A Construction Progress Regulation System and Method in Engineering Construction Based on UAVs

Abstract

The present invention relates to the technical field of construction progress regulation, and discloses a construction progress regulation system and method in engineering construction based on an unmanned aerial vehicle. First, the present invention obtains an inclined image of an engineering construction site, uses the bundle adjustment method and the SURF algorithm to extract feature points and match the feature points, and establishes a real scene model of the engineering construction site; secondly, combines the real scene model of the engineering construction site to cluster the point cloud of the engineering construction site, extracts the point cloud set of the buildings on the construction site, conducts deviation analysis, and adjusts the engineering construction progress; then trains a BP neural network to predict the completion rate of the engineering construction; finally, based on the completion rate of the engineering construction, adjusts and optimizes the engineering construction progress plan based on the golden jackal optimization algorithm to achieve the control of the engineering construction progress. The present invention achieves the purpose of construction progress regulation 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 regulation, and particularly to a construction progress regulation system and method in engineering construction based on an unmanned aerial vehicle (UAV). 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 foundation pit detection data; a data analysis module, which preprocesses the foundation pit detection data and obtains a trend fitting curve of dynamic changes, and performs detection analysis and subsequent data correction on the foundation pit construction through the trend fitting curve; a theoretical analysis interaction module, which compares and analyzes the detection data and theoretical calculation values, selects the optimal calculation model, predicts the change of the foundation pit state in subsequent construction, and predicts the foundation pit segmentation and foundation pit zoning at the same time; a dynamic construction control module, which calculates an alarm value and obtains the dynamic construction change trend of each construction stage, foundation pit segmentation, and foundation pit zoning of the foundation pit through the predicted value; and an alarm module, which automatically alarms when the alarm value exceeds the alarm threshold during the dynamic construction process of the foundation pit.

[0003] Traditional construction progress regulation methods regulate construction progress through on-site command methods. There are problems such as low information transmission efficiency and uncontrollable regulation processes during the process. Moreover, the unreasonable allocation of construction resources brings trouble to the progress of engineering construction. At the same time, without using high-tech such as neural networks, the construction progress and real-time performance of engineering construction cannot be guaranteed. Summary of the Invention

[0004] In view of the problems in the related art, the present invention provides a construction progress regulation system and method in engineering construction based on an unmanned aerial vehicle (UAV) to overcome the above-mentioned technical problems existing in the existing related technologies.

[0005] To solve the above technical problems, the present invention is realized through the following technical solutions:

[0006] The present invention provides a construction progress regulation method in engineering construction based on an unmanned aerial vehicle (UAV), including the following steps:

[0007] S1. The UAV acquires an inclined image of the engineering construction site, corrects the inclined image of the engineering construction site based on the bundle adjustment method, then uses the SURF algorithm to extract feature points and match the feature points. After obtaining the matched inclined image of the engineering construction site, a white film model is generated, and a real scene model of the engineering construction site is established;

[0008] S2. Extract the point cloud of the engineering construction site according to the real scene model of the engineering construction site, cluster the point cloud of the engineering construction site based on an improved clustering algorithm, extract the point cloud set of the buildings in the construction site, obtain the engineering construction progress model, calculate the progress deviation for deviation analysis, and adjust the engineering construction progress;

[0009] S3. Set the engineering construction progress plan, combine the progress deviation, train a BP neural network to obtain a BP neural network model, and predict the completion rate of the engineering construction;

[0010] S4. Combine the completion rate of the engineering construction, and adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm. Obtain the adjusted engineering construction progress plan by finding the optimal fitness function value, and realize the control of the engineering construction progress.

[0011] The present invention obtains the oblique image of the engineering construction site through a drone, corrects the oblique image of the engineering construction site using the bundle adjustment method, extracts feature points through the SURF algorithm, matches the feature points of the oblique image of the engineering construction site, generates a white film model, and then establishes a real scene model of the engineering construction site through texture mapping; secondly, extract the point cloud of the engineering construction site, cluster based on an improved clustering algorithm, and select the point cloud set of the buildings in the construction site by setting a threshold, so as to determine the engineering construction progress model. Analyze the deviation to adjust the engineering construction progress by comparing the daily construction progress of the engineering construction progress model; this method effectively reduces the data volume of the point cloud, accurately selects the buildings, and judges the construction progress through the point cloud set of the buildings in the construction site; then combine the number of construction workers, the engineering construction level and the daily progress deviation to train a BP neural network to predict the future completion rate of the engineering construction; this neural network can effectively process the non-linear relationship between various elements of the engineering construction, and the prediction result has high accuracy; finally, adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm, optimize the daily engineering construction resource quantity, the daily number of construction workers and the daily engineering construction funds, and complete the control of the engineering construction progress; this algorithm imitates the cooperative hunting behavior of the golden jackal to find the optimal solution to the problem, and has the advantages of strong global optimization ability, etc.

