Construction progress regulation and control system and method in engineering construction based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of construction progress regulation and control, and discloses a construction progress regulation and control system and method in engineering construction based on an unmanned aerial vehicle. The method comprises the following steps: firstly, acquiring an inclined image of an engineering construction site, extracting feature points by using a light beam method and an SURF algorithm, matching the feature points, and establishing an engineering construction site live-action model; secondly, clustering site clouds of the engineering construction site in combination with an engineering construction site real scene model, extracting a construction site building point cloud set, performing deviation analysis, and adjusting the engineering construction progress; a BP neural network is trained, and the engineering construction completion rate is predicted; and finally, according to the engineering construction completion rate, the engineering construction progress plan is adjusted and optimized based on a litsea rotundifolia optimization algorithm, and engineering construction progress control is realized. The purpose of regulating and controlling the construction progress is achieved by adjusting and controlling the engineering construction progress, and the method is accurate and objective.

Description

Technical Field

[0001] The present invention relates to the technical field of construction progress control, and in particular to a construction progress control system and method in engineering construction based on unmanned aerial vehicle. Background Art

[0002] Chinese patent application CN115167212A discloses a foundation pit dynamic construction control system and method based on a monitoring platform. The system specifically includes an engineering visualization module, which can dynamically visualize engineering information, detection information and monitoring point numbers in a foundation pit visualization model; a foundation pit antenna monitoring module, which detects the foundation pit and uploads the foundation pit detection data; a data analysis module, which pre-processes the foundation pit detection data and obtains a dynamically changing trend fitting curve, and uses the trend fitting curve to detect and analyze the foundation pit construction and perform subsequent data correction; a theoretical analysis interaction module, which compares and analyzes the detection data and theoretical calculation values, selects the optimal calculation model, and predicts the subsequent construction of the foundation pit state changes, while predicting the foundation pit segmentation and foundation pit partition; a dynamic construction control module, which calculates the alarm value and obtains the dynamic construction change trend of each construction stage, foundation pit segmentation and foundation pit partition through the predicted value; an alarm module, which automatically alarms when the alarm value exceeds the alarm threshold during the dynamic construction of the foundation pit.

[0003] The traditional construction progress control method uses on-site command to control the construction progress. There are problems such as low efficiency of information transmission and uncontrollable control process in the process. In addition, the unreasonable allocation of construction resources brings troubles to the advancement of project construction. At the same time, high-tech technologies such as neural networks are not used, and the progress and real-time performance of the project construction are not guaranteed. Summary of the invention

[0004] In view of the problems in the related technology, the present invention provides a construction progress control system and method in engineering construction based on unmanned aerial vehicle, so as to overcome the above-mentioned technical problems existing in the existing related technology.

[0005] In order to solve the above technical problems, the present invention is achieved through the following technical solutions: The present invention is a construction progress control method in engineering construction based on unmanned aerial vehicle, comprising the following steps: S1. The UAV obtains the inclined image of the construction site, corrects the inclined image of the construction site based on the beam method, and then uses the SURF algorithm to extract and match the feature points. After obtaining the matched inclined image of the construction site, a white film model is generated to establish a real-scene model of the construction site; S2, extracting the construction site point cloud according to the real-scene model of the construction site, clustering the construction site point cloud based on the improved clustering algorithm, extracting the point cloud set of the construction site buildings, obtaining the construction progress model, calculating the progress deviation, performing deviation analysis, and adjusting the construction progress; S3, setting a construction schedule, combining the progress deviation, training a BP neural network, obtaining a BP neural network model, and predicting the completion rate of the construction; S4. In combination with the completion rate of the project construction, the project construction schedule is adjusted and optimized based on the golden jackal optimization algorithm, and the adjusted project construction schedule is obtained by finding the best fitness function value to achieve project construction progress control.

[0006] The invention obtains the inclined image of the construction site through a drone, corrects the inclined image of the construction site using the beam method, extracts feature points through the SURF algorithm, matches the feature points of the inclined image of the construction site, generates a white film model, and then establishes a real-life model of the construction site through texture mapping; secondly, extracts the point cloud of the construction site, clusters it based on the improved clustering algorithm, and selects the point cloud set of the construction site buildings by setting a threshold, thereby determining the construction progress model, and adjusting the construction progress by comparing the daily construction progress of the construction progress model and analyzing the deviation; the method effectively reduces the data volume of the point cloud, accurately selects the building, and judges the construction progress through the point cloud set of the construction site building; then, the BP neural network is trained in combination with the number of construction workers, the construction level and the daily progress deviation to predict the future construction completion rate; the neural network can effectively handle the nonlinear relationship between the various elements of the construction, and the prediction result is highly accurate; finally, the construction progress plan is adjusted and optimized based on the golden jackal optimization algorithm, the daily construction resource volume, the daily number of construction workers and the daily construction funds are optimized, and the control of the construction progress is completed; the algorithm imitates the collaborative hunting behavior of the golden jackal to find the optimal solution to the problem, and has the advantages of strong global optimization ability.

