clustering centers. And generating a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment according to the clustering center, and constructing a power grid engineering point cloud processing model library based on the generated contents.
In the power grid engineering point cloud processing model library, a point cloud association model is discretized point cloud data corresponding to a three-dimensional model of a clustering center; the preprocessing range of the point cloud model data is the point cloud difference point degree, and if the space difference is within 30%; the related environment association range is a cluster center environment such as a transformer substation environment and a transmission line environment; the related equipment association range clusters center association ranges such as transformer model association bus models; the topology range of the related equipment is the connection relation between the cluster center and other structures.
The device cluster model includes: the three-dimensional model based on the power grid equipment business knowledge and the equipment clustering model based on the region growth comprise model topology construction information such as planes, balls, cones and the like in a defined point cloud according to power grid equipment construction, and specifically, the three-dimensional model based on the power grid equipment business knowledge can be understood to comprise construction information of the three-dimensional model of a clustering center, such as construction of a plurality of planes, a plurality of balls, a plurality of cones or the like.
The equipment clustering model based on the regional growth comprises the preset condition of stopping growth among different targets in the point cloud according to the expansion condition of the power grid equipment. For example, when the number of transformers is different, the lengths of the buses are different, and the device cluster model based on region growth defines the number or the length of connected structures corresponding to different numbers of power grid devices, such as the lengths of the buses corresponding to two transformers and the lengths of the buses corresponding to three transformers.
Specifically, the step S202 includes:
and step S2021, comparing the three-dimensional design drawing with the laser point cloud data of the power transmission and transformation project to obtain a reference comparison point.
And step S2022, determining the subareas of the laser point cloud data of the power transmission and transformation project based on the three-dimensional model point cloud association model of the power grid equipment and the reference comparison points.
Specifically, the matching comparison of the three-dimensional design drawing, the point cloud association model and the power transmission and transformation project laser point cloud data can firstly carry out approximate partitioning on the power transmission and transformation project laser point cloud data.
Step S2023, performing discrete point removal, point cloud ground removal, downsampling and smoothing on the power transmission and transformation engineering laser point cloud data based on the point cloud model data preprocessing range, the related environment association range, the related equipment association range and the related equipment topology range.
And step S2024, carrying out matching segmentation on the partitioned power transmission and transformation project laser point cloud data based on the equipment clustering model to obtain single equipment point cloud.
In some alternative embodiments, step S2024 described above comprises:
step a1, determining a point cloud similarity candidate set in partitioned power transmission and transformation project laser point cloud data according to a three-dimensional model based on power grid equipment business knowledge; specifically, according to model combination information such as planes, balls, cones and the like in a three-dimensional model of power grid equipment business knowledge and power transmission and transformation project laser point cloud data, determining a part similar to the three-dimensional model in the power transmission and transformation project laser point cloud data as a point cloud similarity candidate set.
Step a2, removing point clouds which do not meet the condition in the point cloud similarity candidate set based on the point cloud model data preprocessing range; for example, the preprocessing range of the cuboid at the top of the transformer is less than 30%, for example, if a certain cuboid in the candidate set of the point cloud similarity of the transformer exceeds 30%, the cuboid is considered to be not a transformer model, and the cuboid is removed.
And a step a3 of generating a discretized point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretized point cloud model. Specifically, the process of generating the discretized point cloud model can be obtained by discretizing by adopting a two-stage grid sampling method based on Poisson-Disk distribution. And removing irrelevant point cloud data in the point cloud similarity candidate set based on the discretized point cloud model.
Step a4, removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value; specifically, a condition for target growth is defined in the equipment cluster model based on region growth, and according to the condition, point clouds which do not meet the condition in the point cloud similarity candidate set are removed, for example, a bus can be lengthened according to the scale length of a transformer and a circuit breaker, but the length range covers equipment such as the transformer and the circuit breaker, and the equipment has a certain proportionality, so that the equipment which does not meet the proportionality can be deleted. In addition, after the point cloud matching scheme is determined, the user can be sent to confirm first, and after the user confirms, the next processing is carried out.