[0012] Preferably, the S1 includes the following steps:

[0013] S11. Select control points in the engineering construction site, the drone conducts oblique photography above the control points to obtain the initial oblique image of the engineering construction site. After performing distortion correction processing and equalization of illumination and color on the initial oblique image of the engineering construction site, obtain the oblique image of the engineering construction site; establish a three-dimensional rectangular coordinate system in the engineering construction site, and set the pixel point coordinates on the oblique image of the engineering construction site as , the focal length is , the coordinates of the control points in the engineering construction site are , the central projection coordinates of the drone are The coefficients corresponding to the exterior orientation angle elements of the inclined image of the engineering construction site are respectively , , , , , , , and , then the collinearity equation calculation formula is as follows: ,

[0014] ;

[0015] According to the collinearity equation, the pixel coordinates on the inclined image of the engineering construction site are calculated, and the inclined image of the engineering construction site is corrected using the pixel coordinates on the inclined image of the engineering construction site to obtain the processed inclined image of the engineering construction site;

[0016] S12. Select i×j pixel points in the processed inclined image of the engineering construction site for sampling, perform quantization processing on the pixel points, and generate the processed inclined matrix of the engineering construction site A as follows:

[0017] ;

[0018] Among them, represents the i th pixel point in the horizontal direction and the j th pixel point in the vertical direction;

[0019] Use the SURF (Speeded Up Robust Features) algorithm to detect and extract feature points from the processed inclined matrix of the engineering construction site, complete feature point matching, and obtain the matched inclined matrix of the engineering construction site. The specific steps are as follows:

[0020] S121. Select pixel points in the processed inclined matrix of the engineering construction site, calculate the integral corresponding to the pixel points and record it as the integral image, and select feature points in the integral image, calculate the second-order Gaussian differential convolution at the feature point , and use the second-order Gaussian differential convolution to construct the Hessian matrix; set the scale box filter, use the scale box filter to construct the integral image to obtain the spatial scale; the Hessian matrix determines the feature points in the spatial scale, and then locates the feature points in the processed inclined matrix of the engineering construction site;

[0021] S122. Select a sub - matrix centered on the feature point, calculate the four - dimensional vectors of the sub - matrix in the processed tilt matrix of the engineering construction site in the horizontal and vertical directions. The four - dimensional vector is the feature point direction. When the signs of the traces of the Hessian matrices of two feature points are the same in the feature point direction, the corresponding two feature points are matched. Match all the pixel points in the processed tilt matrix of the engineering construction site in turn to obtain the matched tilt matrix of the engineering construction site.

[0022] S13. Obtain the matched tilt image of the engineering construction site according to the matched tilt matrix of the engineering construction site, generate ultra - high - density point cloud data, construct a dense triangular mesh, and then generate a white film model. Based on the white film model, use texture mapping to establish a real - scene three - dimensional model, which is the real - scene model of the engineering construction site.

[0023] The invention obtains the tilt image of the engineering construction site through a drone, corrects the tilt image of the engineering construction site using the bundle adjustment method, extracts feature points, then matches the feature points of the tilt image of the engineering construction site, generates a white film model, and then establishes a real - scene model of the engineering construction site through texture mapping.

[0024] Preferably, S2 includes the following steps:

[0025] S21. Extract the point cloud of the engineering construction site according to the real - scene model of the engineering construction site to obtain a point cloud set of the engineering construction site. Select any point cloud in the point cloud set of the engineering construction site, denote it as the point cloud to be processed, establish the neighborhood of the point cloud to be processed, and set the neighborhood threshold as . Traverse the point clouds in the point cloud set of the engineering construction site whose Euclidean distance from the point cloud to be processed is less than the neighborhood threshold, and add them to the clustered point cloud set. The Euclidean distance calculation formula is as follows:

[0026] ;

[0027] where 、 and represent the three - dimensional coordinates of the th point cloud in the point cloud set of the engineering construction site, 、 and represent the three - dimensional coordinates of the point cloud to be processed, and represents the Euclidean distance.

[0028] Obtain the clustered point cloud set, and then select any clustered point cloud from the clustered point set, denoted as the to-be-processed clustered point cloud. Select the point clouds in the clustered point cloud set whose Euclidean distance from the to-be-processed clustered point cloud is less than the neighborhood threshold to obtain a new clustered point cloud set. Repeat S21 until all the point clouds of the engineering construction site are clustered, and obtain the final clustered point cloud set;

[0029] S22. Set the length, width, and height thresholds of the construction site building to be , and , respectively. If the clustered point cloud in the final clustered point cloud set meets the length, width, and height thresholds of the construction site building, then denote the final clustered point cloud set as the construction site building point cloud set; otherwise, delete the corresponding clustered point cloud. Repeat S22 until the clustered point cloud in the final clustered point cloud set meets the length, width, and height thresholds of the construction site building; Integrate the construction site building point cloud set and the engineering construction site real scene model to obtain the engineering construction progress model;

[0030] Obtain the engineering construction progress plan. Set the expected engineering construction progress and engineering construction stage according to the engineering construction progress plan. After the engineering construction stage, the unmanned aerial vehicle re - photographs the engineering construction site to obtain a new engineering construction progress model. Compare the engineering construction progress model and the new engineering construction progress model, analyze whether there is a deviation, and adjust the engineering construction progress. The specific steps are as follows:

[0031] S221. Set that in the engineering construction stage, the planned engineering construction quantity is , the budget unit price of the engineering construction is , and the completed engineering construction quantity is . Then the engineering construction progress deviation calculation formula is as follows:

[0032] , , ;

[0033] Among them, represents the engineering consumption resources quantity completed in the engineering construction stage, represents the engineering consumption resources quantity that has been completed in the engineering construction stage, represents the progress deviation;

[0034] When the progress deviation is less than 0, at this time the engineering construction progress lags behind the engineering construction progress plan. When the progress deviation is greater than 0, at this time the engineering construction progress is ahead of the engineering construction progress plan. When the progress deviation is equal to 0, at this time the engineering construction progress has no deviation;

[0035] S222. If the engineering construction progress lags behind the engineering construction progress plan and the project cannot be completed on schedule, adjust the engineering construction progress; if the engineering construction progress is ahead of the engineering construction progress plan and the project is completed ahead of schedule, ensure the engineering construction quality; if there is no deviation in the engineering construction progress, do not adjust the engineering construction progress; complete the adjustment of the engineering construction progress.

[0036] The invention extracts the point cloud of the engineering construction site, selects the point cloud set of the buildings at the construction site based on the improved clustering algorithm, determines the engineering construction progress model, compares the daily construction progress of the engineering construction progress model, analyzes the deviation, and realizes the adjustment of the engineering construction progress; this method greatly reduces the data volume of the point cloud, can accurately select the point cloud of the building, and thus effectively judge the construction progress.

[0037] Preferably, the said S3 includes the following steps:

[0038] S31. Obtain the actual number of daily construction workers according to the engineering construction progress plan, denoted as the number of construction workers. Score the engineering construction team daily by professional construction personnel such as the engineering construction project manager and engineer, with the score range between 0 and 1 to quantify the engineering construction level. Calculate the daily progress deviation through the said progress deviation. Combine the number of construction workers, the engineering construction level, and the daily progress deviation to obtain the engineering construction progress data set, denoted as , where represents the number of construction workers on the m th day, represents the engineering construction level on the m th day, represents the daily progress deviation on the m th day;

[0039] Obtain the engineering construction progress plan in previous years, calculate the number of construction workers in previous years, the engineering construction level in previous years, and the daily progress deviation in previous years, and form the engineering construction sample data set. Divide the said engineering construction sample data set into the engineering construction training set and the engineering construction test set; set the number of neurons in the input layer of the BP neural network to 3, the number of neurons in the hidden layer to 9, the number of neurons in the output layer to 1, the learning rate to , the activation function is the sigmoid function. Input the said engineering construction training set into the BP (backpropagation) neural network, set the current number of iterations to , the maximum number of iterations to . When the current number of iterations reaches the maximum number of iterations, stop the iteration to obtain the trained BP neural network; then input the said engineering construction test set into the trained BP neural network, set the error threshold to . When the error of the output result is less than the error threshold, obtain the BP neural network model, otherwise adjust the weights until the error of the output result is less than the error threshold;

[0040] S32. Input 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.

[0041] This invention predicts the engineering construction completion rate by training the BP neural network; this neural network can effectively handle the non-linear relationship between various elements of the engineering construction, and the prediction result has high accuracy.

[0042] Preferably, step S4 includes the following steps:

[0043] S41. The engineering construction completion rate is the ratio of the actual completed quantity of the engineering construction to the planned completed quantity of the engineering construction. When the engineering construction completion rate is greater than or equal to 1, the actual completed quantity of the engineering construction meets the requirements; otherwise, the actual completed quantity of the engineering construction does not meet the requirements. Use the golden jackal optimization algorithm to adjust and optimize the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds in the engineering construction progress plan. The specific steps are as follows:

[0044] S411. Assign weights to the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds, construct an engineering construction influence factor function, and use the engineering construction influence factor function as the fitness function; set the number of golden jackal populations to , which also represents the number of engineering construction days. Each dimension of the golden jackal in the golden jackal population is , then initialize the matrix as follows:

[0045] ;

[0046] where, represents the th golden jackal in the golden jackal population with a dimension of ;

[0047] Each golden jackal individual in the initialized matrix corresponds to a fitness function value; set the upper limit of the golden jackal position in the golden jackal population to , and the lower limit of the golden jackal position in the golden jackal population to , represents a random number and , then the initial position of the golden jackal in the golden jackal population is calculated as follows:

[0048] ;

[0049] S412. In the exploration stage of the golden jackal population, set the current iteration number to h , the maximum iteration number to H , represents a random number and , the prey escape energy is ; when , the prey has enough energy to avoid being hunted by the golden jackal. When , the golden jackal launches a hunt on the prey. The formula for calculating the prey escape energy is as follows:

[0050] , , ;

[0051] Among them, represents the original energy of the prey, represents the decreasing energy of the prey;