[0007] Preferably, the S1 comprises the following steps: S11. A control point is selected at the construction site. The drone performs oblique photography above the control point to obtain an initial oblique image of the construction site. The initial oblique image of the construction site is subjected to distortion correction and uniform light and color processing to obtain an oblique image of the construction site. A three-dimensional rectangular coordinate system is established at the construction site. The coordinates of the pixel points on the oblique image of the construction site are set as , the focal length is , the coordinates of the control points at the construction site are , the center projection coordinates of the UAV are , the coefficients corresponding to the external azimuth elements of the inclined image of the construction site are , , , , , , , and , then the collinear equation calculation formula is as follows: , ; According to the collinear equation, the coordinates of the pixel points on the inclined image of the engineering construction site are calculated, and the inclined image of the engineering construction site is corrected using the coordinates of the pixel points on the inclined image of the engineering construction site to obtain a processed inclined image of the engineering construction site; S12, selecting in the processed engineering construction site inclined image i×j Pixel points are sampled, and the pixel points are quantized to generate the processed engineering construction site tilt matrix A as follows: ; in, Indicates horizontal i , vertical j pixels; The SURF (fast and robust features) algorithm is used to detect and extract feature points of the processed engineering construction site tilt matrix, and feature point matching is completed to obtain the matched engineering construction site tilt matrix. The specific steps are as follows: S121, select pixel points in the processed engineering construction site tilt matrix, calculate the integral corresponding to the pixel points and record it as an integral image, and select feature points in the integral image , calculate the feature points A Gaussian second-order differential convolution is performed at the position, and a Hessian matrix is ​​constructed using the Gaussian second-order differential convolution; a scale box filter is set, and the integral image is constructed using the scale box filter to obtain a spatial scale; the Hessian matrix determines the feature points in the spatial scale, and then the feature points are located in the processed engineering construction site tilt matrix; S122, selecting a submatrix with the feature point as the center, calculating the four-dimensional vector of the submatrix in the processed engineering construction site tilt matrix in the horizontal direction and the vertical direction, the four-dimensional vector is the feature point direction, when the traces of the Hessian matrices of the two feature points have the same sign in the feature point direction, the corresponding two feature points are matched; sequentially matching all pixel points in the processed engineering construction site tilt matrix to obtain a matched engineering construction site tilt matrix; S13, obtaining a matched tilted image of the construction site according to the matched tilted matrix of the construction site, generating ultra-high density point cloud data, forming a dense triangulated network, and then generating a white film model; and establishing a real-scene three-dimensional model based on the white film model using texture mapping, wherein the real-scene three-dimensional model is the real-scene model of the construction site.

[0008] The invention obtains the tilted image of the construction site through a drone, corrects the tilted image of the construction site using the light beam method, extracts feature points, matches the feature points of the tilted image of the construction site, generates a white film model, and then establishes a real-life model of the construction site through texture mapping.

[0009] Preferably, S2 comprises the following steps: S21, extracting the point cloud of the construction site according to the real-life model of the construction site, obtaining a point cloud set of the construction site, selecting any point cloud from the point cloud set of the construction site, recording it as the point cloud to be processed, establishing a neighborhood of the point cloud to be processed, and setting the neighborhood threshold to , traverse the point cloud set of the construction site and the point cloud to be processed whose Euclidean distance is less than the neighborhood threshold, and add them to the clustered point cloud set. The Euclidean distance calculation formula is as follows: ; in, , and Indicates the number of points in the construction site point cloud collection The 3D coordinates of the point cloud of the construction site, , and Represents the 3D coordinates of the point cloud to be processed, represents the Euclidean distance; A cluster point cloud set is obtained, and then any cluster point cloud is selected from the cluster point set, recorded as a to-be-processed cluster point cloud, and a point cloud whose Euclidean distance with the to-be-processed cluster point cloud is less than a neighborhood threshold is selected from the cluster point cloud set to obtain a new cluster point cloud set, and S21 is repeated until all the construction site point clouds are clustered to obtain a final cluster point cloud set; S22. Set the length, width and height thresholds of the construction site buildings to be , and If the clustered point clouds in the final clustered point cloud set meet the length, width and height thresholds of the buildings on the construction site, the final clustered point cloud set is recorded as the construction site building point cloud set; otherwise, the corresponding clustered point cloud is deleted, and S22 is repeated until the clustered point clouds in the final clustered point cloud set meet the length, width and height thresholds of the buildings on the construction site; the construction site building point cloud set is merged with the real scene model of the engineering construction site to obtain an engineering construction progress model; Obtain the construction schedule, set the expected construction schedule and construction phase according to the construction schedule, re-photograph the construction site after the construction phase, obtain a new construction schedule model, compare the construction schedule model with the new construction schedule model, analyze whether there is a deviation, and adjust the construction schedule. The specific steps are as follows: S221. Set the planned construction volume in the construction phase as The unit price of construction budget is The construction work has been completed , then the calculation formula for the construction progress deviation is as follows: , , ; in, Indicates the amount of resources consumed by the project completed during the construction phase. Indicates the amount of resources consumed by the project that has been completed during the construction phase. Indicates progress deviation; When the progress deviation is less than 0, the construction progress is behind the construction progress plan; when the progress deviation is greater than 0, the construction progress is ahead of the construction progress plan; when the progress deviation is equal to 0, there is no deviation in the construction progress; S222. If the construction progress of the project lags behind the construction progress plan and the construction period cannot be completed on schedule, the construction progress of the project will be adjusted; if the construction progress of the project is ahead of the construction progress plan, the project will be completed ahead of schedule to ensure the quality of the construction; if there is no deviation in the construction progress, the construction progress will not be adjusted; the construction progress adjustment is completed.