And a step a5, generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting laser point cloud data of power transmission and transformation engineering to obtain single equipment point clouds. Specifically, after the point cloud matching scheme is determined, further determining a point cloud topological relation, such as a connection topological relation of a transformer, a circuit breaker and a bus, determining a point cloud candidate set according to the topological relation to obtain a final point cloud segmentation scheme, and segmenting laser point cloud data of power transmission and transformation engineering based on the scheme to obtain single equipment.
Step S203, carrying out association mapping on single equipment point clouds based on a three-dimensional equipment model point cloud library of the power grid to obtain a mapping result; please refer to step S103 in the embodiment shown in fig. 1 in detail, which is not described herein.
And step S204, aligning the single device point cloud and the three-dimensional design drawing based on the mapping result, and realizing the registration and reconstruction of the two twin bodies of the single device point cloud and the three-dimensional design drawing. Please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.
According to the intelligent twin body registration and reconstruction method provided by the invention, discrete point removal, ground removal, downsampling, smoothing and other processing steps of the laser point cloud data of the power transmission and transformation project are realized through the pre-constructed network engineering point cloud processing model library, so that the point cloud is easier to segment and extract the monomers. Meanwhile, a point cloud similarity candidate set of the laser point cloud data of the power transmission and transformation project and point cloud removal which does not meet the conditions are selected based on the equipment clustering model, so that the single equipment is segmented, and the segmentation accuracy is ensured.
In this embodiment, a twin intelligent registration and reconstruction method is provided, fig. 3 is a flowchart of the twin intelligent registration and reconstruction method according to an embodiment of the present invention, as shown in fig. 3, and the flowchart includes the following steps:
step S301, acquiring laser point cloud data of power transmission and transformation engineering; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.
Step S302, dividing laser point cloud data of power transmission and transformation projects by adopting a pre-constructed power grid project point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; please refer to step S102 in the embodiment shown in fig. 1 in detail, which is not described herein.
And step S303, carrying out association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result.
Specifically, the step S303 includes:
step S3031, converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library; specifically, the three-dimensional model of the power grid equipment is obtained by modeling and designing the power grid equipment by adopting three-dimensional design software. And storing the three-dimensional models in a model library to obtain a three-dimensional model library of the power grid equipment, performing triangular meshing on the three-dimensional models in the model library, designing three-dimensional model sampling rules aiming at business rules of the three-dimensional models of the power grid equipment, increasing sampling aiming at edges and model joints, adopting simplified point clouds for corresponding rule cubes and cylinders, and finally obtaining a point cloud library of the power grid equipment model. And then, the point cloud can be rotated and stretched according to the service rule to obtain equipment model point clouds with different situations, and the equipment model point clouds are stored in a point cloud library.
Step S3032, generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment obtained by scanning and the equipment point cloud in the equipment point cloud library; specifically, the three-dimensional point cloud of the power grid equipment obtained by scanning can be obtained by scanning the power grid equipment under different angles in different environments, such as day or night, front or back, and the like. Matching the three-dimensional point cloud of the power grid equipment with the equipment point cloud in the equipment point cloud library, and generating noise data according to the difference of the three-dimensional point cloud and the equipment point cloud.
Step S3033, based on the equipment point cloud library, the noise data and the grid equipment three-dimensional point cloud training siamese deep learning model, obtaining compact learning local feature descriptors and generating a classifier; specifically, a siamese deep learning architecture is adopted for model training, a compact learning local feature descriptor 3DSmoothNet is obtained, a device point cloud base, noise data and power grid device point cloud matching relation are established, and then a classifier is built by combining the type or model of the power grid device.
Step S3034, determining a class of the point cloud of the single device based on the classifier. Specifically, based on the classifier, the single equipment point cloud obtained by segmentation is matched with equipment in a point cloud library, and the single equipment point cloud can be identified by combining the type or model of the power grid equipment in the classifier, such as a transformer or a disconnecting switch.