[0052] Levy flight is introduced. Let represent the fitness function value of Levy flight. Then, for any vector of the Levy distribution, the position vector of the prey at the h -th iteration is . The position of the male golden jackal at the h -th iteration in the golden jackal population is . The position of the female golden jackal at the h -th iteration in the golden jackal population is . The positions of the male and female golden jackals at the h -th iteration in the golden jackal population are updated, denoted as and respectively. The calculation formulas are as follows:

[0053] ,

[0054] ;

[0055] Then, the position of the golden jackal at the h + 1-th iteration in the golden jackal population is . By iterating the position of the golden jackal, the fitness function values of the golden jackal individuals in the current golden jackal population are obtained, and the best fitness function value is continuously searched for;

[0056] S413. In the attack stage of the golden jackal population, the golden jackals conduct hunting. The positions of the male and female golden jackals at the h -th iteration in the golden jackal population are updated again. The calculation formulas are as follows:

[0057] ,

[0058] ;

[0059] According to the positions of the male and female golden jackals at the h -th iteration and their updated positions, the positions of the male and female golden jackals at the hThe position of the golden jackal in the +1st iteration is updated, and the iteration is continuously carried out. When the current iteration number reaches the maximum iteration number, 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 respectively correspond to the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds, and the adjusted engineering construction progress plan is obtained in sequence;

[0060] S42. According to the adjusted engineering construction progress plan, control the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds to achieve the control of the engineering construction progress.

[0061] The invention adjusts and optimizes the engineering construction progress plan by using the golden jackal optimization algorithm, and completes the control of the engineering construction progress by optimizing the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds; the algorithm has fewer control parameters and has the advantages of strong global optimization ability, etc.

[0062] This embodiment also discloses a system for a construction progress regulation and control method in engineering construction based on an unmanned aerial vehicle, which specifically includes: an engineering construction site real scene model establishment module, a construction site building point cloud set extraction module, an engineering construction completion rate prediction module, and an engineering construction progress control module;

[0063] The engineering construction site real scene model establishment module is used to construct an engineering construction site real scene model based on the bundle adjustment method and the SURF algorithm;

[0064] The construction site building point cloud set extraction module is used to extract the construction site building point cloud by using an improved clustering algorithm to obtain an engineering construction progress model;

[0065] The engineering construction completion rate prediction module is used to predict the engineering construction completion rate by using a BP neural network;

[0066] The engineering construction progress control module is used to adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm.

[0067] The present invention has the following beneficial effects:

[0068] 1. The invention corrects the inclined image of the engineering construction site by using the bundle adjustment method, and uses the SURF algorithm to extract and match the feature points, which greatly improves the matching speed of the feature points;

[0069] 2. The invention establishes an engineering construction site real scene model by using the method of texture mapping, and this method can establish the model quickly and efficiently.

[0070] 3. The invention extracts the point cloud of the engineering construction site, selects the point cloud set of the buildings on the construction site based on an improved clustering algorithm, and adjusts the engineering construction progress by analyzing the deviation; this method effectively reduces the data volume of the point cloud, can accurately select the building point cloud, and thus effectively judge the construction progress.

[0071] 4. The invention trains a BP neural network to predict the completion rate of engineering construction; this neural network can effectively process the non-linear relationship between various elements of engineering construction, and the prediction result has high accuracy.

[0072] 5. The invention adjusts and optimizes the engineering construction progress plan by using the golden jackal optimization algorithm, and realizes the control of the engineering construction progress by finding the optimal parameters; this algorithm has fewer control variables and has the advantages of strong global optimization ability, etc.

[0073] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. Brief Description of the Drawings

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0075] Figure 1 It is a schematic flow chart of the construction progress regulation of an engineering construction by a construction progress regulation system based on an unmanned aerial vehicle according to the present invention. Detailed Embodiments

[0076] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0077] In the description of the present invention, it should be understood that the terms "openings", "upper", "lower", "top", "middle", "inner", etc. indicating the orientation or position relationship are only for the convenience of describing the invention and simplifying the description, rather than indicating or implying that the components or elements referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as a limitation of the invention.

[0078] Embodiment 1

[0079] This embodiment discloses a construction progress regulation method in engineering construction based on an unmanned aerial vehicle, which specifically includes the following contents:

[0080] S1. The drone acquires the tilted image of the engineering construction site, corrects the tilted image of the engineering construction site based on the bundle adjustment method, then uses the SURF algorithm to extract feature points and match the feature points. After obtaining the tilted image of the engineering construction site after matching, a white film model is generated and a real-scene model of the engineering construction site is established;

[0081] The said S1 includes the following steps:

[0082] S11. Select control points in the engineering construction site. The drone conducts oblique photography above the control points to obtain the initial tilted image of the engineering construction site. After performing distortion correction processing and equalization of illumination and color on the initial tilted image of the engineering construction site, the tilted image of the engineering construction site is obtained; Establish a three-dimensional rectangular coordinate system in the engineering construction site, and set the pixel point coordinates on the tilted image of the engineering construction site as , the focal length is , the control point coordinates in the engineering construction site are , the central projection coordinates of the drone are , and the coefficients corresponding to the exterior orientation angle elements of the tilted image of the engineering construction site are respectively , , , , , , , and , then the collinearity equation calculation formula is as follows: ,