[0010] The invention extracts point clouds of construction sites, selects point cloud sets of buildings at construction sites based on an improved clustering algorithm, determines a construction progress model, compares the daily construction progress of the construction progress model, and analyzes deviations to adjust the construction progress. This method greatly reduces the amount of point cloud data and can accurately select building point clouds, thereby effectively judging the construction progress.

[0011] Preferably, S3 comprises the following steps: S31. According to the construction schedule, the actual number of construction workers per day is obtained, recorded as the number of construction workers. The construction team is scored daily by professional construction personnel such as the construction project manager and engineers, with the score ranging from 0 to 1 to quantify the construction level. The daily progress deviation is calculated by the progress deviation. The construction progress data set is obtained by combining the number of construction workers, the construction level and the daily progress deviation, recorded as ,in Indicates m Number of construction workers per day, Indicates m The level of engineering construction of the day, Indicates m daily progress deviation of the day; The construction schedule of previous years was obtained, the number of construction workers, construction level and daily progress deviation of previous years were calculated, and a sample data set of construction was formed. The sample data set of construction was divided into a training set of construction and a test set of construction. The number of neurons in the input layer of the BP neural network was set to 3, the number of neurons in the hidden layer was set to 9, the number of neurons in the output layer was set to 1, and the learning rate was set to , the activation function is the sigmoid function, the engineering construction training set is input into the BP (back propagation) neural network, and the current number of iterations is set to , the maximum number of iterations is When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain a trained BP neural network; the engineering construction test set is then input into the trained BP neural network, and the error threshold is set to , when the output result error is less than the error threshold, the BP neural network model is obtained, otherwise the weight is adjusted until the output result error is less than the error threshold; S32, inputting the engineering construction progress data set into the BP neural network model, and the BP neural network model outputs a predicted value, which is the engineering construction completion rate.

[0012] The invention predicts the completion rate of engineering construction by training the BP neural network; the neural network can effectively handle the nonlinear relationship between various elements of engineering construction, and the prediction results are highly accurate.