And step S304, aligning the single device point cloud and the three-dimensional design drawing based on the mapping result, and realizing the registration and reconstruction of the two twin bodies of the single device point cloud and the three-dimensional design drawing. Please refer to step S104 in the embodiment shown in fig. 1 in detail, which is not described herein.
According to the intelligent twin body registration and reconstruction method provided by the invention, a deep neural network is introduced, a point cloud key point extraction method and a feature descriptor suitable for complex power equipment are researched, and the accuracy of three-dimensional matching and retrieval of a power grid based on appearance is improved. Noise data is introduced at the same time, so that the recognition efficiency and accuracy are further improved.
In this embodiment, a twin intelligent registration and reconstruction method is provided, and fig. 4 is a flowchart of the twin intelligent registration and reconstruction method according to an embodiment of the present invention, as shown in fig. 4, where the flowchart includes the following steps:
step S401, acquiring laser point cloud data of power transmission and transformation engineering; please refer to step S101 in the embodiment shown in fig. 1 in detail, which is not described herein.
Step S402, dividing laser point cloud data of power transmission and transformation engineering by adopting a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud; please refer to step S102 in the embodiment shown in fig. 1 in detail, which is not described herein.
Step S403, carrying out association mapping on the point cloud of the single equipment based on the point cloud library of the three-dimensional equipment model of the power grid to obtain a mapping result; please refer to step S103 in the embodiment shown in fig. 1 in detail, which is not described herein.
And step S404, aligning the single device point cloud and the three-dimensional design drawing based on the mapping result, and realizing the registration and reconstruction of the two twin bodies of the single device point cloud and the three-dimensional design drawing.
Specifically, the step S404 includes:
step S4041, matching the single equipment point cloud with the three-dimensional model in the three-dimensional design drawing based on the mapping result; specifically, for example, a transformer with a certain model of single equipment point cloud is identified, and then the transformer is matched with a corresponding model of transformer in the three-dimensional design drawing.
And step S4042, adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the single equipment point cloud and the three-dimensional model. Specifically, according to the difference between the two matching results, a least square method is adopted, and the alignment of the single equipment point cloud and the three-dimensional model gesture in the three-dimensional design drawing is realized by minimizing the square of the error. For example, if the single device point cloud is inclined by 5 degrees relative to the three-dimensional model, the three-dimensional model is adjusted by 5 degrees, so that the two gestures are aligned.
The embodiment provides a twin body intelligent registration and reconstruction device, as shown in fig. 5, including:
the data acquisition module 501 is used for acquiring laser point cloud data of power transmission and transformation engineering;
the segmentation module 502 is configured to segment laser point cloud data of a power transmission and transformation project by using a pre-constructed power grid engineering point cloud processing model library and a three-dimensional design drawing, so as to obtain a single equipment point cloud;
the mapping module 503 is configured to perform association mapping on the point cloud of the single device based on the point cloud library of the three-dimensional device model of the power grid, so as to obtain a mapping result;
and the registration module 504 is used for aligning the single device point cloud and the three-dimensional design drawing based on the mapping result, so that the registration and reconstruction of the single device point cloud and the two twin bodies of the three-dimensional design drawing are realized.
The twin body intelligent registration and reconstruction device in this embodiment is presented in the form of functional units, where the units refer to ASIC circuits, processors and memories executing one or more software or fixed programs, and/or other devices that can provide the above described functionality.
In an alternative embodiment, the power grid engineering point cloud processing model library is constructed in the following manner: acquiring a power grid equipment point cloud model obtained by discretizing a power grid equipment point cloud scanning model and a power grid equipment three-dimensional model; matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data; carrying out depth feature extraction on point cloud data in a point cloud scanning model of the power grid equipment and three-dimensional model difference space data by adopting a three-dimensional deep learning network model; clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed.
In an alternative embodiment, the device cluster model includes: the three-dimensional model based on the power grid equipment business knowledge comprises plane, sphere and cone model topology construction information in the defined point cloud according to power grid equipment construction, and the equipment clustering model based on the area growth comprises preset conditions for stopping growth between different targets in the defined point cloud according to the expansion condition of the power grid equipment.