[0083] ;

[0084] According to the said collinearity equation, calculate the pixel point coordinates on the tilted image of the engineering construction site, and use the pixel point coordinates on the tilted image of the engineering construction site to correct the tilted image of the engineering construction site to obtain the processed tilted image of the engineering construction site;

[0085] S12. Select i×j pixel points in the processed tilted image of the engineering construction site for sampling, perform quantization processing on the pixel points, and generate the processed tilted matrix A of the engineering construction site as follows:

[0086] ;

[0087] Among them, represents the i th pixel point in the horizontal direction and the j th pixel point in the vertical direction;

[0088] Use the SURF algorithm to detect and extract feature points from the processed tilt matrix of the engineering construction site, complete feature point matching, and obtain the tilt matrix of the engineering construction site after matching. The specific steps are as follows:

[0089] S121. Select pixel points in the processed tilt matrix of the engineering construction site, calculate the integral corresponding to the pixel points and record it as the integral image, and select feature points in the integral image , calculate the feature points Perform Gaussian second-order differential convolution at, use the Gaussian second-order differential convolution to construct a Hessian matrix; set a scale box filter, use the scale box filter to construct the integral image, and obtain the spatial scale; the Hessian matrix determines the feature points in the spatial scale, and then locate the feature points to the processed tilt matrix of the engineering construction site;

[0090] S122. Select a submatrix centered on the feature points, calculate the four-dimensional vectors of the submatrix in the horizontal and vertical directions in the processed tilt matrix of the engineering construction site. The four-dimensional vectors are the feature point directions. When the signs of the traces of the Hessian matrices of two feature points are the same in the feature point directions, the corresponding two feature points are matched; match all the pixel points in the processed tilt matrix of the engineering construction site in turn to obtain the tilt matrix of the engineering construction site after matching;

[0091] S13. Obtain the tilt image of the engineering construction site after matching according to the tilt matrix of the engineering construction site after matching, generate ultra-high density point cloud data, form a dense triangular network, and then generate a white film model; use texture mapping based on the white film model to establish a real scene three-dimensional model. The real scene three-dimensional model is the real scene model of the engineering construction site;

[0092] S2. Extract the point cloud of the engineering construction site according to the real scene model of the engineering construction site, perform clustering on the point cloud of the engineering construction site based on an improved clustering algorithm, extract the point cloud set of the buildings on the construction site, obtain the engineering construction progress model, calculate the progress deviation for deviation analysis, and adjust the engineering construction progress;

[0093] The S2 includes the following steps:

[0094] S21. Extract the point cloud of the engineering construction site according to the real scene model of the engineering construction site to obtain the point cloud set of the engineering construction site. Select any point cloud in the point cloud set of the engineering construction site, record it as the point cloud to be processed, establish the neighborhood of the point cloud to be processed, and set the neighborhood threshold to , traverse the point clouds in the point cloud set of the engineering construction site whose Euclidean distance from the point cloud to be processed is less than the neighborhood threshold, and add them to the clustering point cloud set. The Euclidean distance calculation formula is as follows:

[0095] ;

[0096] Among them, 、 and represent the three-dimensional coordinates of the point cloud of the th construction site in the point cloud set of the construction site, 、 and represent the three-dimensional coordinates of the point cloud to be processed, represents the Euclidean distance;

[0097] Obtain the clustered point cloud set, then select any clustered point cloud from the clustered point set, denote it as the clustered point cloud to be processed, and select the point clouds in the clustered point cloud set whose Euclidean distance from the clustered point cloud to be processed is less than the neighborhood threshold to obtain a new clustered point cloud set. Repeat S21 until all the point clouds of the construction site are clustered to obtain the final clustered point cloud set;

[0098] S22. Set the length, width, and height thresholds of the construction site building to 、 and respectively. If the clustered point cloud in the final clustered point cloud set meets the length, width, and height thresholds of the construction site building, then denote the final clustered point cloud set as the point cloud set of the construction site building; otherwise, delete the corresponding clustered point cloud. Repeat S22 until the clustered point cloud in the final clustered point cloud set meets the length, width, and height thresholds of the construction site building; Integrate the point cloud set of the construction site building and the actual scene model of the construction site to obtain the construction progress model;

[0099] Obtain the construction progress plan, set the expected construction progress and construction stages according to the construction progress plan. After the construction stage, the unmanned aerial vehicle re - shoots the construction site to obtain a new construction progress model. Compare the construction progress model and the new construction progress model, analyze whether there is a deviation, and adjust the construction progress. The specific steps are as follows:

[0100] S221. Set that in the construction stage, the planned construction quantity is , the budget unit price of the construction is , and the completed construction quantity is . Then the calculation formula for the construction progress deviation is as follows:

[0101] , , ;

[0102] Among them, represents the amount of resources consumed for the project completed in the construction stage, Indicates the amount of resources consumed in the project construction stage that has been completed. Indicates the schedule deviation.

