Search

CN-121980965-A - Engineering BIM model accuracy verification method and system based on AR and AI cooperation

CN121980965ACN 121980965 ACN121980965 ACN 121980965ACN-121980965-A

Abstract

The invention discloses an engineering BIM model accuracy verification method and system based on AR and AI cooperation, which relates to the technical field of information transmission, and comprises the following steps of setting execution actions into a sequence form, calculating invalidity degree, regulating and controlling an acquisition period and acquisition frequency, registering the coordinate system of the built BIM model with the coordinate system of the entity model, judging whether a matrix corresponding to the execution action is transmitted or not on the basis of registration, and if not, judging whether readjustment is needed according to the acquisition period and the acquisition frequency. According to the invention, the acquisition period and frequency are automatically adjusted by introducing the invalidity index, unnecessary resource consumption is reduced, the system has self-adaptability and more intelligently responds to the change by dynamically regulating and controlling the acquisition strategy, the cosine similarity is utilized to judge whether the execution action corresponding matrix is successfully transmitted, the abnormality in the data transmission or model updating process is effectively detected, such as data loss, delay or error, and the accuracy of information synchronization is ensured.

Inventors

  • FANG LIWEI

Assignees

  • 温州职业技术学院

Dates

Publication Date
20260505
Application Date
20260401

Claims (10)

  1. 1. The engineering BIM model accuracy verification method based on AR and AI cooperation is characterized by comprising the following steps of S100, setting execution actions into a sequence form according to preset execution rules; Step S200, calculating invalidity degree according to the execution action in the sequence form, and regulating and controlling the acquisition period and the acquisition frequency according to the comparison relation between the invalidity degree and a preset invalidity degree threshold value, wherein the step S200 regulates and controls the acquisition period and the acquisition frequency on the basis of the invalidity degree of the execution operation; and step S300, registering the coordinate system of the constructed BIM model with the coordinate system of the entity model, judging whether a matrix corresponding to the execution action is transmitted or not based on cosine similarity on the basis of registration, and if not, judging whether readjustment is needed according to the acquisition period and the acquisition frequency.
  2. 2. The method for verifying accuracy of engineering BIM model based on AR and AI cooperation as claimed in claim 1, wherein an execution rule is set, the execution rule is expressed as a correspondence between an execution condition and an execution action, and the execution condition includes A i =A i+1 、A i A i+1 、A i+1 A i , and (a i )- =B, And is also provided with Performing actions includes not operating, adding, deleting, and modifying.
  3. 3. The method for verifying the accuracy of an engineering BIM model based on AR and AI collaboration according to claim 2, wherein the execution actions are represented as a sequence in an acquisition period, and the sequence of the execution actions is , wherein, Denoted as the kth acquisition period, The execution action is obtained according to the execution rule and is represented by the ith acquisition time and the (i+1) th acquisition time in the kth acquisition period.
  4. 4. The method for verifying the accuracy of an engineering BIM model based on the cooperation of AR and AI according to claim 3, wherein the sequence of executing actions is assigned; calculating invalidity according to the assigned execution action, setting a first value as an invalidity threshold, comparing the invalidity with the first value, and regulating and controlling the acquisition period and the acquisition frequency according to a comparison result; The method for calculating the invalidity degree comprises the steps of selecting any acquisition period and a corresponding sequence of executing actions, counting the number of elements in the sequence of executing actions, marking the number as a first number, counting the number of elements with value of 0 in the sequence of executing actions, marking the number as a second number, calculating the ratio of the second number to the first number, and setting the ratio of the second number to the first number as the invalidity degree.
  5. 5. The method for verifying the accuracy of an engineering BIM model based on the cooperation of AR and AI according to claim 4, wherein after the acquisition frequency and the acquisition period are set, a coordinate system is constructed, corresponding execution actions of addition, deletion and modification are verified according to the coordinate system and the scanned image, and transmission data is selected according to the verification result.
  6. 6. The method for verifying the accuracy of an engineering BIM model based on AR and AI collaboration according to claim 1, wherein the pixel coordinates are mapped to coordinates among the coordinates according to a conversion relation between a camera coordinate system and a three-dimensional coordinate system.
  7. 7. The method for verifying the precision of an engineering BIM model based on the cooperation of AR and AI according to claim 6, wherein the method for constructing a coordinate system is characterized in that the method comprises the steps of arbitrarily selecting a point in the BIM model as an origin, taking any direction as an x-axis, taking any direction vertical to the x-axis as a y-axis, taking the direction vertical to a plane formed by the x-axis and the y-axis as a z-axis, and constructing an initial three-dimensional coordinate system; Converting the BIM model into a cylindrical superposition body with a first size, wherein the first size is expressed as a first cross-sectional area and is first high, the first size meets a first constraint condition, the first constraint condition is expressed as that the area filled by the segmented BIM model is minimum, and the difference between the total volume of the cylindrical superposition body with the first size and the original volume of the BIM model is minimum; Setting the parameter structure of the cylinder as [ x, y, z, a, b, c, r ], wherein (x, y, z) is any point in the cylinder, (a, b, c) is the coordinate of the end point of the unit vector in the axial direction of the cylinder, and r is the radius of the cross section of the cylinder; Randomly selecting points in a coordinate system, marking the points as local points, calculating first distances between the local points and (a, b and c) according to an Euclidean distance formula, calculating an average value of the first distances, traversing the average value of the first distances corresponding to the local points, carrying out ascending order on the average value of the first distances, selecting the average value of the first distances with the smallest numerical value as an original point, taking a first direction as an x axis, taking a second direction as a y axis and taking a third direction as a z axis, and constructing the coordinate system, wherein the first direction, the second direction and the third direction are mutually perpendicular; the coordinate system meets a second constraint condition, wherein the second constraint condition is expressed as that the x-axis coordinates in the BIM model are distributed in 0-positive infinity, and the number of points with y-axis coordinates distributed in 0-positive infinity is the largest, so that the construction of the coordinate system is completed.
  8. 8. The method for verifying the accuracy of an engineering BIM model based on the cooperation of AR and AI according to claim 1, wherein after the registration operation of a coordinate system is completed, corresponding execution actions of addition, deletion and modification are verified, and the verification method is realized based on a cosine similarity formula.
  9. 9. The method for verifying the accuracy of an engineering BIM model based on the cooperation of AR and AI according to claim 1, wherein a transmission operation is set according to a verification result, when the verification result is that matching is performed, a corresponding set is transmitted, when the verification result is that matching is not performed, a value of the corresponding set is set to 0, a calculation step of invalidation is skipped, and whether the acquisition period and the acquisition frequency need to be regulated again is judged according to the calculated invalidation.
  10. 10. The engineering BIM model accuracy verification system based on AR and AI cooperation is used for executing the engineering BIM model accuracy verification method based on AR and AI cooperation, and is characterized by comprising a setting module, a regulating module and a verification module; The setting module sets the execution action into a sequence form according to a preset execution rule; the regulation and control module calculates the invalidity degree according to the execution action in the sequence form, and regulates and controls the acquisition period and the acquisition frequency according to the comparison relation between the invalidity degree and a preset invalidity degree threshold value; The verification module registers the coordinate system of the built BIM model with the coordinate system of the entity model, judges whether a matrix corresponding to the execution action is transmitted or not based on cosine similarity on the basis of registration, and if not, judges whether readjustment is needed according to the acquisition period and the acquisition frequency.