In an alternative embodiment, the segmentation module includes: the comparison module is used for comparing the three-dimensional design drawing with the laser point cloud data of the power transmission and transformation project to obtain a reference comparison point; the partitioning module is used for determining the partitioning of the power transmission and transformation project laser point cloud data based on the three-dimensional model point cloud association model of the power grid equipment and the reference comparison points; the preprocessing module is used for performing discrete point removal, point cloud ground removal, downsampling and smoothing on the laser point cloud data of the power transmission and transformation project based on the point cloud model data preprocessing range, the related environment association range, the related equipment association range and the related equipment topology range; and the segmentation sub-module is used for carrying out matching segmentation on the partitioned power transmission and transformation project laser point cloud data based on the equipment clustering model to obtain single equipment point cloud.
In an alternative embodiment, the segmentation submodule is specifically configured to: determining a point cloud similarity candidate set in the partitioned power transmission and transformation project laser point cloud data according to a three-dimensional model based on power grid equipment business knowledge; removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value; and generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting laser point cloud data of the power transmission and transformation project to obtain single equipment point clouds.
In an alternative embodiment, the segmentation sub-module is further configured to: removing point clouds which do not meet the conditions in the point cloud similarity candidate set based on the point cloud model data preprocessing range; and generating a discretization point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretization point cloud model.
In an alternative embodiment, the mapping module is specifically configured to: converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library; generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment and the equipment point cloud in the equipment point cloud library; based on the equipment point cloud library, the noise data and the three-dimensional point cloud training siamese deep learning model of the power grid equipment, obtaining compact learning local feature descriptors and generating a classifier; the class of the point cloud of the single device is determined based on the classifier.
In an alternative embodiment, the registration module is specifically configured to: matching the single equipment point cloud with a three-dimensional model in a three-dimensional design drawing based on the mapping result; and adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the point cloud of the single equipment and the posture of the three-dimensional model.
Further functional descriptions of the above respective modules and units are the same as those of the above corresponding embodiments, and are not repeated here.
The embodiment of the invention also provides computer equipment, which is provided with the twin body intelligent registration and reconstruction device shown in the figure 5.
Referring to fig. 6, fig. 6 is a schematic structural diagram of a computer device according to an alternative embodiment of the present invention, as shown in fig. 6, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting the various components, including high-speed interfaces and low-speed interfaces. The various components are communicatively coupled to each other using different buses and may be mounted on a common motherboard or in other manners as desired. The processor may process instructions executing within the computer device, including instructions stored in or on memory to display graphical information of the GUI on an external input/output device, such as a display device coupled to the interface. In some alternative embodiments, multiple processors and/or multiple buses may be used, if desired, along with multiple memories and multiple memories. Also, multiple computer devices may be connected, each providing a portion of the necessary operations (e.g., as a server array, a set of blade servers, or a multiprocessor system). One processor 10 is illustrated in fig. 6.
The processor 10 may be a central processor, a network processor, or a combination thereof. The processor 10 may further include a hardware chip, among others. The hardware chip may be an application specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general-purpose array logic, or any combination thereof.
Wherein the memory 20 stores instructions executable by the at least one processor 10 to cause the at least one processor 10 to perform a method for implementing the embodiments described above.
The memory 20 may include a storage program area that may store an operating system, at least one application program required for functions, and a storage data area; the storage data area may store data created from the use of the computer device of the presentation of a sort of applet landing page, and the like. In addition, the memory 20 may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, memory 20 may optionally include memory located remotely from processor 10, which may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
Memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk, or solid state disk; the memory 20 may also comprise a combination of the above types of memories.
The computer device also includes a communication interface 30 for the computer device to communicate with other devices or communication networks.