[0103] When the schedule deviation is less than 0, the project construction progress lags behind the project construction schedule. When the schedule deviation is greater than 0, the project construction progress is ahead of the project construction schedule. When the schedule deviation is equal to 0, there is no deviation in the project construction progress.

[0104] S222. If the project construction progress lags behind the project construction schedule and the project duration cannot end on schedule, then adjust the project construction progress. If the project construction progress is ahead of the project construction schedule and the project is completed ahead of schedule, ensure the project construction quality. If there is no deviation in the project construction progress, then do not adjust the project construction progress. Complete the adjustment of the project construction progress.

[0105] S3. Set the project construction schedule, combine the schedule deviation, train the BP neural network, obtain the BP neural network model, and predict the project construction completion rate.

[0106] The S3 includes the following steps:

[0107] S31. Obtain the actual number of construction workers per day according to the project construction schedule, denoted as the number of construction workers. Score the project construction team daily by professional construction personnel such as the project construction manager and engineer. The score range is between 0 and 1 to quantify the project construction level. Calculate the daily schedule deviation through the schedule deviation. Combine the number of construction workers, the project construction level, and the daily schedule deviation to obtain the project construction progress data set, denoted as , where Indicates the number of construction workers on the m th day, Indicates the project construction level on the m th day, Indicates the daily schedule deviation on the m th day;

[0108] Obtain the project construction schedules of previous years, calculate the number of construction workers in previous years, the project construction level in previous years, and the daily schedule deviation in previous years, and form the project construction sample data set. Divide the project construction sample data set into the project construction training set and the project construction test set. Set the number of neurons in the input layer of the BP neural network to 3, the number of neurons in the hidden layer to 9, the number of neurons in the output layer to 1, the learning rate to , the activation function is the sigmoid function, input the project construction training set into the BP neural network, set the current number of iterations to , and the maximum number of iterations to , when the current iteration number reaches the maximum iteration number, stop the iteration to obtain the trained BP neural network; then input the engineering construction test set into the trained BP neural network, and set the error threshold to , when the error of the output result is less than the error threshold, obtain the BP neural network model, otherwise adjust the weights until the error of the output result is less than the error threshold;

[0109] S32. Input 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;

[0110] S4. Combine the engineering construction completion rate, and based on the golden jackal optimization algorithm, adjust and optimize the engineering construction progress plan, and obtain the adjusted engineering construction progress plan by finding the best fitness function value to achieve the control of the engineering construction progress;

[0111] The S4 includes the following steps:

[0112] S41. The engineering construction completion rate is the ratio of the actual completed quantity of the engineering construction to the planned completed quantity of the engineering construction. When the engineering construction completion rate is greater than or equal to 1, the actual completed quantity of the engineering construction meets the requirements, otherwise the actual completed quantity of the engineering construction does not meet the requirements. Use the golden jackal optimization algorithm to adjust and optimize the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds in the engineering construction progress plan. The specific steps are as follows:

[0113] S411. Assign weights to the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds, construct an engineering construction influencing factor function, and use the engineering construction influencing factor function as the fitness function; set the number of golden jackal populations to , At the same time, it represents the number of days of the engineering construction. Each dimension of the golden jackal in the golden jackal population is , then initialize the matrix as follows:

[0114] ;

[0115] Among them, represents the th golden jackal in the golden jackal population with a dimension of ;

[0116] Each golden jackal individual in the initialized matrix corresponds to a fitness function value; set the upper limit of the golden jackal position in the golden jackal population to , and the lower limit of the golden jackal position in the golden jackal population to , represents a random number and , then the initial position of the golden jackal in the golden jackal population The calculation formula is as follows:

[0117] ;

[0118] S412. During the exploration stage of the golden jackal population, set the current iteration number as h , the maximum iteration number as H , represents a random number and , the prey escape energy is ; When , the prey has enough energy to avoid the golden jackal's hunting. When , the golden jackal launches a hunt on the prey. The calculation formula for the prey escape energy is as follows:

[0119] , , ;

[0120] Among them, represents the original energy of the prey, represents the decreasing energy of the prey;

[0121] Introduce Levy flight, set represents the Levy flight fitness function value. Then, for any vector of the Levy distribution, the position vector of the prey at the h -th iteration is , the position of the male golden jackal at the h -th iteration in the golden jackal population is , the position of the female golden jackal at the h -th iteration in the golden jackal population is . Update the positions of the male and female golden jackals at the h -th iteration in the golden jackal population, and denote them as and respectively. The calculation formula is as follows:

[0122] ,

[0123] ;

[0124] Then, the position of the golden jackal at the h +1-th iteration in the golden jackal population is . Obtain the fitness function values of the golden jackal individuals in the current golden jackal population by iterating the golden jackal positions, and continuously search for the best fitness function value;

[0125] S413. During the attack stage of the golden jackal population, the golden jackals conduct hunting. Update the positions of the male and female golden jackals at the h -th iteration in the golden jackal population again. The calculation formula is as follows:

[0126] ,

[0127] ;

[0128] According to the positions of male golden jackals and the updated positions of female golden jackals in the h th iteration of the golden jackal population, update the positions of the golden jackal population in the h +1th iteration. Keep iterating. When the current iteration number reaches the maximum iteration number, stop iterating to obtain the final positions of the golden jackals. The three-dimensional coordinate values of the final positions of the golden jackals respectively correspond to the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds, and then obtain the adjusted engineering construction progress plan in sequence;

[0129] S42. Control the daily engineering construction resource quantity, the daily number of construction workers, and the daily engineering construction funds according to the adjusted engineering construction progress plan to achieve the control of the engineering construction progress.