Description

Engineering BIM model accuracy verification method and system based on AR and AI cooperation Technical Field The invention relates to the technical field of information transmission, in particular to an engineering BIM model accuracy verification method and system based on AR and AI cooperation. Background In recent years, for the rapid development of modeling technology of engineering, BIM technology aims at the situation that the field changes frequently and data update is needed in time, and the situation of calibration lag is easy to generate, while three-dimensional laser scanning technology can collect the point cloud data of a field building, compare, convert and coordinate BIM model with the data of an actual field, and eliminate errors, so that the purposes of assisting engineering quality inspection, rapid modeling and reducing rework are achieved, and the method plays an important role in the aspect of three-dimensional modeling and verification in reverse engineering. At present, in the chinese patent of publication No. CN120317022a, a three-dimensional simulation verification method and system based on BIM are disclosed, and the method is implemented by performing high-precision welding management on real data through virtual simulation, so that thermal deformation of a workpiece seam at different positions in a welding process can be simulated and fitted on the premise of synchronous conditions, when the current welding spot is displaced to the next welding spot after welding, a small amount of change of the seam position in a cooling state is judged, and further, a deviation correction value which should be made by the welding spot is calculated. Disclosure of Invention The technical problem to be solved by the invention is that in the related art, all data transmission is carried out aiming at verification data acquired by slow BIM updating response, whether the data need to be transmitted or not is judged or whether the data are effective or not is checked, the updating rapidity is not facilitated, the updating precision is not facilitated, and a certain limitation exists. In order to solve the technical problems, the present invention provides a technical scheme, in a first aspect, an engineering BIM model accuracy verification method based on cooperation of AR and AI, comprising the following steps, step S100, setting an execution action to a sequence form according to a preset execution rule; Step S200, calculating invalidity degree according to the execution action in the sequence form, and regulating and controlling the acquisition period and the acquisition frequency according to the comparison relation between the invalidity degree and a preset invalidity degree threshold value, wherein the step S200 regulates and controls the acquisition period and the acquisition frequency on the basis of the invalidity degree of the execution operation; and step S300, registering the coordinate system of the constructed BIM model with the coordinate system of the entity model, judging whether a matrix corresponding to the execution action is transmitted or not based on cosine similarity on the basis of registration, and if not, judging whether readjustment is needed according to the acquisition period and the acquisition frequency. As the optimal scheme of the engineering BIM model accuracy verification method based on AR and AI cooperation, an execution rule is set, wherein the execution rule is expressed as the corresponding relation between an execution condition and an execution action, and the execution condition comprises A i=Ai+1、AiAi+1、Ai+1A i, and (a i)-=B,And is also provided withPerforming actions includes not operating, adding, deleting, and modifying. As the optimal scheme of the engineering BIM model accuracy verification method based on AR and AI cooperation, in the invention, the execution action is expressed as a sequence in the acquisition period, and the sequence of the execution action is that, wherein,Denoted as the kth acquisition period,The execution action is obtained according to the execution rule and is represented by the ith acquisition time and the (i+1) th acquisition time in the kth acquisition period. As the optimal scheme of the engineering BIM model accuracy verification method based on the cooperation of AR and AI, the sequence of executing actions is assigned; calculating invalidity according to the assigned execution action, setting a first value as an invalidity threshold, comparing the invalidity with the first value, and regulating and controlling the acquisition period and the acquisition frequency according to a comparison result; The method for calculating the invalidity degree comprises the steps of selecting any acquisition period and a corresponding sequence of executing actions, counting the number of elements in the sequence of executing actions, marking the number as a first number, counting the number of elements with value of 0 in the sequence of