The embodiments of the present invention also provide a computer readable storage medium, and the method according to the embodiments of the present invention described above may be implemented in hardware, firmware, or as a computer code which may be recorded on a storage medium, or as original stored in a remote storage medium or a non-transitory machine readable storage medium downloaded through a network and to be stored in a local storage medium, so that the method described herein may be stored on such software process on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, a solid state disk or the like; further, the storage medium may also comprise a combination of memories of the kind described above. It will be appreciated that a computer, processor, microprocessor controller or programmable hardware includes a storage element that can store or receive software or computer code that, when accessed and executed by the computer, processor or hardware, implements the methods illustrated by the above embodiments.
Although embodiments of the present invention have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope of the invention as defined by the appended claims.

Claims (18)

1. A twin intelligent registration and reconstruction method, the method comprising:
acquiring laser point cloud data of power transmission and transformation engineering;
dividing the laser point cloud data of the power transmission and transformation project by adopting a pre-constructed power grid project point cloud processing model library and a three-dimensional design drawing to obtain single equipment point cloud;
performing association mapping on the single equipment point cloud based on a power grid three-dimensional equipment model point cloud library to obtain a mapping result;
and carrying out posture alignment on the single equipment point cloud and the three-dimensional design drawing based on the mapping result, so as to realize registration and reconstruction of two twin bodies of the single equipment point cloud and the three-dimensional design drawing.
2. The method according to claim 1, wherein the grid engineering point cloud processing model library is constructed by adopting the following modes:
acquiring a power grid equipment point cloud model obtained by discretizing a power grid equipment point cloud scanning model and a power grid equipment three-dimensional model;
Matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data;
adopting a three-dimensional deep learning network model to extract depth features of point cloud data in the point cloud scanning model of the power grid equipment and difference space data of the three-dimensional model;
clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed.
3. The method of claim 2, wherein the device cluster model comprises: the three-dimensional model based on the power grid equipment business knowledge comprises plane, sphere and cone model topology construction information in a defined point cloud according to power grid equipment construction, and the equipment clustering model based on the area growth comprises preset conditions for stopping growth between different targets in the defined point cloud according to the expansion condition of the power grid equipment.
4. The method of claim 3, wherein the dividing the power transmission and transformation project laser point cloud data by using a pre-constructed power grid project point cloud processing model library and a three-dimensional design drawing to obtain a single device point cloud comprises:
comparing the three-dimensional design drawing with the power transmission and transformation project laser point cloud data to obtain a reference comparison point;
determining the subareas of the power transmission and transformation project laser point cloud data based on the three-dimensional model point cloud association model of the power grid equipment and the reference comparison points;
performing discrete point removal, point cloud ground removal, downsampling and smoothing on the power transmission and transformation engineering laser point cloud data based on the point cloud model data preprocessing range, the related environment association range, the related equipment association range and the related equipment topology range;
and carrying out matching segmentation on the partitioned laser point cloud data of the power transmission and transformation project based on the equipment clustering model to obtain a single equipment point cloud.
5. The method of claim 4, wherein the obtaining the single device point cloud based on the device cluster model by matching and dividing the partitioned power transmission and transformation project laser point cloud data comprises:
Determining a point cloud similarity candidate set in the partitioned power transmission and transformation project laser point cloud data according to the three-dimensional model based on the power grid equipment business knowledge;
removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value;
and generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting the laser point cloud data of the power transmission and transformation project to obtain the point cloud of the single equipment.
6. The method of claim 4, wherein removing point clouds in the candidate set of point cloud similarity according to the device cluster model based on region growing, before obtaining a point cloud matching scheme with similarity greater than a threshold value, further comprises:
removing point clouds which do not meet the conditions in the point cloud similarity candidate set based on the point cloud model data preprocessing range;
and generating a discretization point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretization point cloud model.
7. The method of claim 1, wherein performing the associative mapping on the single device point cloud based on the grid three-dimensional device model point cloud library to obtain a mapping result comprises:
Converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library;
generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment and the equipment point cloud in the equipment point cloud library;
based on the equipment point cloud library, the noise data and the three-dimensional point cloud training siamese deep learning model of the power grid equipment, obtaining compact learning local feature descriptors and generating a classifier;
and determining the category of the single device point cloud based on the classifier.