[0130] Embodiment 2

[0131] This embodiment also discloses a system for regulating and controlling the construction progress in engineering construction based on drones, specifically including: an engineering construction site real-scene model establishment module, a construction site building point cloud set extraction module, an engineering construction completion rate prediction module, and an engineering construction progress control module;

[0132] The engineering construction site real-scene model establishment module is used to construct an engineering construction site real-scene model based on the bundle adjustment method and the SURF algorithm;

[0133] 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 an engineering construction progress model;

[0134] The engineering construction completion rate prediction module is used to predict the engineering construction completion rate by a BP neural network;

[0135] The engineering construction progress control module is used to adjust and optimize the engineering construction progress plan based on the golden jackal optimization algorithm.

[0136] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0137] The preferred embodiments of the invention disclosed above are only used to help illustrate the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of this specification. These embodiments are selected and specifically described in this specification to better explain the principle and practical application of the invention, so that those skilled in the art can well understand and utilize the invention.

Claims

1. A method for regulating construction progress in engineering construction based on drones, characterized in that, It includes the following steps: S1. The drone acquires the inclined image of the engineering construction site, corrects the inclined image of the engineering construction site based on the bundle adjustment method, then uses the SURF algorithm to extract feature points and match the feature points. After obtaining the matched inclined image of the engineering construction site, a white film model is generated, and a real scene model of the engineering construction site is established; S2. Extract the point cloud of the engineering construction site according to the real scene model of the engineering construction site, cluster the point cloud of the engineering construction site based on the improved clustering algorithm, extract the point cloud set of the buildings on the construction site, obtain the engineering construction progress model, calculate the progress deviation for deviation analysis, and adjust the engineering construction progress; S3. Set the engineering construction progress plan, combine the progress deviation, train the BP neural network, obtain the BP neural network model, and predict the completion rate of the engineering construction; S4. Combine the completion rate of the engineering construction, and adjust and optimize the engineering construction progress plan. By finding the optimal fitness function value, obtain the adjusted engineering construction progress plan to achieve the control of the engineering construction progress; The S1 includes the following steps: S11. The drone acquires the initial inclined image of the engineering construction site, and corrects the inclined image of the engineering construction site using the bundle adjustment method to obtain the processed inclined image of the engineering construction site; S12. Generate the processed inclined matrix of the engineering construction site according to the processed inclined image of the engineering construction site; use the SURF algorithm to detect and extract feature points from the processed inclined matrix of the engineering construction site, complete the feature point matching, and obtain the matched inclined matrix of the engineering construction site; S13. Obtain the matched inclined image of the engineering construction site according to the matched inclined matrix of the engineering construction site, generate ultra-high density point cloud data, form a dense triangular network, and then generate a white film model; establish a real scene model of the engineering construction site based on the white film model using texture mapping.

2. The method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 1, wherein The use of the SURF algorithm to detect and extract feature points from the processed inclined matrix of the engineering construction site and complete the feature point matching includes the following steps: Select pixel points in the processed inclined matrix of the engineering construction site, and construct a Hessian matrix based on the Gaussian second-order differential convolution to determine the position of the feature points in the Hessian matrix, and then locate the feature points to the processed inclined matrix of the engineering construction site; Select a sub-matrix with the feature point as the center, calculate the direction of the feature point. When the signs of the traces of the Hessian matrices of two feature points are the same in the direction of the feature point, the corresponding two feature points are matched; match all pixel points in the processed inclined matrix of the engineering construction site in turn to obtain the matched inclined matrix of the engineering construction site.

3. The method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 2, wherein The S2 includes the following steps: S21. Extract the point cloud of the engineering construction site according to the real scene model of the engineering construction site, and cluster the point cloud of the engineering construction site based on the improved clustering algorithm to obtain the final clustered point cloud set; S22. Set the building threshold of the construction site, compare it with the final clustered point cloud set, extract the building point cloud set of the construction site, and fuse the building point cloud set of the construction site with the actual scene model of the construction site to obtain the construction progress model; calculate the progress deviation by comparing the construction progress model, conduct deviation analysis, and adjust the construction progress.

4. A method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 3, characterized in that, The adjustment of the construction progress includes the following steps: When the progress deviation is less than 0, at this time the construction progress lags behind the construction progress plan and the construction period cannot end on schedule, then adjust the construction progress; when the progress deviation is greater than 0, at this time the construction progress is ahead of the construction progress plan to ensure the construction quality; when the progress deviation is equal to 0, then do not adjust the construction progress; complete the adjustment of the construction progress.

5. The method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 4, wherein, The said S3 includes the following steps: S31. Obtain the number of construction workers, the construction level of the project, and the daily progress deviation according to the construction progress plan, and form a construction progress data set; Obtain the construction progress plan of previous years, calculate the number of construction workers in previous years, the construction level of previous years, and the daily progress deviation in previous years, form a construction sample data set, and divide the construction sample data set into a construction training set and a construction test set; set the number of neurons in the input layer of the BP neural network to 3, the number of neurons in the hidden layer to 9, the number of neurons in the output layer to 1, the learning rate to χ, and the activation function to the sigmoid function. Input the construction training set into the BP neural network, set the current iteration number to a″, and the maximum iteration number to A″. When the current iteration number reaches the maximum iteration number, stop the iteration to obtain a trained BP neural network; then input the construction test set into the trained BP neural network, set the error threshold to ξ. When the error of the output result is less than the error threshold, obtain the BP neural network model, otherwise adjust the weights until the error of the output result is less than the error threshold. S32. Input the construction progress data set into the BP neural network model, and the BP neural network model outputs a predicted value, which is the construction completion rate of the project.

6. A method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 5, characterized in that, The said S4 includes the following steps: S41. When the construction completion rate is less than 1, adjust and optimize the construction progress plan based on the golden jackal optimization algorithm to obtain an adjusted construction progress plan; S42. Control the daily construction resource quantity, the number of daily construction workers, and the daily construction funds according to the adjusted construction progress plan to achieve construction progress control.

7. A method for regulating construction progress in engineering construction based on an unmanned aerial vehicle according to claim 6, characterized in that, The said S41 includes the following steps: S411. Assign weights to the daily construction resource quantity, the number of daily construction workers, and the daily construction funds, construct a construction influencing factor function, and use the construction influencing factor function as the fitness function; set the number of golden jackal populations to i′, and i′ also represents the number of construction days. Each golden jackal in the golden jackal population has a dimension of j′, and an initial matrix is established. Each golden jackal individual in the initialization matrix corresponds to a fitness function value; it is set that the upper limit of the golden jackal position in the golden jackal population is c′, the lower limit of the golden jackal position in the golden jackal population is c″, g1 represents a random number and g1 ∈ [0, 1], then the calculation formula for the initial position D0 of the golden jackal in the golden jackal population is as follows: D0 = c″ + g1·(c′ - c″); S412. In the exploration stage of the golden jackal population, set the current iteration number as h, the maximum iteration number as H, g2 represents a random number and g2 ∈ [0, 1], the prey escape energy is δ, and the calculation formula for the prey escape energy is as follows: Among them, δ1 represents the original energy of the prey, and δ2 represents the decreasing energy of the prey; Introduce Levy flight, set Levy to represent the fitness function value of Levy flight, then any vector ε of the Levy distribution is ε = 0.05·Levy, the prey position vector at the h-th iteration is φ(h), the position of the male golden jackal at the h-th iteration in the golden jackal population is D′(h), the position of the female golden jackal at the h-th iteration in the golden jackal population is D″(h), and the positions of the male golden jackal and the female golden jackal at the h-th iteration in the golden jackal population are updated, denoted as D1(h) and D2(h) respectively, and the calculation formulas are as follows: D1(h) = D′(h) - δ·|D′(h) - ε·φ(h)|, D2(h) = D″(h) - δ·|D″(h) - ε·φ(h)|; The position of the golden jackal in the (h + 1)-th iteration of the golden jackal population By iterating the positions of the golden jackals, the fitness function values of the golden jackal individuals in the current golden jackal population are obtained, and the optimal fitness function value is continuously searched for; S413. In the attack stage of the golden jackal population, the golden jackals hunt, and the positions of the male golden jackal and the female golden jackal at the h-th iteration in the golden jackal population are updated again, and the calculation formulas are as follows: D1(h) = D′(h) - δ·|ε·D′(h) - φ(h)|, D2(h) = D″(h) - δ·|ε·D″(h) - φ(h)|; According to the positions of the male golden jackal and the female golden jackal at the h-th iteration in the golden jackal population and their updated positions again, the positions of the golden jackals at the (h + 1)-th iteration in the golden jackal population are updated, and iterate continuously. When the current iteration number reaches the maximum iteration number, stop iterating to obtain the final position of the golden jackal. The three-dimensional coordinate values of the final position of the golden jackal respectively correspond to the daily engineering construction resources, the daily number of construction workers, and the daily engineering construction funds, and the adjusted engineering construction progress plan is obtained in sequence.

8. A system for implementing the construction progress regulation method in engineering construction based on unmanned aerial vehicles according to any one of claims 1-7, characterized in that, Specifically, it includes: An engineering construction site real-scene model establishment module, a construction site building point cloud set extraction module, an engineering construction completion rate prediction module, and an engineering construction progress control module; The engineering construction site real-scene model establishment module is used to construct an engineering construction site real-scene model based on the bundle adjustment method and the 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 an engineering construction progress model; The engineering construction completion rate prediction module is used to predict the engineering construction completion rate by a 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