[0013] Preferably, S4 comprises the following steps: S41, the construction completion rate is the ratio of the actual completion amount of the construction to the planned completion amount of the construction. When the construction completion rate is greater than or equal to 1, the actual completion amount of the construction meets the requirements. Otherwise, the actual completion amount of the construction does not meet the requirements. The daily construction resource quantity, daily construction personnel and daily construction funds in the construction schedule are adjusted and optimized using the Golden Jackal optimization algorithm. The specific steps are as follows: S411, assigning weights to the daily construction resource volume, the daily construction personnel, and the daily construction funds, constructing a construction influencing factor function, and using the construction influencing factor function as a fitness function; setting the number of golden jackal populations to , It also represents the number of construction days. The dimension of each golden jackal in the golden jackal population is , then initialize the matrix as follows: ; in, Indicates the golden jackal population Only, dimension is of the golden jackal; Each golden jackal individual in the initialization matrix corresponds to a fitness function value; the upper limit of the position of the golden jackal in the golden jackal population is set to , the lower limit of the position of golden jackals in the golden jackal population is , represents a random number and , then the initial position of the golden jackal in the golden jackal population is The calculation formula is as follows: ; S412, the golden jackal population is in the exploration phase, and the current number of iterations is set to h , the maximum number of iterations is H , represents a random number and , the prey escape energy is ;when When the prey has enough energy to avoid the golden jackal's hunting, When the golden jackal attacks the prey, the energy calculation formula for the prey's escape is as follows: , , ; in, Represents the original energy of the prey, It indicates that the prey has decreasing energy; Introducing Levi Flight, setting represents the value of the Levy flight fitness function, then any vector of the Levy distribution , No. h The prey position vector of the iteration is , the first in the golden jackal population h The position of the male golden jackal in the next iteration is , the first in the golden jackal population h The position of the female golden jackal in the next iteration is , for the golden jackal population h The positions of the male and female golden jackals are updated in the iteration, and are recorded as and , the calculation formula is as follows: , ; The golden jackal population h +1 iteration Golden Jackal location , by iterating the position of the golden jackal, the fitness function value of the golden jackal individual in the current golden jackal population is obtained, and the optimal fitness function value is continuously sought; S413, During the attack phase, the golden jackals hunt and h In the next iteration, the male and female golden jackal positions are updated again, and the calculation formula is as follows: , ; According to the golden jackal population h The position of the male golden jackal and the female golden jackal are updated again in the iteration. h +1 iteration to update the position of the golden jackal, and continue to iterate. When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain the final position of the golden jackal. The three-dimensional coordinate values ​​of the final position of the golden jackal correspond to the daily construction resource volume, daily construction personnel and daily construction funds, and the adjusted construction schedule is obtained in turn; S42. According to the adjusted construction schedule, the daily construction resource quantity, daily construction personnel and daily construction funds are controlled to achieve construction schedule control.

[0014] The invention adjusts and optimizes the construction schedule by using the Golden Jackal optimization algorithm, and controls the construction progress by optimizing the daily construction resources, daily construction personnel and daily construction funds. The algorithm has fewer control parameters and has the advantages of strong global optimization capabilities.

[0015] This embodiment also discloses a system for a method for controlling the construction progress in engineering construction based on a drone, which specifically includes: a module for establishing a real-scene model of an engineering construction site, a module for extracting a point cloud set of buildings at the construction site, a module for predicting the completion rate of engineering construction, and a module for controlling the progress of engineering construction; The construction site real scene model building module is used to build a construction site real scene model based on the beam method and SURF algorithm; The construction site building point cloud set extraction module is used to extract the construction site building point cloud using an improved clustering algorithm to obtain a project construction progress model; The project construction completion rate prediction module is used for predicting the project construction completion rate by BP neural network; The engineering construction progress control module is used to adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm.

[0016] The present invention has the following beneficial effects: 1. The invention uses the beam method to correct the tilted image of the construction site, and uses the SURF algorithm to extract and match the feature points, which greatly improves the matching speed of the feature points; 2. The invention establishes a real-scene model of the construction site through a texture mapping method, which can establish the model quickly and efficiently.

[0017] 3. The invention extracts point clouds from construction sites, selects point cloud sets of buildings at construction sites based on an improved clustering algorithm, and adjusts the construction progress by analyzing deviations. This method effectively reduces the amount of point cloud data, and can accurately select building point clouds, thereby effectively judging the construction progress.

[0018] 4. The invention predicts the completion rate of engineering construction by training the BP neural network; the neural network can effectively handle the nonlinear relationship between various elements of engineering construction, and the prediction results are highly accurate.

[0019] 5. The invention uses the Golden Jackal optimization algorithm to adjust and optimize the construction schedule, and controls the construction progress by finding the optimal parameters. The algorithm has fewer control variables and has the advantages of strong global optimization capabilities.