8. The method of claim 1, wherein the posing the single device point cloud and three-dimensional design drawing based on the mapping result comprises:
matching the single equipment point cloud with a three-dimensional model in a three-dimensional design drawing based on the mapping result;
and adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the point cloud of the single equipment and the posture of the three-dimensional model.
9. A twin intelligent registration and reconstruction device, the device comprising:
the data acquisition module is used for acquiring laser point cloud data of power transmission and transformation engineering;
the segmentation module is used for segmenting the laser point cloud data of the power transmission and transformation project by adopting a pre-built power grid project point cloud processing model library and a three-dimensional design drawing to obtain a single equipment point cloud;
The mapping module is used for carrying out association mapping on the single equipment point cloud based on the three-dimensional equipment model point cloud library of the power grid to obtain a mapping result;
and the registration module is used for aligning the gesture of the single equipment point cloud and the three-dimensional design drawing based on the mapping result, so that the registration and reconstruction of the two twin bodies of the single equipment point cloud and the three-dimensional design drawing are realized.
10. The apparatus of claim 9, wherein the grid engineering point cloud processing model library is constructed by: acquiring a power grid equipment point cloud model obtained by discretizing a power grid equipment point cloud scanning model and a power grid equipment three-dimensional model; matching the grid equipment point cloud scanning model with the grid equipment point cloud model to obtain three-dimensional model difference space data; carrying out depth feature extraction on point cloud data in a point cloud scanning model of the power grid equipment and three-dimensional model difference space data by adopting a three-dimensional deep learning network model; clustering is carried out based on the extracted features, a three-dimensional model point cloud association model, a point cloud model data preprocessing range, a related environment association range, a related equipment topology range and an equipment clustering model of the power grid equipment are generated, and a power grid engineering point cloud processing model library is constructed.
11. The apparatus of claim 10, wherein the device cluster model comprises: the three-dimensional model based on the power grid equipment business knowledge comprises plane, sphere and cone model topology construction information in the defined point cloud according to power grid equipment construction, and the equipment clustering model based on the area growth comprises preset conditions for stopping growth between different targets in the defined point cloud according to the expansion condition of the power grid equipment.
12. The apparatus of claim 11, wherein the partitioning module comprises: the comparison module is used for comparing the three-dimensional design drawing with the laser point cloud data of the power transmission and transformation project to obtain a reference comparison point; the partitioning module is used for determining the partitioning of the power transmission and transformation project laser point cloud data based on the three-dimensional model point cloud association model of the power grid equipment and the reference comparison points; the preprocessing module is used for performing discrete point removal, point cloud ground removal, downsampling and smoothing on the laser point cloud data of the power transmission and transformation project based on the point cloud model data preprocessing range, the related environment association range, the related equipment association range and the related equipment topology range; and the segmentation sub-module is used for carrying out matching segmentation on the partitioned power transmission and transformation project laser point cloud data based on the equipment clustering model to obtain single equipment point cloud.
13. The apparatus of claim 12, wherein the segmentation submodule is specifically configured to: determining a point cloud similarity candidate set in the partitioned power transmission and transformation project laser point cloud data according to a three-dimensional model based on power grid equipment business knowledge; removing point clouds which do not meet the conditions in the point cloud similarity candidate set according to the equipment clustering model based on the region growth, and obtaining a point cloud matching scheme with similarity larger than a threshold value; and generating a point cloud segmentation scheme according to the point cloud matching scheme and a preset topological relation, and segmenting laser point cloud data of the power transmission and transformation project to obtain single equipment point clouds.
14. The apparatus of claim 12, wherein the segmentation submodule is further configured to: removing point clouds which do not meet the conditions in the point cloud similarity candidate set based on the point cloud model data preprocessing range; and generating a discretization point cloud model based on the three-dimensional design drawing, and removing point clouds which do not meet the condition in the point cloud similarity candidate set according to the discretization point cloud model.