[0020] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0022] Figure 1 The present invention provides a flow chart of construction progress control in engineering construction by a construction progress control system based on unmanned aerial vehicles. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0025] Example 1 This embodiment discloses a construction progress control method in engineering construction based on drones, which specifically includes the following contents: S1. The UAV obtains the inclined image of the construction site, corrects the inclined image of the construction site based on the beam method, and then uses the SURF algorithm to extract and match the feature points. After obtaining the matched inclined image of the construction site, a white film model is generated to establish a real-scene model of the construction site; The S1 comprises the following steps: S11. A control point is selected at the construction site. The drone performs oblique photography above the control point to obtain an initial oblique image of the construction site. The initial oblique image of the construction site is subjected to distortion correction and uniform light and color processing to obtain an oblique image of the construction site. A three-dimensional rectangular coordinate system is established at the construction site. The coordinates of the pixel points on the oblique image of the construction site are set as , the focal length is , the coordinates of the control points at the construction site are , the center projection coordinates of the UAV are , the coefficients corresponding to the external azimuth elements of the inclined image of the construction site are , , , , , , , and , then the collinear equation calculation formula is as follows: , ; According to the collinear equation, the coordinates of the pixel points on the inclined image of the engineering construction site are calculated, and the inclined image of the engineering construction site is corrected using the coordinates of the pixel points on the inclined image of the engineering construction site to obtain a processed inclined image of the engineering construction site; S12, selecting in the processed engineering construction site inclined image i×j Pixel points are sampled, and the pixel points are quantized to generate the processed engineering construction site tilt matrix A as follows: ; in, Indicates horizontal i , vertical j pixels; The SURF algorithm is used to detect and extract feature points of the processed engineering construction site tilt matrix, complete feature point matching, and obtain the matched engineering construction site tilt matrix. The specific steps are as follows: S121, select pixel points in the processed engineering construction site tilt matrix, calculate the integral corresponding to the pixel points and record it as an integral image, and select feature points in the integral image , calculate the feature points A Gaussian second-order differential convolution is performed at the position, and a Hessian matrix is ​​constructed using the Gaussian second-order differential convolution; a scale box filter is set, and the integral image is constructed using the scale box filter to obtain a spatial scale; the Hessian matrix determines the feature points in the spatial scale, and then the feature points are located in the processed engineering construction site tilt matrix; S122, selecting a submatrix with the feature point as the center, calculating the four-dimensional vector of the submatrix in the processed engineering construction site tilt matrix in the horizontal direction and the vertical direction, the four-dimensional vector is the feature point direction, when the traces of the Hessian matrices of the two feature points have the same sign in the feature point direction, the corresponding two feature points are matched; sequentially matching all pixel points in the processed engineering construction site tilt matrix to obtain a matched engineering construction site tilt matrix; S13, obtaining a matched tilted image of the construction site according to the matched tilted matrix of the construction site, generating ultra-high density point cloud data, forming a dense triangulated network, and then generating a white film model; using texture mapping to establish a real-scene three-dimensional model based on the white film model, the real-scene three-dimensional model being the real-scene model of the construction site; S2, extracting the construction site point cloud according to the real-scene model of the construction site, clustering the construction site point cloud based on the improved clustering algorithm, extracting the point cloud set of the construction site buildings, obtaining the construction progress model, calculating the progress deviation, performing deviation analysis, and adjusting the construction progress; The S2 comprises the following steps: S21, extracting the point cloud of the construction site according to the real-life model of the construction site, obtaining a point cloud set of the construction site, selecting any point cloud from the point cloud set of the construction site, recording it as the point cloud to be processed, establishing a neighborhood of the point cloud to be processed, and setting the neighborhood threshold to , traverse the point cloud set of the construction site and the point cloud to be processed whose Euclidean distance is less than the neighborhood threshold, and add them to the clustered point cloud set. The Euclidean distance calculation formula is as follows: ; in, , and Indicates the number of points in the construction site point cloud collection The 3D coordinates of the point cloud of the construction site, , and Represents the 3D coordinates of the point cloud to be processed, represents the Euclidean distance; A cluster point cloud set is obtained, and then any cluster point cloud is selected from the cluster point set, recorded as a to-be-processed cluster point cloud, and a point cloud whose Euclidean distance with the to-be-processed cluster point cloud is less than a neighborhood threshold is selected from the cluster point cloud set to obtain a new cluster point cloud set, and S21 is repeated until all the construction site point clouds are clustered to obtain a final cluster point cloud set; S22. Set the length, width and height thresholds of the construction site buildings to be , and If the clustered point clouds in the final clustered point cloud set meet the length, width and height thresholds of the buildings on the construction site, the final clustered point cloud set is recorded as the construction site building point cloud set; otherwise, the corresponding clustered point cloud is deleted, and S22 is repeated until the clustered point clouds in the final clustered point cloud set meet the length, width and height thresholds of the buildings on the construction site; the construction site building point cloud set is merged with the real scene model of the engineering construction site to obtain an engineering construction progress model; Obtain the construction schedule, set the expected construction schedule and construction phase according to the construction schedule, re-photograph the construction site after the construction phase, obtain a new construction schedule model, compare the construction schedule model with the new construction schedule model, analyze whether there is a deviation, and adjust the construction schedule. The specific steps are as follows: S221. Set the planned construction volume in the construction phase as The unit price of construction budget is The construction work has been completed , then the calculation formula for the construction progress deviation is as follows: , , ; in, Indicates the amount of resources consumed by the project completed during the construction phase. Indicates the amount of resources consumed by the project that has been completed during the construction phase. Indicates progress deviation; When the progress deviation is less than 0, the construction progress is behind the construction progress plan; when the progress deviation is greater than 0, the construction progress is ahead of the construction progress plan; when the progress deviation is equal to 0, there is no deviation in the construction progress; S222. If the construction progress of the project lags behind the construction progress plan and the construction period cannot be completed on schedule, the construction progress of the project will be adjusted; if the construction progress of the project is ahead of the construction progress plan, the project will be completed ahead of schedule to ensure the quality of the construction; if there is no deviation in the construction progress of the project, the construction progress of the project will not be adjusted; the construction progress adjustment of the project is completed; S3, setting a construction schedule, combining the progress deviation, training a BP neural network, obtaining a BP neural network model, and predicting the completion rate of the construction; The S3 comprises the following steps: S31. According to the construction schedule, the actual number of construction workers per day is obtained, recorded as the number of construction workers. The construction team is scored daily by professional construction personnel such as the construction project manager and engineers, with the score ranging from 0 to 1 to quantify the construction level. The daily progress deviation is calculated by the progress deviation. The construction progress data set is obtained by combining the number of construction workers, the construction level and the daily progress deviation, recorded as ,in Indicates m Number of construction workers per day, Indicates m The level of engineering construction of the day, Indicates m daily progress deviation of the day; The construction schedule of previous years was obtained, the number of construction workers, construction level and daily progress deviation of previous years were calculated, and a sample data set of construction was formed. The sample data set of construction was divided into a training set of construction and a test set of construction. The number of neurons in the input layer of the BP neural network was set to 3, the number of neurons in the hidden layer was set to 9, the number of neurons in the output layer was set to 1, and the learning rate was set to , the activation function is the sigmoid function, the engineering construction training set is input into the BP neural network, and the current number of iterations is set to , the maximum number of iterations is When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain a trained BP neural network; the engineering construction test set is then input into the trained BP neural network, and the error threshold is set to , when the output result error is less than the error threshold, the BP neural network model is obtained, otherwise the weight is adjusted until the output result error is less than the error threshold; S32, inputting the engineering construction progress data set into the BP neural network model, and the BP neural network model outputs a predicted value, which is the engineering construction completion rate; S4, combining the completion rate of the project construction, and adjusting and optimizing the project construction schedule based on the golden jackal optimization algorithm, and obtaining the adjusted project construction schedule by finding the best fitness function value, so as to realize the project construction progress control; The S4 comprises the following steps: S41, the construction completion rate is the ratio of the actual completion amount of the construction to the planned completion amount of the construction. When the construction completion rate is greater than or equal to 1, the actual completion amount of the construction meets the requirements. Otherwise, the actual completion amount of the construction does not meet the requirements. The daily construction resource quantity, daily construction personnel and daily construction funds in the construction schedule are adjusted and optimized using the Golden Jackal optimization algorithm. The specific steps are as follows: S411, assigning weights to the daily construction resource volume, the daily construction personnel, and the daily construction funds, constructing a construction influencing factor function, and using the construction influencing factor function as a fitness function; setting the number of golden jackal populations to , It also represents the number of construction days. The dimension of each golden jackal in the golden jackal population is , then initialize the matrix as follows: ; in, Indicates the golden jackal population Only, dimension is of the golden jackal; Each golden jackal individual in the initialization matrix corresponds to a fitness function value; the upper limit of the position of the golden jackal in the golden jackal population is set to , the lower limit of the position of golden jackals in the golden jackal population is , represents a random number and , then the initial position of the golden jackal in the golden jackal population is The calculation formula is as follows: ; S412, the golden jackal population is in the exploration phase, and the current number of iterations is set to h , the maximum number of iterations is H , represents a random number and , the prey escape energy is ;when When the prey has enough energy to avoid the golden jackal's hunting, When the golden jackal attacks the prey, the energy calculation formula for the prey's escape is as follows: , , ; in, Represents the original energy of the prey, It indicates that the prey has decreasing energy; Introducing Levi Flight, setting represents the value of the Levy flight fitness function, then any vector of the Levy distribution , No. h The prey position vector of the iteration is , the first in the golden jackal population h The position of the male golden jackal in the next iteration is , the first in the golden jackal population h The position of the female golden jackal in the next iteration is , for the golden jackal population h The positions of the male and female golden jackals are updated in the iteration, and are recorded as and , the calculation formula is as follows: , ; The golden jackal population h +1 iteration Golden Jackal location , by iterating the position of the golden jackal, the fitness function value of the golden jackal individual in the current golden jackal population is obtained, and the optimal fitness function value is continuously sought; S413, During the attack phase, the golden jackals hunt and h In the next iteration, the male and female golden jackal positions are updated again, and the calculation formula is as follows: , ; According to the golden jackal population h The position of the male golden jackal and the female golden jackal are updated again in the iteration. h +1 iteration to update the position of the golden jackal, and continue to iterate. When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain the final position of the golden jackal. The three-dimensional coordinate values ​​of the final position of the golden jackal correspond to the daily construction resource volume, daily construction personnel and daily construction funds, and the adjusted construction schedule is obtained in turn; S42. According to the adjusted construction schedule, the daily construction resource quantity, daily construction personnel and daily construction funds are controlled to achieve construction schedule control.

[0026] Example 2 This embodiment also discloses a system for a method for controlling the construction progress in engineering construction based on a drone, which specifically includes: a module for establishing a real-scene model of an engineering construction site, a module for extracting a point cloud set of buildings at the construction site, a module for predicting the completion rate of engineering construction, and a module for controlling the progress of engineering construction; The construction site real scene model building module is used to build a construction site real scene model based on the beam method and SURF algorithm; The construction site building point cloud set extraction module is used to extract the construction site building point cloud using an improved clustering algorithm to obtain a project construction progress model; The project construction completion rate prediction module is used for predicting the project construction completion rate by BP neural network; The engineering construction progress control module is used to adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm.

[0027] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0028] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A construction progress control method based on drones in engineering construction, characterized in that: The steps include: S1. The UAV obtains the inclined image of the construction site, corrects the inclined image of the construction site based on the beam method, and then uses the SURF algorithm to extract and match the feature points. After obtaining the matched inclined image of the construction site, a white film model is generated to establish a real-scene model of the construction site; S2, extracting the construction site point cloud according to the real-scene model of the construction site, clustering the construction site point cloud based on the improved clustering algorithm, extracting the point cloud set of the construction site buildings, obtaining the construction progress model, calculating the progress deviation, performing deviation analysis, and adjusting the construction progress; S3, setting a construction schedule, combining the progress deviation, training a BP neural network, obtaining a BP neural network model, and predicting the completion rate of the construction; S4. In combination with the completion rate of the project construction, the project construction schedule is adjusted and optimized, and the adjusted project construction schedule is obtained by finding the best fitness function value to achieve project construction progress control.

2. The method for controlling the construction progress in engineering construction based on drones according to claim 1 is characterized in that: The S1 comprises the following steps: S11, the UAV obtains an initial tilted image of the construction site, and uses a beam method to correct the tilted image of the construction site to obtain a processed tilted image of the construction site; S12, generating a processed engineering construction site tilt matrix according to the processed engineering construction site tilt image; using the SURF algorithm to detect and extract feature points of the processed engineering construction site tilt matrix, completing feature point matching, and obtaining a matched engineering construction site tilt matrix; S13, obtaining a matched tilted image of the construction site according to the matched tilted matrix of the construction site, generating ultra-high density point cloud data, forming a dense triangulated network, and then generating a white film model; and establishing a real-scene model of the construction site using texture mapping based on the white film model.

3. The method for controlling the construction progress in engineering construction based on drone according to claim 2 is characterized in that: The SURF algorithm is used to detect and extract feature points from the processed engineering construction site tilt matrix, and the feature point matching is completed, which includes the following steps: Selecting pixel points in the processed engineering construction site tilt matrix, constructing a Hessian matrix based on Gaussian second-order differential convolution, determining the position of feature points in the Hessian matrix, and then locating the feature points to the processed engineering construction site tilt matrix; A submatrix is ​​selected with the feature point as the center, and the direction of the feature point is calculated. When the traces of the Hessian matrices of the two feature points have the same sign in the direction of the feature point, the corresponding two feature points are matched; all pixel points in the processed engineering construction site tilt matrix are matched in turn to obtain the matched engineering construction site tilt matrix.

4. The method for controlling the construction progress in engineering construction based on drones according to claim 3 is characterized in that: The S2 comprises the following steps: S21, extracting the construction site point cloud according to the construction site real scene model, clustering the construction site point cloud based on the improved clustering algorithm to obtain a final clustered point cloud set; S22. Set a threshold value for the construction site buildings, compare it with the final clustered point cloud set, extract the point cloud set of the construction site buildings, merge the point cloud set of the construction site buildings with the real-life model of the construction site, and obtain a construction progress model; calculate the progress deviation by comparing the construction progress model, perform deviation analysis, and adjust the construction progress.

5. The method for controlling the construction progress in engineering construction based on drones according to claim 4 is characterized in that: The adjustment of the construction progress includes the following steps: When the progress deviation is less than 0, the project construction progress lags behind the project construction schedule and the construction period cannot be completed on schedule, so the project construction progress is adjusted; when the progress deviation is greater than 0, the project construction progress is ahead of the project construction schedule to ensure the quality of the project construction; when the progress deviation is equal to 0, no adjustment is made to the project construction progress; the project construction progress adjustment is completed.

6. A method for controlling the construction progress in engineering construction based on drones according to claim 5, characterized in that: The S3 comprises the following steps: S31. According to the construction schedule, the number of construction workers, the construction level and the daily progress deviation are obtained to form a construction progress data set; The construction schedule of previous years was obtained, the number of construction workers, construction level and daily progress deviation of previous years were calculated, and a sample data set of construction was formed. The sample data set of construction was divided into a training set of construction and a test set of construction. The number of neurons in the input layer of the BP neural network was set to 3, the number of neurons in the hidden layer was set to 9, the number of neurons in the output layer was set to 1, and the learning rate was set to , the activation function is the sigmoid function, the engineering construction training set is input into the BP neural network, and the current number of iterations is set to , the maximum number of iterations is When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain a trained BP neural network; the engineering construction test set is then input into the trained BP neural network, and the error threshold is set to , when the output result error is less than the error threshold, the BP neural network model is obtained, otherwise the weight is adjusted until the output result error is less than the error threshold; S32, inputting the engineering construction progress data set into the BP neural network model, and the BP neural network model outputs a predicted value, which is the engineering construction completion rate.

7. The method for controlling the construction progress in engineering construction based on drones according to claim 4 is characterized in that: The S4 comprises the following steps: S41, when the project construction completion rate is less than 1, adjusting and optimizing the project construction schedule based on the golden jackal optimization algorithm to obtain an adjusted project construction schedule; S42. According to the adjusted construction schedule, the daily construction resource quantity, daily construction personnel and daily construction funds are controlled to achieve construction schedule control.

8. The method for controlling the construction progress in engineering construction based on drones according to claim 7 is characterized in that: The S41 comprises the following steps: S411, assigning weights to the daily construction resource volume, the daily construction personnel, and the daily construction funds, constructing a construction influencing factor function, and using the construction influencing factor function as a fitness function; setting the number of golden jackal populations to , It also represents the number of construction days. The dimension of each golden jackal in the golden jackal population is , establish the initialization matrix; Each golden jackal individual in the initialization matrix corresponds to a fitness function value; the upper limit of the position of the golden jackal in the golden jackal population is set to , the lower limit of the position of golden jackals in the golden jackal population is , represents a random number and , then the initial position of the golden jackal in the golden jackal population is The calculation formula is as follows: ； S412, the golden jackal population is in the exploration phase, and the current number of iterations is set to h , the maximum number of iterations is H , represents a random number and , the prey escape energy is , the prey escape energy calculation formula is as follows: ， ， ； in, Represents the original energy of the prey, Indicates that the prey is decreasing in energy; Introducing Levi Flight, setting represents the value of the Levy flight fitness function, then any vector of the Levy distribution , No. h The prey position vector of the iteration is , the first in the golden jackal population h The position of the male golden jackal in the next iteration is , the first in the golden jackal population h The position of the female golden jackal in the next iteration is , for the golden jackal population h The positions of the male and female golden jackals are updated in the iteration, and are recorded as and , the calculation formula is as follows: ， ； The golden jackal population h +1 iteration Golden Jackal location , by iterating the position of the golden jackal, the fitness function value of the golden jackal individual in the current golden jackal population is obtained, and the optimal fitness function value is continuously sought; S413, During the attack phase, the golden jackals hunt and h In the next iteration, the male and female golden jackal positions are updated again, and the calculation formula is as follows: ， ； According to the golden jackal population h The position of the male golden jackal and the female golden jackal are updated again in the iteration. h The position of the golden jackal is updated after +1 iteration, and the iteration is continued. When the current number of iterations reaches the maximum number of iterations, the iteration is stopped to obtain the final position of the golden jackal. The three-dimensional coordinate values ​​of the final position of the golden jackal correspond to the daily construction resource volume, the daily number of construction personnel and the daily construction funds, respectively, and the adjusted construction schedule is obtained in turn.

9. A system for implementing the method for controlling the construction progress in engineering construction based on drones as described in any one of claims 1 to 8, characterized in that: Specifically include: The module for establishing the real-scene model of the construction site, the module for extracting the point cloud of the buildings at the construction site, the module for predicting the completion rate of the construction project, and the module for controlling the progress of the construction project; The construction site real scene model building module is used to build a construction site real scene model based on the beam method and SURF algorithm; The construction site building point cloud set extraction module is used to extract the construction site building point cloud using an improved clustering algorithm to obtain a project construction progress model; The project construction completion rate prediction module is used for predicting the project construction completion rate by BP neural network; The engineering construction progress control module is used to adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm.

Images