15. The apparatus of claim 9, wherein the mapping module is specifically configured to: converting a three-dimensional model library of the power grid equipment constructed based on the three-dimensional model of the power grid equipment into an equipment point cloud library; generating noise data based on the difference between the three-dimensional point cloud of the power grid equipment and the equipment point cloud in the equipment point cloud library; based on the equipment point cloud library, the noise data and the three-dimensional point cloud training siamese deep learning model of the power grid equipment, obtaining compact learning local feature descriptors and generating a classifier; the class of the point cloud of the single device is determined based on the classifier.
16. The apparatus according to claim 9, wherein the registration module is specifically configured to: matching the single equipment point cloud with a three-dimensional model in a three-dimensional design drawing based on the mapping result; and adjusting the three-dimensional model in the three-dimensional design drawing according to the matching result to realize the alignment of the point cloud of the single equipment and the posture of the three-dimensional model.
17. A computer device, comprising:
a memory and a processor communicatively coupled to each other, the memory having stored therein computer instructions that, upon execution, perform the twin intelligent registration and reconstruction method of any of claims 1-8.
18. A computer readable storage medium having stored thereon computer instructions for causing a computer to perform the twin body intelligent registration and reconstruction method of any one of claims 1 to 8.
CN202310827456.4A 2023-07-06 2023-07-06 Intelligent twin body registration and reconstruction method, device, equipment and medium Pending CN116843704A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310827456.4A CN116843704A (en) 2023-07-06 2023-07-06 Intelligent twin body registration and reconstruction method, device, equipment and medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310827456.4A CN116843704A (en) 2023-07-06 2023-07-06 Intelligent twin body registration and reconstruction method, device, equipment and medium

Publications (1)

Publication Number Publication Date
CN116843704A true CN116843704A (en) 2023-10-03

Family

ID=88161278

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310827456.4A Pending CN116843704A (en) 2023-07-06 2023-07-06 Intelligent twin body registration and reconstruction method, device, equipment and medium

Country Status (1)

Country Link
CN (1) CN116843704A (en)

Similar Documents

Publication Publication Date Title
CN108228798B (en) Method and device for determining matching relation between point cloud data
CN110472585B (en) VI-S L AM closed-loop detection method based on inertial navigation attitude track information assistance
CN113487730B (en) Urban three-dimensional automatic modeling method based on laser radar point cloud data
CN110544298B (en) Substation modeling method, device, computer equipment and storage medium
CN105183834A (en) Ontology library based transportation big data semantic application service method
CN111159897A (en) Target optimization method and device based on system modeling application
CN112819199A (en) Precipitation prediction method, device, equipment and storage medium
CN117453937B (en) Automatic generation method and device for power grid graph data model
CN103970901A (en) Geographic information graphic data integration method
CN116977549A (en) Construction method and system of distribution network digital twin body
Zha et al. ENGD-BiFPN: A remote sensing object detection model based on grouped deformable convolution for power transmission towers
CN110569546B (en) Traffic cell division method and device
Yuan et al. Automatic cables segmentation from a substation device based on 3D point cloud
CN116843704A (en) Intelligent twin body registration and reconstruction method, device, equipment and medium
CN116403235A (en) Electrical wiring diagram recognition system and method based on computer vision
Li et al. 3D scene reconstruction using a texture probabilistic grammar
CN113128006A (en) High-voltage power distribution method and device based on neural network
CN114913330A (en) Point cloud component segmentation method and device, electronic equipment and storage medium
Ding et al. Study on the automatic modeling method of 3D information model for substations
US20230040195A1 (en) Three-dimensional point cloud identification device, learning device, three-dimensional point cloud identification method, learning method and program
CN113723405A (en) Method and device for determining area outline and electronic equipment
Liu et al. A multi-seed dynamic local graph matching model for tracking of densely packed cells across unregistered microscopy image sequences
Takkala et al. CHIP: Clustering hotspots in layout using integer programming
CN117826116B (en) Method and device for determining opening and closing states of double-column horizontal rotary isolating switch
CN112579841B (en) Multi-mode database establishment method, retrieval method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination